Ask the Scholar
Document scope · 1 page
Scholar
Ask about this object, its catalog metadata, its source description, or the page inventory.
For page-specific OCR and visual context, open one of the page chats.
Scholar Source Context
Document identity
localId
225358539
label
IPCC [Intergovernmental Panel on Climate Change] - Volume II [1]
core
doc
dtoType
document
citationUrl
pageCount
1
Source metadata
id
225358539
contentType
document
title
IPCC [Intergovernmental Panel on Climate Change] - Volume II [1]
citationUrl
collections
Records of the Council of Economic Advisers (Clinton Administration)
Joseph Stiglitz's Files
imageCount
1
hasImages
yes
source
import
hasTranscription
no
Source extras
naId
225358539
levelOfDescription
fileUnit
otherTitles
42-t-4739354-20171095F-025-001-2019
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
8de462316762a268
ocrText
FOIA Number: 2017-1095-F
FOIA
MARKER
This is not a textual record. This is used as an
administrative marker by the William J. Clinton
Presidential Library Staff.
Collection/Record Group:
Clinton Presidential Records
Subgroup/Office of Origin:
Council of Economic Advisers
Series/Staff Member:
Joseph Stiglitz
Subseries:
OA/ID Number:
9569
FolderID:
Folder Title:
IPCC [Intergovernmental Panel on Climate Change] - Volume II [1]
Stack:
Row:
Section:
Shelf:
Position:
S
20
5
9
3
Chapter 4
Intertemporal Equity and Discounting
Lead Authors:
Kenneth J. Arrow, William R. Cline, Karl-Goran Maler,
Mohan Munasinghe, and Joseph E. Stiglitz
January, 1995
TABLE OF CONTENTS
ii
SUMMARY
1
4.1
Introduction
4.1.1 Importance of the Discount Rate
1
4.1.2 Toward an Emerging Consensus
2
3
4.2 Intergenerational Equity
4.3 Discounting and the Social Rate of Time Preference
5
4.3.1 Formulation of Prescriptive Approach
6
4.3.2 The Pure Rate of Time Preference Across Generations
7
4.3.3 Discounting Utility for Human Extinction
8
4.3.4 Discounting Utility for Acceptable Optimization Results
9
4.3.5 Discounting Utility for Empathic Distance
10
4.3.6 Discounting for Rising Levels of Consumption and Utility
11
4.3.7 Relationship between the Social Rate of Time Preference and Market
Rates of Interest
12
4.3.8 Discount Rate Estimates -- Prescriptive Approach
13
4.4 Formulation of Descriptive Approach
14
4.4.1 Returns to Investment
14
4.5 Discounting for Risk
17
4.6 Producer or Consumer Interest Rates?
17
4.7 Discounting Environmental Impacts
19
4.8 Implications for Sustainable Development
20
Annex 4A
23
Endnotes
27
References
28
1
SUMMARY
2
3
The methodology for comparing effects at different points in time, and especially establishing
4
trade-offs across different generations is a matter of philosophy (and for some, religion and morals)
5
because it involves reaching judgements about what is fair or just. The issue is also a matter of
6
economics, because comparisons across time are appropriately judged in the light of changing
7
standards of living over time as well as opportunities for productive investment.
8
9
Intergenerational Equity
10
11
Climate policy raises particular questions of equity among generations, because future generations
12
are not able to influence directly the policies being chosen today that will affect their well-being and
13
because it might conceivably not be possible to compensate future generations for reductions in
14
well-being that might be caused by current policies. A proper formal treatment of fairness among
15
generations poses special challenges to economic science.
16
17
The report of the Commission on Environment and Development called for "sustainable
18
development," defined as economic activity that "meets the needs of the present without
19
compromising the ability of future generations to meet their own needs." Similarly, Brown-Weiss
20
(1989, 25) has argued from the standpoint of international law that "each generation is entitled to
21
inherit a planet and cultural resource base at least as good as that of previous generations."
22
23
Among economists the standard approach has been to summarize views about interpersonal equity
24
in the form of a social welfare function. Arguments then turn on the ability to derive a particular
25
social welfare function from sound theoretical principles, and about the resulting reasonableness of
26
its derived implications.
27
28
Discount Rate
29
30
The discount rate is the analytical tool economists use to compare economic effects that occur at
31
different points in time. Identifying the proper discount rate is probably the single most important
32
analytical step in economic analysis of climate change. It is also the most profound ethical question,
33
since it inherently confronts present pain of abatement cost against damages suffered by future
34
generations if no action is taken.
35
36
The literature on the appropriate discount rate for climate change analysis can be divided into two
37
approaches. The prescriptive approach asks whether a particular program would make society better
38
off, given a plausible social welfare function. The discounting rules specify how consumption at
39
different dates should be valued to provide the correct answer to that question. The descriptive
40
approach focuses on intertemporal efficiency and opportunity costs.
41
42
Prescriptive Approach
43
44
The prescriptive approach discounts consumption accruing to different generations using the "social
45
rate of time preference" (SRTP):
46
47
SRTP = P + Θg
ii
where p is the rate of "pure" time preference, Θ is the absolute value of the elasticity of marginal
utility, and g is the growth rate of per capita consumption.
3
4
The first term (p), the rate of pure time preference, is the rate at which utility is discounted because
5
it occurs later rather than sooner. The second term (Θg) is the rate at which changes in consumption
6
levels are discounted to be translated into the resulting changes in utility or welfare levels. If per
7
capita consumption is growing a rate g, then an extra unit of consumption in the future should be
8
discounted by the term Θg to take account of the lower marginal to utility of consumption at higher
9
consumption levels. This approach takes into account opportunity costs in calculating changes in
10
consumption, but in valuing the changes in consumption, uses the SRTP.
11
12
With high rates of productivity increase (and high g), of the order of 1.5%, and high values of the
13
elasticity of marginal utility (Θ), the social discount rate is of the order of 3%; with low rates of
14
productivity increase, of the order of .5%, and low values of the elasticity of marginal utility, the
15
social discount rate is of the order of .5%. There is no reason why the discount rate should be
16
constant over time even if P and Θ are constant, since g need not be constant. Developing countries,
17
with high rates of per capita income growth (5 to 8% per year in some countries), and elasticities of
18
marginal utility of 2, could justify discount rates of the order of 10% to 16%. Countries with per
19
capita incomes close to subsistence levels could have high elasticities of marginal utility (rapid
20
dropoff of marginal utility from initially extremely high levels associated with privation), and so
21
justify high discount rates even if they were experiencing slow growth of per capita income.
22
23
Descriptive Approach
24
25
The descriptive approach to the discount rate looks at returns to investments in the real world, and
26
with this information seeks intertemporal efficiency. It asks: could consumption be increased at one
27
date without decreasing it at any other?
28
Market rates of return usually exceed the SRTP. Conceptually, funds could then be invested in
29
projects that provide a higher return with the proceeds being used to increase future consumption. In
30
the context of climate change, however, society is not likely to be able to set aside investments over
31
the next three centuries and earmark the proceeds the eventual compensation of those adversely
32
affected.
33
34
35
iii
1
Chapter 4
2
Intertemporal Equity and Discounting
3
4
5
4.1
Introduction
6
7
This chapter concerns the methodology for comparing effects at different points in time,
8
and especially establishing trade-offs across different generations. How we think of these
9
trade-offs involves issues of intertemporal equity. This issue is a matter of philosophy (and
10
for some, religion and morals) because it involves reaching judgements about what is fair or
11
just. The issue is also a matter of economics, because comparisons across time are
12
appropriately judged in the light of changing standards of living over time as well as
13
opportunities for productive investment. To reflect the balance of issues, an alternative
14
chapter title would be Intertemporal Equity, Discounting, and Economic Efficiency.
15
16
4.1.1 Importance of the Discount Rate
17
18
The discount rate is the analytical tool economists use to compare economic effects that
19
occur at different points in time. Identifying the proper discount rate is probably the single
20
most important analytical step in economic analysis of climate change. It is also the most
21
profound ethical question, since it inherently confronts present pain of abatement cost
22
against damages suffered by future generations if no action is taken.
23
24
The issue of choosing an appropriate discount rate has been discussed in the context of
25
general cost-benefit analysis (chapter 5) for many years (Dasgupta et al. 1972, Harberger
26
1976, Little and Mirrlees 1974, Sen 1967, Stiglitz 1982). The long term perspective
27
required for sustainable development suggests that the discount rate might play a critical
28
role in intertemporal decisions concerning the use of environmental resources.
29
30
The emission of carbon dioxide causes effects likely to persist on a time scale of centuries.
31
IPCC (1990a) estimates that the first half-life of CO2 is 50 years and the second half-life
32
another 250 years (with the decline to 1/e taking 120 years). In the absence of policy
33
intervention carbon dioxide concentrations could continue to rise for the next 300 years,
34
given current fossil fuel resources (Sundquist, 1990). Climate change thus poses unusual
35
problems for economic analysis, which more typically considers issues with time scales of
36
only a few decades at most.
37
38
The usual formulation for discounting over time involves exponential growth of the
39
discount (or value-shrinkage) factor. Thus, one unit of consumption that arrives t years from
40
today is equivalent to the fraction l/(1+i)' of one unit of consumption available today,
41
discounting at rate i.
42
43
Social scientists have debated the precise rate to use for global climate analysis (Broome,
44
1992; Cline, 1992; Nordhaus, 1993). There is general agreement that the choice of the
1
1
discount rate has extremely powerful effects on the economic evaluation of policy on
2
climate change because of the unusually long time horizons involved (e.g. Schmalensee,
3
1993). With benefits of greenhouse abatement accruing decades or even centuries into the
4
future, if a high discount rate is used, the benefits get heavily discounted relative to the
5
expenditures incurred today in trying to mitigate climate change.
6
7
For example, rates of 8% - 10% per annum are often used in project analysis. Such rates
8
cause the future to vanish rapidly. Thus, discounting at 8%, damage of $1 billion 50 years
9
from now is valued at only [$lx109]/[1.08⁵⁰] = $21.3 million; and the same damage 200
10
years from now, at only $200. This collapse of future values suggests the need to think
11
especially carefully about proper discounting for climate change analysis.
12
13
14
4.1.2 Toward an Emerging Consensus
15
16
The framework of an emerging consensus on discounting for climate change analysis would
17
include the following.
18
19
First, although economists have disagreed about (a) whether future generations at levels of
20
welfare or utility equal to those of today should nonetheless be discounted relative to the
21
present, and (b) how precisely to relate changes in the discount factor to changes in income,
22
there is widespread (but not universal) agreement about the following propositions:
23
24
(i) Rejection of environmentally specific egalitarianism: future generations can be
25
compensated for some environmental/natural resource effects by offsetting
26
accumulations of capital.
27
28
(ii) Acceptance of more generalized (but not, usually, absolute) egalitarianism:
29
differences in the overall standard of living should be reflected in the discount rate.
30
There is, in fact, remarkable consensus in the quantitative magnitudes by which
31
future discounting should be affected by rising per capita income. Given the best
32
available forecasts, expected increases in the standard of living should cause
33
discounting of changes in consumption of future generations on the order of 1/2 to 3
34
percent annually.
35
36
Second, trade-offs of consumption across generations, while affected by observed market
37
(producer) rates of interest, through the effect of market rates on the value of foregone
38
investment alternatives, are not in general equal to these market rates.
39
40
Third, trade-offs across generations in environmental goods (or public goods) need not
41
correspond to trade-offs across generations in private consumption goods. The standard
42
economist's procedure for dealing with this problem is to convert other goods into their
43
equivalent in private consumption goods, the apply the discount rate to the equivalents.
44
This procedure incorporates market rates of return, by including as a cost of a project the
2
1
foregone benefits of possible investments not undertaken.
2
3
Fourth, issues of discounting involve the relative values of goods at different dates, and
4
thus intertemporal pricing. Issues of pricing of risk need to be carefully separated out from
5
those involving intertemporal pricing. The standard procedure for conventional models,
6
involving impacts over a single individual's life, involves converting probabilistic
7
consumption patterns into their certainty equivalents, and then discounting at the social rate
8
of time preference -- not the incorporation of an incremental component into the discount
9
rate itself. This procedure does not, however, deal adequately with uncertainties about
10
future rates of growth of per capita income. Such uncertainties are much greater over
11
centuries-scale horizons.
12
13
Fifth, issues of equity can be treated in a way analogous to those of risk, through the use of
14
certainty equivalents (Atkinson 1970, Rothschild and Stiglitz 1974), though the likely effect
15
on the appropriate rate of discount have yet to be thoroughly studied.
16
17
4.2 Intergenerational Equity
18
19
Climate policy raises particular questions of equity among generations, because future
20
generations are not able to influence directly the policies being chosen today that will affect
21
their well-being (Mishan, 1975; Broome, 1992), and because it might conceivably not be
22
possible to compensate future generations for reductions in well-being that might be caused
23
by current policies.
24
25
The concern for fairness to future generations has long undergirded the environmental
26
movement. This concern spans a broad range of political, social, and ethical considerations
27
and viewpoints; a proper formal treatment poses special challenges to economic science.
28
Perhaps best known among recent formulations, the report of the Commission on
29
Environment and Development (United Nations, 1987), led by Norwegian Prime Minister
30
Gro Harlem Brundtland, called for "sustainable development," defined as economic activity
31
that "meets the needs of the present without compromising the ability of future generations
32
to meet their own needs" (United Nations, 1987; for a discussion of sustainable
33
development, see sections 1.6 above and 3.6 below). Similarly, Brown-Weiss (1989, 25) has
34
argued from the standpoint of international law that "each generation is entitled to inherit a
35
planet and cultural resource base at least as good as that of previous generations."
36
37
A consensus exists among economists that this does not imply that future generations
38
should inherit a world with at least as much of every resource; such a view would preclude
39
consuming any exhaustible natural resource. The common interpretation is that if the stock
40
of some natural resource be depleted, it be compensated by an increase in the stock of
41
capital (physical or human). Under most calculations, given the savings rates of all but the
42
lowest-saving countries in the world, most countries now pass this test of sustainability.
43
Discussions of intergenerational equity must thus go beyond a focus on sustainable
44
development.
3
1
Economists have long debated the equity of discounting distant future benefits (Ramsey,
2
1928; Mishan, 1975; Rawls, 1971; Sen, 1982). The standard approach to issues of equity
3
since Bergson (1938) has been to summarize views about interpersonal equity in the form
4
of a social welfare function, an algebraic formulation relating welfare to levels of
5
consumption of the society's members at a given point in time and across time. Arguments
6
about the choice among alternative social welfare functions then turn on the ability to
7
derive a particular function from sound theoretical principles (seemingly plausible axioms),
8
and about the resulting reasonableness of its derived implications.
9
10
While all social welfare functions have been criticized for assuming interpersonal
11
comparability of utility, there seems to be no way of addressing the ethical issues involved
12
in making decisions affecting different generations without making some assumptions
13
implicitly or explicitly about interpersonal comparability. Two polar views are represented
14
by the utilitarian approach, in which social welfare is the sum of utilities; and the Rawlsian
15
approach, in which social welfare reflects the welfare of the worst-off individuals. While
16
the utilitarian approach can be derived from what many view as a persuasive axiomatic
17
(theoretical) structure (Harsanyi, 1955), the Rawlsian approach is derived from a "max-min"
18
strategy (maximize the minimum outcome for any given party) popular in game theory, but
19
which itself does not rest on widely accepted axiomatic principles.
20
21
The Rawlsian max-min
22
principle is the strongest
A Dissenting View on Discounting
23
in assuring (the least
24
fortunate groups of)
One prominent economist, Thomas Schelling, has rejected the
25
future generations levels
approach presented here, arguing that it is requires the
assumption of utility maximization by a single representative
26
of consumption at least as
individual who lives forever. Schelling emphasizes that the
27
great as that of (the least
essence of climate change is that those likely to bear the cost of
28
fortunate groups of) the
mitigation--the developed countries-differ from those likely to
29
current generation. It is
enjoy the benefits-the currently developing countries. Thus he
30
consistent with the
emphasizes the importance of distributional considerations.
31
Brown-Weiss (1989)
Supporters of the approach set out in this chapter note that the
32
approach noted above.
model of one immortal individual is used only for exposition, and
33
The max-min criterion
that the results derived from the social welfare function approach
34
permits inequality in
presented here remain true even with many individuals (or
35
consumption between
countries) within each generation.
36
individuals (or in this
37
case, between
38
generations) only if it improves the position of the poorest. (For the remainder of this
39
chapter, we shall simplify by ignoring inequality within a generation.) In the absence of
40
technical change this would imply that consumption per head should be the same for all
41
generations. By contrast, the utilitarian criterion allows, in principle, future consumption to
42
fall below current consumption, provided the current generation is made sufficiently better
43
off as a result. Correspondingly, it also allows for decreases in present consumption,
44
provided the future generation is made sufficiently better off as a result.
4
1
The Rawls and utilitarian social welfare functions can be viewed as limiting cases of more
2
general social welfare functions embracing social values of equality (Atkinson, 1971;
3
Rothschild. and Stiglitz, 1973). In practice, so long as there is sufficient scope for
4
technological change, optimizing any egalitarian social welfare function over time yields
5
increases in consumption per capita. Moreover, with any of the approaches, earlier
6
generations are entitled to draw down the pool of exhaustible resources so long as they add
7
to the stock of reproducible capital.
8
9
The individualistic social welfare function, accepted by most economists as the basis of
10
ethical judgments, accepts individuals' own relative valuations of different goods. It does
11
not place separate valuations on unequal access to particular goods, other than through their
12
effects on the affected individuals. For example, Thurow (date) argued that income
13
distribution is a public good, but that its value could be captured in individuals' valuations.
14
Although this represents the consensus view, some economists have insisted that for
15
particular goods, individuals' valuations need not be the basis of societal valuations. For
16
instance, Tobin (1970), in what he called specific egalitarianism, argued that society might
17
argue for greater equality in distribution of health care than would be reflected in
18
individuals' own evaluations. Most economists, however, reject this view.
19
20
Sen (1982) similarly suggests a basis for not discounting when environmental effects are
21
in question. He argues that a fundamental right of the future generation may be violated
22
when the environment is degraded by the present generation, and that the resulting
23
"oppression" of the future generation is inappropriate even if that generation is richer than
24
the present and has a lower marginal utility of consumption. In this framework,
25
intertemporal equity for environmental questions requires "a rejection of 'welfarism,' which
26
judges social states exclusively by their personal welfare characteristics."
27
28
4.3 Discounting and the Social Rate of Time Preference
29
30
The approach described in this section is sometimes called the prescriptive approach. It
31
asks whether a particular program would make society better off, given a plausible social
32
welfare function. The discounting rules specify how consumption at different dates should
33
be valued to provide the correct answer to that question. The alternative view, described in
34
Section 4.4, focuses on intertemporal efficiency and opportunity costs. To distinguish it
35
from the prescriptive approach, it is sometimes referred to as the descriptive approach.
36
37
The question of the appropriate discount rate involves issues in normative as well as
38
positive economics. How ethically should we value impacts on future generations? And to
39
what extent, for instance, will investments made to reduce greenhouse gas emissions
40
displace investments elsewhere? The debate is often confused both because the ethical and
41
efficiency aspects are intermingled, and because there are, in fact, three separate issues that
42
have to be addressed: How do we discount the welfare or utility of future generations?
43
How do we discount future dollars? And how do we discount future pollution? The
44
following discussion summarizes some of the salient issues in the debate, reaching the
5
1
following conclusions:
2
3
1. Total impacts on consumption (or welfare) across generations: the social rate of time
4
preference, a reflection of societal views concerning trade-offs of consumption across
5
generations, should be employed.
6
7
2. Direct costs and benefits of a greenhouse mitigation project: Suppose that a mitigation
8
project would displace private investment, and that returns to both projects accrue to the
9
same generations. Then it is appropriate to use the opportunity cost of capital - the private
10
return - in discounting.
11
12
Investments in both physical capital (e.g. machines) and human capital (e.g. education)
13
yield on average a positive real return. That is, money invested today can be transformed
14
into more money later, even after adjusting for inflation. Discounting converts each future
15
dollar amount associated with the project into the equivalent present dollar amount that
16
must be invested today in the general economy in order to yield the same future amount.
17
18
Greenhouse gas emission control may be viewed as an investment: money is spent today on
19
emission controls today so that more money (or other benefits) may be received later in the
20
form of reduced economic costs of climate change. If the real rate of return on investment
21
in emission reduction exceeds the rate on investment in machines and education, then future
22
generations would be better off if less were invested today in machines and education and
23
more in controlling greenhouse gas emissions; the converse also holds.
24
25
3. Competing investments: The cost of a greenhouse mitigation project must include the
26
foregone benefits of other competing investments not undertaken. Only then will it be
27
appropriate to use the social rate of time preference.
28
29
4. Evaluating risk: uncertainty about the impacts (changes in consumption) can be
30
incorporated by analyzing their certainty equivalents (the certain result that would make an
31
individual indifferent between it and the uncertain outcome), and discounting the certainty
32
equivalents in the manner described above.
33
34
5. Evaluating environmental impacts: environmental impacts can be incorporated in an
35
analogous way--converting them to consumption equivalents, and discounting at the rate
36
described above. Under plausible assumptions, the relative price of environmental goods
37
will increase over time, which would have consequences equivalent to adopting a lower
38
discount rate for such goods at unchanged prices.
39
40
4.3.1 Formulation of Prescriptive Approach
41
42
The first issue in the analysis of the discount rate is the question of how we value
43
consumption accruing to different generations. Within the predominant utilitarian (as
44
opposed to Rawlsian) framework, with standard simplifications¹, the fundamental equation
6
1
for the discount rate of consumption, or the "social rate of time preference" (SRTP) is:
2
3
4.1) SRTP = p + Θg
4
5
where p is the rate of "pure" time preference, Θ is the absolute value of the elasticity of
6
marginal utility, and g is the growth rate of per capita consumption.
7
8
The first term (p) is the rate at which utility is discounted because it occurs later rather than
9
sooner. Sometimes this term, the rate of pure time preference, is referred to as discounting
10
for impatience or myopia. The second term (Θg) is the rate at which changes in
11
consumption levels are discounted to be translated into the resulting changes in utility or
12
welfare levels. The idea is that if per capita consumption is growing a rate g, then an extra
13
unit of consumption in the future should be discounted by the term Θg to take account of
14
the lower marginal to utility of consumption at higher consumption levels.
15
16
Annex 4A sets out the mathematics of discounting. Along an optimal path, with no
17
distortions, the marginal product of capital (equal to the market rate of interest) and the
18
discount rate on consumption both equal p + Θg. It is important to emphasize that in using
19
the SRTP approach, we value the total change in consumption at each date, not just the
20
direct outputs of the project. If the project displaces private investment that has a high
21
marginal productivity, then there will be a large loss in output and future consumption
22
resulting from the displacement. The prescriptive view takes into account opportunity costs
23
in calculating changes in consumption, but in valuing the changes in consumption, uses the
24
SRTP. In short, all effects are converted to their consumption equivalents (as discussed in
25
section 4.6), then discounted at the SRTP.
26
27
4.3.2 The Pure Rate of Time Preference Across Generations
28
29
Consider pure time preference first. In a society in which income levels are not expected to
30
rise, impatience may still cause a household (or the present generation) to discount the
31
future (generation), that is, to equate a smaller amount of consumption today with a larger
32
amount in the future. In his classic paper on optimal saving, Ramsey (1928, p. 543) judged
33
that any allowance for pure time preference (p> 0) "is ethically indefensible and arises
34
merely from the weakness of the imagination." Correspondingly, he argued that future
35
generations should have equal standing with the current generation; there was no moral
36
or ethical basis for weighing the welfare of future generations less than that of the current
37
generation.
38
39
For an individual, some non-zero value of pure time preference can make sense, because he
40
or she has a finite life and thus uncertainty about being alive to enjoy future consumption.
41
Nonetheless, for a life span of 70 years, pure time preference at even 1 percent per annum
42
implies that consumption at the end of life is worth only half that at the beginning.
43
Evidence also suggests that individuals' discount rates can change over time, with lower
44
discount rates being used for longer time horizons (Thaler and Lowenstein 1989).
7
1
Considerations for society as a whole are different. The approach described earlier asks: if
2
society values different generations in a particular way (the social welfare function), how
3
should changes in consumption in different generations be compared? How society values
4
different generations can be viewed then from two different perspectives: 1) How should
5
society value different generations (an ethical approach); or 2) How does the current
6
generation value the consumption of future generations. Ramsey's analysis focused on the
7
ethical presumption that consumption by all generations should have equal value. But this
8
does not exclude the possibility that as a matter of description the current generation gives
9
less value to consumption of future generations.
10
11
4.3.3 Discounting Utility for Human Extinction
12
13
One argument focuses on the fact that there is no absolute certainty that future generations
14
will exist since any number of catastrophic events could occur. This argument, which
15
might be called the "asteroid effect," might argue that it would make no sense to give up
16
consumption today for the generation 200 years from now because it may not come into
17
existence. Some economists believe it is appropriate for this reason to discount future
18
consumption to reflect the inherent uncertainty in future conditions. However, the
19
probability of human extinction would seem sufficiently small (especially within a time
20
frame of two to three centuries) that the quantitative magnitude of discounting for this
21
purpose is likely to be extremely small.
22
23
Thus, according to some estimates the probability that an individual will die because an
24
asteroid hits the earth during his or her lifespan is of the same order of magnitude as the
25
probability that he or she will die in an airplane crash (about one in ten thousand). The
26
chances of earth's collision with an asteroid large enough to cause species extinction (such
27
as the 10-km-diameter object believed by some to have caused the Cretaceous/Tertiary
28
dinosaur extinction) is far smaller (Wetherill and Shoemaker, 1982). However, to be
29
generous to the extinction argument and for purposes of illustration, consider the individual
30
risk at one in 10,000. The three-century horizon appropriate for climate change analysis is
31
about four life spans, so set the probability of asteroid impact over this period at 1/2,500,
32
for purposes of illustration. Then at the end of this period, a given economic effect should
33
be discounted enough to shrink its value by one- 2,500th. The discount rate required for
34
this shrinkage is vanishingly small (= 0.133x10⁴ percent per annum).
35
36
Perhaps one reason the optimal growth literature includes discounting for extinction is that
37
most of the literature emerged during the cold war, when nuclear annihilation was more
38
plausible than today. For public policy purposes, even under those conditions it is
39
arguably inappropriate to incorporate analytical assumptions that take as a premise the
40
failure of policy (breakdown into nuclear war). In any event this motive for
41
extinction-based discounting would seem even less appropriate for the future.
42
43
44
8
1
4.3.4 Discounting Utility for Acceptable Optimization Results
2
3
Another argument for non-zero pure time preference is that setting the rate at zero could
4
imply that the present generation should accept near-starvation consumption levels, and
5
correspondingly low utility, because with even very small returns on investment, an endless
6
stream of future generations could enjoy increased consumption and (to a lesser degree)
7
utility as a result.
8
9
To some extent, however, this concern is already addressed in the overall discount rate
10
equation (4.1). As noted, the first term in that equation discounts utility (pure time
11
preference), but the second term additionally discounts consumption to take account of
12
falling marginal utility. The present generation is protected against an optimizing program
13
setting its consumption near zero if the term for elasticity of marginal utility (Θ) is large
14
enough and marginal utility drops off fast enough to rule out impoverishment of the present
15
generation for gains to future generations.
16
17
More fundamentally, a basic concern about zero pure time preference has to do with the
18
mathematics of maximization over an infinite time horizon. If the utility function has no
19
upper limit, any savings-investment optimization problem is not well defined, because the
20
sum of utilities over time is infinite. There is some literature (Weizsacker 1965) proposing
21
criteria (the "overtaking criterion") to address this problem, but the implications of these
22
criteria for climate policy have not been examined.
23
24
Even when the optimization problem is made well-defined, for the purposes here, e.g., if
25
there is an upper bound to utility (as in the function in annex equation 4A.3), there is a
26
tendency toward knife-edge results, with one extreme outcome for zero pure time
27
preference and the opposite extreme for non-zero rates. Thus, whereas the concern of
28
Koopmans (1965), Mirlees (1974), and Chakravarty (1969) is that zero pure time
29
preference can suppress current consumption to unacceptably low levels, just the opposite
30
result can be reached in optimization models that take exhaustible resources into account. In
31
some specifications of such models (Dasgupta and Heal, 1974; Solow, 1974a), any pure
32
time preference rate in excess of zero generates the unacceptable result that optimal
33
consumption falls to zero over the very-long term (although Stiglitz, 1974, shows that this
34
awkward conclusion is not robust with respect to alternative assumptions about
35
technological change, production functions, and utility functions).
36
37
Knife-edge results can of course result from other properties of utility maximization models
38
as well (depending, for example, on whether the production function has non-unitary
39
elasticity of substitution between factors, or whether capital-augmenting technical change is
40
zero or positive). Focusing on the mathematical and knife-edge complications of
41
optimization approaches, they do not go far toward specifying the proper magnitude for the
42
rate of pure time preference, even though they might be seen as providing a set of
43
arguments that the rate is greater than zero. Thus, considering the infinite horizon in such
44
models, an infinitesimally small but non-zero rate of pure time preference could suffice to
9
1
avoid prejudicing the optimal consumption path against the present generation, depending
2
on other assumptions of the model.
3
4
A stronger attack on very low discount rates is provided by those models concluding that
5
low discount rates imply unreasonably high savings rates, particularly in poor economies.
6
Only by raising the discount rate to higher, "more reasonable" numbers can savings rates of
7
the kind actually observed be obtained (e.g. Mirlees, 1967; Chakravarty, 1969). This
8
illustrates a general problem with models founded on utilitarianism: they may imply very
9
large sacrifices from one generation or other group. This argument also assumes that the
10
model captures accurately the structure of the economy; typically, however, these models
11
imply much higher rates of return on capital than are in fact observed in LDCs. For
12
example, Chakravarty (1969) assumes constant return on capital at 33 percent. Instead, the
13
small differences between advanced and developing countries in observed rates of return in
14
spite of large differences in capital-labor ratios suggests either that the economies are not
15
on the same production function, or that the constant returns to scale production function
16
models employed are inappropriate. (See Stiglitz, 1988, Lucas, 1988). Manne (1994) uses
17
a standard growth model to examine the relation between discount rates and savings rates in
18
the context of developed economies. He finds that discount rates of one or two percent
19
imply an unrealistically rapid near-term increase in the rate of investment. Manne thus
20
concludes that a rate this low is grossly inconsistent with observed or plausibly anticipated
21
behavior.
22
23
Manne's analysis can, alternatively, be interpreted as showing that the intertemporal
24
equilibrium established by market economies differs markedly from that corresponding to
25
the solution of an intertemporal maximization problem based on a reasonable social welfare
26
function. Such results are consistent with standard life cycle models, without government
27
intervention. On the other hand, they are not consistent with dynastic models, in which
28
each generation incorporates into its own utility function the utility of future generations in
29
a manner exactly analogous to the way future generations are incorporated into the social
30
welfare function (Barro date). However, considerable evidence weighs against Barro's
31
hypothesis.
32
33
4.3.5 Discounting Utility for Empathic Distance
34
35
Rothenberg (1993) and Schelling (1993) have suggested that although non-zero pure time
36
preference might make sense for an initial two or three decades, beyond a certain future
37
point it makes no sense to apply further discounting of consumption for pure time
38
preference. Thus: "as the future recedes. single generations come to be perceived more
39
and more as homogeneous entities (Rothenberg). Similarly: "time may serve as a kind of
40
measure of distance Beyond certain distances there may be no further depreciation for
41
time, culture, geography, race, or kinship" (Schelling). A graph of the fraction of face
42
value accorded to each successive generation (for constant real consumption) would thus be
43
a series of declining, successively shallower steps that eventually reach a horizontal plateau.
44
A deep plateau signifies major discounting for empathic distance; a horizontal line
10
1
beginning and remaining at unity is the zero pure time preference rate across generations
2
recommended by Ramsey. Policy based on empathetic distance (a shelf lower than unity)
3
may be more defensible in a normative sense when the action is refraining from conferring
4
a windfall gain (as in penurious aid budgets) than when it involves the imposition of
5
windfall damage (as in climate change's effects on future generations).
6
7
4.3.6 Discounting for Rising Levels of Consumption and Utility
8
9
Even if future generations are given weight equal to that for current generations, so that
10
pure time preference (p) is zero, there could still be an important basis for discounting
11
future consumption: people are expected to be better off, so that an extra unit of
12
consumption would not be worth as much in the future as today. Thus, if technological
13
change continues to proceed at the rate at which it has over the last century, with
14
productivity and living standards doubling every thirty years or so, then marginal changes
15
in consumption of those future generations should be weighed much less. This
16
consideration is captured by the second right-hand term in equation 4.1.
17
18
There are two questions, corresponding to the two elements in the final term of the
19
equation. First, what are reasonable expectations concerning increases in per capita income
20
(growth rate g in the equation)? Second, how should intertemporal differences in expected
21
consumption per capita be translated into social weights, that is, marginal valuations of
22
dollars of future income. This second question refers to the parameter Θ, the elasticity of
23
marginal utility. This parameter essentially tells how rapidly the additional utility from an
24
extra unit of consumption drops off as consumption rises.
25
26
No consensus on the first question has emerged. The low levels of productivity increase
27
over the past two decades have raised the question: were the high rates of technological
28
change experienced in the decades following World War II an aberration? Or are the low
29
rates experienced from 1973 until 1993 an aberration?
30
31
While it is also the case that no consensus has emerged on the answer to the second
32
question, there is a generally accepted methodology for approaching the issue. The
33
evaluation of any individual's consumption can be summarized by a utility function of the
34
form U = U(c) where the parentheses indicate that U, utility, is a function of c, per capita
35
consumption. Marginal utility is positive (U'(c) > 0), but it declines as consumption rises
36
(U"(c) < 0). That is why if consumption of some future generation is higher, the marginal
37
valuation of its consumption will be lower. The question is, how much lower? Formally,
38
the answer is given by the elasticity of marginal utility (Θ) or: [dU'/U']/[dc/c].
39
40
Individuals in their day-to-day decision making reveal information about their perceptions
41
concerning their own utility functions, in at least two different contexts: behavior towards
42
risk and intertemporal allocation of consumption. In both contexts, there seems to be a
43
consensus on elasticities of marginal utilities in the range of 1 to 2, even though the
44
empirical studies require strong assumptions about the specific form of the utility function
11
1
(symmetric and time-separable).
2
3
Thus, one of the most commonly used utility functions, logarithmic utility, implies Θ =1,
4
meaning that if income rises by 1% the marginal utility of consumption falls by 1%.
5
Attempts to estimate this elasticity by Fellner (1967) and Scott (1989) both place it
6
somewhat higher, at 1.5; whereas recent estimates reviewed by Pearce and Ulph (1994)
7
place it in the vicinity of 0.8.
8
9
Just as the choice of the rate of pure time preference (p) has important implications for
10
intergenerational equity, as discussed above, so does the choice of the elasticity of marginal
11
utility. The more weight the society gives to intergenerational equity, the higher the value
12
of Θ. Thus, a value of, say, 3, would mean that it would require a 30 percent rise in the
13
next generation's per capita consumption to warrant a 10 percent reduction in that of the
14
present generation; or, under a bleaker outlook, that if the future generation is expected to
15
be poorer than the present, the present would be prepared to accept a 30 percent reduction
16
in consumption to secure a 10 percent increase in that of the future generation (so long as
17
the two relative consumption levels did not reverse). Even a unitary value is equity
18
oriented, however. When Θ= 1, a 10 percent reduction in the richer generation's income
19
will be an acceptable trade-off for a 10 percent increase in that of the poorer generation,
20
even though the absolute reduction of the one exceeds the absolute increase of the other
21
(because the absolute consumption base of the one is larger than that of the other)².
22
23
4.3.7 Relationship between the Social Rate of Time Preference and Market Rates of Interest
24
25
Economists have long recognized that a competitive market equilibrium yields a (Pareto)
26
efficient outcome, under appropriate conditions (perfect competition, no externalities, etc.).
27
The distribution of income that it yields, however, has no special merit; that is, it does not,
28
in general, maximize any particular social welfare function. It is a well recognized
29
function of government to intervene in the distribution of income, e.g. by establishing
30
programs for the very poor. By the same token, the intertemporal distribution of welfare
31
that emerges from the market will not, in general, maximize any particular social welfare
32
function discussed in the preceding section. While it is a legitimate function of government
33
to intervene to change the intergenerational distribution of welfare, there is no presumption
34
that the government has in fact intervened so that the observed resource allocations are
35
those that maximize intertemporal social welfare. Moreover, in the case of climate change,
36
no one government exists to make these decisions.
37
38
In general, the market rate of interest--the relative price of consumption of one generation
39
in one year of its life to consumption in another year--will not equal the SRTP. In standard
40
life cycle models, even with no technological progress and an economy in steady state,
41
there would be no discounting for society's purposes -- each generation is identical, so the
42
marginal utility of consumption of each is the same. Nonetheless, the market rate of
43
interest will be positive in any efficient equilibrium under certain reasonable assumptions
44
about utility functions (such as individual impatience and zero bequest motive; Diamond,
12
1
1965). In such models the market rate of interest would thus always overestimate the SRTP.
2
Under some special conditions, with governments intervening with non-distortionary
3
taxation to optimally redistribute income across generations, then observed market rates of
4
interest will accord with the SRTP. But these are highly specialized conditions. (See
5
Stiglitz, 1985, Pestieau, 1972). The market rate of interest, of course, remains relevant,
6
because it reflects the opportunity cost of capital; the changes in consumption generated by
7
any change in policy will be strongly affected by the opportunity cost of capital, as the
8
discussion below will illustrate.
9
10
4.3.8 Discount Rate Estimates -- Prescriptive Approach
11
12
The approach described in the previous sections, under plausible assumptions, leads to low
13
discount rates for changes in consumption of future generations. With high rates of
14
productivity increase (and high g), of the order of 1.5%, and high values of the elasticity of
15
marginal utility (Θ), the social discount rate is of the order of 3%; with low rates of
16
productivity increase, of the order of .5%, and low values of the elasticity of marginal
17
utility, the social discount rate is of the order of .5%.
18
19
It must be emphasized that these discount rates apply to consumption only, and that they
20
can be applied only after the foregone benefits of other investments have been included in
21
the costs of the program.
22
23
In general, there is no reason why the discount rate should be constant over time even if p
24
and Θ are constant, since g need not be constant. In fact, if we assume a gloomy scenario
25
where future output and consumption decline, then g and thus the SRTP may be negative
26
(Munasinghe, 1993).
27
28
Developing countries, with higher rates of productivity increase, can justify higher rates of
29
discount, at least until the gap between their standards of living and those of the more
30
developed countries have been closed. With labor productivity increases and per capita
31
income growth of the order experienced by the Asian miracle countries of 5 to 8% per year,
32
and elasticities of marginal utility of 2, discount rates of the order of 10% to 16% could be
33
justified. Similarly, low-income countries close to subsistence levels could have high
34
elasticities of marginal utility (rapid dropoff of marginal utility from initially extremely high
35
levels associated with privation), so that their SRTPs could be high even if they were
36
experiencing slow growth over long periods. These distinctions have important
37
implications for climate change policy, because they would tend to mean that the calculus
38
of trading off present abatement costs against future benefits from avoidance of climate
39
change damage could be less attractive for developing countries than for industrial
40
countries. However, there are other elements in the calculus that could go the other way,
41
such as the likelihood of higher relative future damage from climate change for the
42
developing countries (see chapter 6).
43
44
Specific applications in the still new economic literature on climate change have adopted
13
1
important differences in estimation of the discount rate. To follow the approach suggested
2
by Cline (1992), with a zero rate of pure time preference (p), and using the consumption
3
growth rate of 1.6% per capita from the IPCC scenarios (IPCC 1992) multiplied by an
4
elasticity of marginal utility (Θ) of 1.5, gives an SRTP of 2.4%. If instead if is assumed that
5
per capita growth is only 1% (perhaps because of slower growth after 100 years), or if Θ =
6
1, then the SRTP becomes 1.5%. After taking account of the share of resources coming out
7
of capital (20% economy-wide, versus 80% out of consumption) and taking into account
8
the opportunity cost of displaced capital and depreciation, the effective discount rate
9
becomes 2% to 3%.
10
11
4.4 Formulation of Descriptive Approach
12
13
As noted above, on ethical grounds, it is difficult to support a rate of pure time preference
14
much above zero. On the same ethical grounds, however, it is equally difficult to support
15
development assistance budgets for the OECD countries that average as low as one quarter
16
of one percent of GDP, to cite one example. Yet this is what occurs in the real world. The
17
descriptive approach to the discount rate looks at returns to investments in the real world,
18
and with this information seeks intertemporal efficiency. It asks: could consumption be
19
increased at one date without decreasing it at any other? The descriptive approach says
20
that if a policy is not intertemporally efficient, it should not be adopted.
21
22
An important difference between the two approaches is that the descriptive approach
23
includes the opportunity cost of capital directly, while the prescriptive approach includes it
24
indirectly, as the box illustrates (and as discussed below with respect to shadow pricing
25
capital). Most climate change optimization models (e.g. Nordhaus, 1993a, b; Peck and
26
Teisberg, 1992; Manne et al., 1993) rely on the descriptive approach. In practice, both
27
approaches may lead to similar results, and similar policy recommendations.
28
29
4.4.1 Returns to Investment
30
31
Nordhaus (1994), Lind (1994), Birdsall and Steer (1993), and Manne (1994), among others,
32
have all stressed the importance of the opportunity cost of capital. Most investments return
33
far more than the 1% to 2% suggested above as the appropriate long-term discount rate for
34
consumption. A review of World Bank projects estimated a real rate of return of 16 percent
35
at project completion; one study found returns of 26% for primary education in developing
36
countries. Even in the OECD countries, equities have yielded over 5% for many decades,
37
after corporate and other taxes, comparable to a pretax rate of at least 7%. An investment
38
at even 5% returns 18 times more than one at 2% over 100 years. Thus, society would be
39
foolish to forgo a 5% investment for a 2% investment.
40
41
Further, selecting a discount rate of 2% implies far more investment than actually occurs in
42
any country now, and thus would require a big jump in savings rates to finance the
43
increased investment. But even if savings could be increased enough to drive the discount
44
rate to one or two percent, climate change investments would still have to compete with
14
Example: Project Evaluation using prescriptive and descriptive approaches
Suppose a greenhouse mitigation project is under consideration. If undertaken now, it will
cost $1 million. If not undertaken, a new sea wall might be required in year 50, costing $10
million. If it is necessary, building a sea wall would avoid damages of $ 1 million per year.
capital cost:
$ 1 million
time until damages begin
50 years
cost of sea wall, year 50
$10 m
avoided damages, years 50,51,52,53
1 m/yr
opportunity cost of capital:
5%
The decision maker has 4 options:
a. Do nothing (year 0), do nothing (year 50)
b. Do nothing (0), build sea wall if necessary (50) --
C. Mitigation project (0), do nothing (50)
d. Other investment (0), build sea wall if necessary (50)
The stream of benefits is as follows:
Option (year)
0
50
51
52
a.
0
0
0
0
b.
0
-10
1.0
1.0
C.
-1
1.0
1.0
1.0
d.
-1
11.5
1.0
1.0
-10
=1.5
At most plausible discount rates, option b dominates option a -- i.e. if the sea level rises, it is
better to build the sea wall than do nothing. Option d dominates option C, as investing the $1
million in year 0 at 5% yields $11.5 million in year 50, enough to build the sea wall with $1.5
million left over. But option d may be institutionally infeasible, as there may be no way to put
aside $1 million today and leave it untouched for 50 years in a sort of Fund for Future
Greenhouse Victims. If d is infeasible, then the choice between b and C will depend on the
value attached to the extra consumption along path b in years 0 to 49; this will depend on the
consumption rate of discount.
1
many other public and private investments offering higher returns. Birdsall and Steer of the
2
World Bank (1993) explain the problem:
3
15
1
we feel that
2
meeting the needs
Estimated returns on financial assets and direct investment
3
of future
4
generations will
Asset
Period
Real return (%)
5
only be possible if
High-income industrial
6
investable
countries:
7
resources are
equities
1960-84
5.4
8
channeled to
bonds
1960-84
1.6
9
projects and
nonresidential capital
1975-90
15.1
10
programs with the
gvt. short-term bonds
1960-90
0.3
11
highest
U.S.
12
environmental,
equities
1925-92
6.5
13
social, and
all private
14
economic rates of
capital, pretax
1963-85
5.7
15
return. This is
corporate capital,
16
much less likely to
posttax
1963-85
5.7
17
real estate
1960-84
5.5
happen if the
farmland
1947-84
5.5
18
discount rate is set
Treasury bills
1926-86
0.3
19
significantly lower
20
than the cost of
Developing countries:
21
capital.
primary education
various
26
22
higher education
various
13
23
That opportunity costs
Sources: Ibbotson and Brinson 1987; Stockfisch 1982, 1989;
24
(market rates of return)
Brainard, Shapiro, and Shoven 1991; Psacharopoulos 1985;
25
usually exceed the SRTP
Nordhaus 1994, Cline 1992.
26
suggests the existence of
27
better alternatives than
28
those barely satisfying,
29
say, a two percent rate of
30
return criterion. At this
31
point the question becomes: why has the SRTP not been brought into accord with observed
32
market rates of return? In the context of climate change, the central argument of the
33
prescriptionists is that other alternatives are not feasible. As suggested earlier, society is
34
not likely to be able to set aside investments over the next three centuries and earmark the
35
proceeds the eventual compensation of those adversely affected by climate change. Even if
36
such a commitment could somehow be made, there would be no guarantee that the rate of
37
return on capital would remain where it is today over this period.
38
39
Accordingly, if the long-term consumption rate of discount is 1% to 2%, say the
40
prescriptionists, then a climate change investment returning 2% is better than no investment
41
at all. To the argument that a discount rate of 2% is glaringly inconsistent with observed
42
behavior (e.g., government spending on education or research, development assistance by
43
donor countries), prescriptionists reply that just because the government fails to allocate
44
resources properly in one area is no reason to insist that decisions in other areas be
16
1
consistent with that initial decision.
2
3
4.5 Discounting for Risk
4
5
The standard treatment of risk in models involving impacts over a single individual's life is
6
not to raise the discount rate for riskier projects, but instead to convert probabilistic
7
consumption patterns into their certainty equivalents and then discount the results at the
8
standard rate. The same should be true for the pure time preference component of the
9
SRTP when discounting across generations. This component should remain unchanged
10
with respect to risk, and the influence of risk should be incorporated in the stream of
11
expected consumption effects instead.
12
13
There would seem to be an argument for varying the growth-based component of the SRTP
14
with respect to risk, however. If there is uncertainty about the rate of per capita income
15
growth, g, then consider the effect on the component Θg in the SRTP. Suppose there are
16
two scenarios each with 50 percent probability: per capita income growth of 1 percent and
17
per capita growth of 2 percent. There will be two resulting possible streams of marginal
18
utility over time. The stream of expected value of marginal utility will be the average of
19
these two streams. But if marginal utility is a convex function of consumption³, this
20
average will be greater than the stream of marginal utility generated by considering the
21
simple average growth rate over time, 1.5 percent. That is, with diminishing marginal
22
utility, at any point in time marginal utility along the path for 1.5 percent growth will be
23
closer to that of the 2 percent growth path than to that of the 1 percent growth path.
24
Correspondingly, the expected marginal utility path lying halfway between the two
25
scenarios will coincide with the marginal utility stream for a growth rate closer to 1 percent
26
than to 2 percent. Essentially, the expected value of marginal utility is greater than the
27
marginal utility of expected income. On this basis, there would be grounds for reducing the
28
growth-based component of the SRTP under circumstances of risk.
29
30
Because the risk in predicting per capita growth on centuries-scale horizons is high, this
31
consideration is particularly relevant for the problem of climate change.
32
33
4.6 Producer or Consumer Interest Rates?
34
35
A large literature has debated whether, for small changes in consumption levels, observed
36
rates of interest provide the appropriate basis of trading off government expenditures and
37
changes in consumption of individuals of different generations at different dates. In a
38
world in which there was no taxation, no market distortions, and a single individual living
39
forever (or else "dynastic" utility functions in which individuals take full account of their
40
descendants' welfare), society's intertemporal discount rate should presumably correspond
41
to that of the representative individual, and his trade-offs across time would be given by the
42
market rate of interest.
43
44
But these assumptions are not generally satisfied, as evidenced by the marked discrepancy
17
1
between the lower interest rates on savings typically facing consumers and the higher rates
2
earned on investments by producers.
3
4
Part of the source of the frequent confusion about appropriate discount rates is a confusion
5
about what is being discounted. In the social discount rate approach, what is being
6
discounted is changes in consumption at different dates. Typically, in the producer interest
7
rate approach, what is being discounted are the direct cash flows from the project. The two
8
need not be inconsistent; under certain conditions, using producer interest rates in
9
evaluating direct cash flows and using the social discount rate in evaluating changes in
10
consumption will give the same results -- but only under the specialized conditions
11
specified earlier.
12
13
Four schools of thought have emerged on the proper discount rate for public policy. One
14
suggests that it should be the producer interest rate, representing the private rate of
15
transformation between investment today and output in the future; one, the consumer
16
interest rate; one, that it should be some combination of the two; and one, that it should
17
have nothing to do with either. In the absence of any non-distortionary intertemporal social
18
welfare maximization by the government, we have already noted the general result that the
19
rate of discount should be neither.
20
21
In special cases, however, one of the other three schools of thought may be correct.
22
For instance, if the government were comparing two projects, both of which cost the same,
23
and both of which yielded their output in the same year, then a comparison of the rates of
24
return would provide an appropriate basis of choosing among projects. Cline (1992)
25
proposes a shadow price of capital set equal to the present discounted value of an annuity
26
paying equal annual installments over a lifetime of N years (set at 15 years for the lifetime
27
of typical capital equipment), with a return of r equal to the rate of return on capital, and
28
discounted at the social rate of time preference (SRTP). With plausible ranges for N, r, and
29
SRTP, the shadow price of capital tends to be in the range of 1-1/2 to 2 units of
30
consumption equivalent per unit of capital.
31
32
If a public project displaced a private project of equal cost, the same reasoning
33
would imply that the government should only undertake the public project if the rate of
34
return exceeded the rate of return in the private sector (Stiglitz, 1982). More generally,
35
when the government undertakes a project, complex general equilibrium effects can be
36
expected. The full consumption effects of these changes (or their consumption equivalents)
37
need to be calculated, and then discounted using the SRTP (Arrow and Kurz, 1970;
38
Feldstein, 1970; Bradford, 1975; Stiglitz, 1982). Implementation of this approach can
39
apply a shadow price of capital to convert all investment effects into their (magnified)
40
consumption equivalents, and then apply the social rate of time preference for consumption
41
to discount the resulting stream of consumption equivalents (Lind, 1982; Gramlich, 1990).
42
A shadow price of capital greater than unity reflects the fact that the rate of return on
43
capital exceeds the SRTP; care must be exercised in evaluating the shadow price and its
44
path over time (Cline, 1992).
18
1
Sometimes, as just discussed, one can look at the direct expenditures and apply an adjusted
2
discount factor, the public sector discount rate. There is a large literature emphasizing
3
different aspects of the adjustment methodology. One body of literature emphasizes the
4
effects on consumption versus investment, deriving a weighted average of the consumption
5
and investment rates of return, with weights depending on the respective importance of the
6
sources of finance (Sandmo and Dreze, 1971). Within the literature on optimal taxation and
7
production (where the discrepancy between producer and consumer rates of interest arises
8
from optimally determined tax rates), if the government is relatively unrestricted in the set
9
of commodity taxes that it can impose, the producer rate of interest should be used to
10
discount (Diamond and Mirrlees, 1971; Pestieau, 1974). However, in the more relevant
11
regime in which government faces constraints on the sets of taxes that are imposed, there is
12
no simple relationship between the appropriate public sector discount rate and the producer
13
interest rate (Stiglitz and Dasgupta, 1971, Stiglitz, 1985, 1988 ).
14
15
4.7 Discounting Environmental Impacts
16
17
The essence of social discounting is to convert all effects into their consumption equivalents
18
at the proper relative prices, and then to discount the resulting stream of consumption
19
equivalents at the social rate of time preference. Incorporating environmental effects thus
20
does not change the SRTP itself, but requires special attention to the proper relative pricing
21
of environmental goods over time. While there is a generally accepted approach to valuing
22
goods, there is less consensus concerning valuation of environmental impacts, other than
23
those valued solely for their impacts on the production of goods. The question is addressed
24
within the public finance literature in terms of the valuation of public goods. Assume
25
consumers have utility functions of the form U = U(c,G) where G is some public good
26
(e.g., quality of the environment). Then marginal rates of substitution between c at
27
different dates may bear no correspondence to marginal rates of substitution between G at
28
different dates. This implies that there is no justification for discounting environmental
29
degradation at market rates of interest The appropriate procedure entails converting the
30
environmental change into contemporaneous consumption benefits, and discounting those.
31
32
Technical progress and structural change over the past several decades have resulted in
33
improvements in several measures of environmental quality in the developed countries
34
(World Bank 1992). Moreover, recorded reserves of many "exhaustible resources" have
35
actually increased over the last century, accompanied by a fall in their real prices. This
36
provides evidence that continued growth in per capita incomes will result in improved
37
environmental quality in at least some dimensions. Some have supposed, however, that
38
environmental degradation will occur as society grows (Weitzman 1993). If this occurs or,
39
more likely, if the environment is an income elastic good on which people are willing to
40
spend relatively more as their income rises, then the marginal rate of substitution between
41
environmental quality and private goods will systematically change over time, toward a
42
higher relative marginal value of the environment. The result equivalent to using a low (or
43
even negative) discount rate for environmental amenities (see annex 4A), with prices
44
unchanged. However, it is important to reiterate that this process involves properly valuing
19
1
future environmental benefits in arriving at the future flow of consumption, and does not
2
change the appropriate discount rate itself at which the consumption stream should be
3
discounted.
4
5
Much of the environmental literature critical of cost-benefit analysis, in contrast, argues for
6
a zero discount rate without seeming to recognize the distinction between a zero rate of
7
pure time preference (p) and a zero social rate of time preference (SRTP; see, e.g., Daly
8
and Cobb, 1991; Norgaard and Howarth, 1991). But from equation 4.1, so long as
9
consumption growth is positive there will be a nonzero SRTP. Similarly, some modern
10
philosophers make the same mistake (e. g. Parfit, 1984; Cowen and Parfit, 1992).
11
12
Finally, there has been considerable discussion about the proper discounting method for
13
environmental projects of institutions such as the Global Environmental Facility of the
14
World Bank (see e.g. Munasinghe, 1993). The method that follows from the social
15
cost-benefit approach is to obtain consumption equivalents of the environmental effects over
16
time and then discount at the SRTP. The consumption equivalents of carbon emissions, for
17
example, can be evaluated by applying the carbon shadow price from models of climate
18
change damage and optimal abatement, so long as those models themselves are
19
implemented with discounting based on appropriate values for the rate of pure time
20
preference and the elasticity of marginal utility (Cline, 1993). Within a fixed institutional
21
investment budget, it may be that the collection of potential projects that successfully passes
22
a cost-benefit test on this basis more than exhausts available funds. If so, efficient
23
trade-offs within the menu of projects will appropriately involve cutoffs at a higher shadow
24
price in funds drawn from the institutional budget but always with benefits evaluation
25
based on the consumption equivalence principle just outlined.
26
27
28
4.8 Implications for Sustainable Development
29
30
Economics has long recognized the concept of sustainability. Hicks (1938) used the
31
idea in defining net national income. Neoclassical growth theory (Phelps date, Meade
32
date, Robinson date) advanced the idea of sustainability in its formulation of the "Golden
33
Rule": that configuration of the economy giving the highest level of consumption per head
34
that can be maintained indefinitely. A recent extension has proposed the "Green golden
35
rule" (Beltratti, Chichilnitsky, and Heal date). The recent economic debate on sustainable
36
development has focused on two issues: 1) intertemporal equity and 2) capital accumulation
37
and substitutability.
38
39
Intertemporal equity. Robert Solow's definition (Solow, 1992) focuses on
40
intertemporal equity: sustainable development requires that future generations be able to be
41
at least as well off as current generations. Sustainable development does not preclude the
42
use of exhaustible natural resources, but requires that any use be appropriately offset.
43
Likewise, any environmental degradation must be offset by an increase in productive capital
44
sufficient to enable future generations to obtain at least the same standard of living as those
20
1
alive today.
2
3
Capital accumulation and substitutability. Solow's definition, and much of
4
economic theory to date, implicitly assumes that substitutes exist or could be found for all
5
resources. If substitution possibilities are high, as most evidence from economic history
6
indicates, then no single resource is indispensable, and intertemporal equity stands as the
7
only crucial issue (Pearce 1988). If on the other hand, human and natural capital are
8
complements or only partial substitutes then different classes of assets must be treated
9
differently, and some assets are to be preserved at all costs.
10
11
Some have argued that the future damages from climate change are on the order of 1 to 2
12
percent of GDP or less, whereas aggressive abatement costs are larger (Nordhaus, 1993). If
13
this is the case, then taking costly actions now to mitigate global climate change later must
14
rest on one of two arguments: First, that prudence calls for avoiding a large-scale
15
experiment with the planet, and avoiding climate change lies beyond normal economic
16
calculus; or second, that the potential exists for large, sudden, irreversible non-linearities
17
with major effects on the economy, particularly the economy of certain countries or regions.
18
19
Others have argued that if a longer time horizon is considered than that of the conventional
20
benchmark of 2xCO₂, and if upper bound warming and damages are taken into account, and
21
considering the range of estimates for abatement costs, then even a standard economic
22
analysis using the social discounting approach outlined here can conclude that the benefits
23
of aggressive action outweigh the costs on economic grounds (Cline, 1992). This result
24
obtains even though technical progress permits future per capita incomes eventually to be
25
much higher than present, because given conventional estimates of the elasticity of marginal
26
utility, the intergenerational bargain of exchanging modest costs today for large avoided
27
damages in the future remains attractive. In this case, the two considerations just noted
28
simply reinforce this conclusion.
29
30
In many developing countries, Solow's definition would not be viewed as acceptable, since
31
it seems to place no weight on their aspirations for growth and development. Further,
32
formal models analyzing optimal development paths using a min-max (Rawlsian) criterion
33
would, of course, focus exclusively on the welfare of the less developed countries in the
34
first place (Note that in Rawls' formulation, Θ= ∞). But the prescription would be simple:
35
massive redistribution from the North to the South immediately, without introducing the
36
complication of long-term environmental problems. Even if there were limits on the
37
transfers, it would suggest that all of the costs of mitigation--including those occurring
38
within the South-- be borne by the North.
39
40
Of course, even the utilitarian approach (Θ< ∞) would tend to lead to higher general income
41
transfers to poor countries than presently observed. Adherents of the descriptive approach
42
would ask why the utilitarian construct is appropriate when considering intergenerational
43
equity (as in the identification of the SRTP suggested in equation 4.1) if it is not applied in
44
practice across (or, for that matter, within) countries at the present. In one sense, this
21
1
question is another application of the principle suggested above that in the absence of
2
optimal redistribution intervention by the government, observed market rates (in this case of
3
North-South transfers) will not necessarily or likely equal social rates. Alternatively, the
4
equity norm suggested here may not be widely shared by governments or voters.
5
6
Some might seek to explain the paucity of present-day North-South transfers on grounds of
7
incentive effects. Thus, the massive marginal tax rates on the North that would be
8
required to equalize income levels with the South could seriously reduce the amount of
9
output available for redistribution. Perhaps more relevant (in view of the modest but
10
unfulfilled international targets for grant aid), the problems of still unaddressed poverty
11
within the North and sharp inequality within many developing countries complicate
12
attempts to improve overall welfare through North-South transfers. The fundamental
13
explanation, however, is that each country tends to consider primarily the welfare of its
14
own citizens, and only secondarily that of others.
15
16
Despite the political constraints on present-day North-South transfers that would otherwise
17
be recommended by the utilitarian approach, the time-discounting concepts of that
18
approach, and the SRTP in particular, remain valid subject to these constraints. Thus,
19
consider a matrix with two rows North and South -- and two columns -- present and
20
future. The SRTP can appropriately be applied between the two columns along each row,
21
even if there is a barrier to its application between the two rows. Leaders and publics in
22
developing countries have cause for concern about their descendants just as do their
23
counterparts in developed countries. As noted above, however, the value of the SRTP is
24
likely to be higher for the South row than for the North row.
25
26
At the same time, both the Rawlsian and utilitarian approaches imply that the South should
27
actively engage in mitigation financed by transfers from the North, so that efficient
28
mitigation can be obtained. Any Pareto efficient development path (including any
29
reasonable definition of sustainable development) must have this property of efficient
30
mitigation. Thus, the level of mitigation should be set based on the eventual target of
31
greenhouse gas stabilization and implied levels of world emissions (without regard to where
32
those emissions originate). The magnitudes of actions required to attain those levels of
33
world emissions will, of course, depend on the rates of growth of the less developed
34
countries as well as the increased energy efficiency among both the developed and
35
developing countries.
22
1
Annex 4A
2
Methodological Notes on Discounting
3
4
4A.1 Intertemporal Maximization of Well-Being
5
6
In an influential series of articles, Koopmans (Koopmans, e.g., 1960) conducted a series of
7
thought-experiments on intertemporal choice so as to see the implications of alternative sets
8
of ethical assumptions in plausible worlds. He suggested that we can have no direct
9
intuition about the validity of discounting future well-beings, unless we know something
10
concrete about feasible development paths. Applying this approach, Mirrlees (1967) and
11
Chakravarty (1969) showed that in plausible economic models for developing countries, not
12
to discount future well-being could imply that the present generation be asked to save and
13
invest around 50 per cent of gross national product -- a stiff requirement when GNP is low.
14
Nonetheless, these models tended to assume high rates of return on capital (a constant 33
15
percent rate in Chakravarty, 1969) and to consider time periods of decades rather than
16
centuries. It is unclear that their findings hold for the centuries-scale problem
17
of climate change, in part because of the much lower likely average return to capital over
18
this horizon.
19
20
Koopmans (1960) considered the set of feasible consumption paths (from the present to the
21
indefinite future) and the corresponding et of welfare or "well-being" paths. These paths
22
could then be ordered to select the optimum path of well-being, according to the criterion:
23
24
4A.1) z=√ᵣ₀ W(c,) e⁻ptdt
25
26
where p > 0.
27
28
Correspondingly, the discount rate for the time path of consumption is:
29
30
4A.2) i,=i(c,) =p+0(c,) [dc,/dt]/c,
31
32
where Θ(c) is the elasticity of marginal well-being, or marginal utility, at time t (Arrow and
33
Kurz, 1970). (Note that whereas the main text treats this term as a constant, it is explicitly
34
considered to vary with the level of consumption in the treatment here.) Along a full
35
optimum path, the consumption rate of discount equals the productivity of capital (i.e. the
36
social rate of return on investment). This is the famous Ramsey Rule (Ramsey 1928).
37
38
A convenient form of W is that giving a constant elasticity of marginal utility, such as:
39
40
4A.3) W(c) = c-Θ
41
42
As discussed in the text, the larger is the rate of pure time preference (p) the lower is the
43
weight accorded to future generations' well-being relative to that of the present generation.
44
Mirrlee's (1967) computations essentially introduced this sort of bias (p > 0) as a way of
23
1
countering the advantages to be enjoyed by future generations, should the productivity of
2
capital and technological progress prove to be powerful engines of growth.
3
4
As noted in the text, a higher value of Θ means greater emphasis on intergenerational
5
equity. As Θ
∞, the well-being functional in 4A.1) resembles more and more the
6
Rawlsian max-min principle; in the limit, optimal growth is zero.
7
8
In 4A.3), W(c) is has no minimum value. If p = 0, this ensures that very low consumption
9
rates are penalized by the optimality criterion reflected in (4A.1). On the other hand, if
10
p were positive, low consumption rates by generations sufficiently far in the future would
11
not penalized by (4A.1). This means that unless the economy is sufficiently productive,
12
optimal consumption will tend to zero in the very long run. Dasgupta and Heal (1974) and
13
Solow (1974a) showed in a model economy with exhaustible resources that optimal
14
consumption declines to zero in the very long run if p > 0 and in the absence of technical
15
change, but that it increases to infinity if p=0.
16
17
It is in such examples that notions of sustainable development can offer some analytical
18
guidance. If by sustainable development we mean that the chosen consumption path should
19
never fall short of some stipulated, positive level, then it follows that the value of P would
20
need to be adjusted downward in a suitable manner to ensure that the optimal consumption
21
path meet the requirement. This was the substance of Solow's remark (see Solow, 1974b)
22
that, in the economies of exhaustible resources the choice of P can be a matter of
23
considerable moment.
24
25
So far an assumption underlying this discussion has been that well-being or utility has not
26
been bounded. If we restrict well-being to be bounded, other results obtain, because of the
27
mathematical properties of the space of bounded sequences. For such sequences present
28
value calculations are not rich enough to capture all of the subtleties of evaluation of a
29
utility stream. Instead, one has to add to the present value another term. Chichilnisky
30
(1994) has suggested that the present value term represents the requirement that the future
31
should not be a dictator over the present; and that the second term represents the
32
requirement that the present should not be a dictator over the future. This second term will
33
in general have the form of a long-term average. It could be approximated by minimum
34
requirements for the long run stocks of environmental resources. This formulation attempts
35
to account for both basic levels of human needs and limitations on total resources.
36
37
Resource Sourcing and the Consumption versus Investment Discount Rate
38
39
Sandmo and Dreze (1971) address the choice of the correct rate of discount to use in the
40
public sector when there are distortions in the economy, e.g. in the form of taxes, which
41
prevent the equalization of marginal rates of substitution and transformation in the private
42
sector. Under certain assumptions, the corporate tax drives a wedge between the marginal
43
rate of time preference of consumers and the marginal rate of transformation in private
44
firms.
24
1
They find that for a closed economy:
2
3
4A.4) 1+r<1+i< 1 r/l-t
4
5
where r is the consumer interest rate, t is the tax rate, and i is the public sector's discount
6
rate. This rate should thus be a weighted average of the rate facing consumers and the tax-
7
distorted rate used by firms. Since 1+r measures the marginal opportunity cost of
8
transferring a unit of resources from private consumption, and since 1+ r/(1-t) is the
9
measure for transfers from private investment, a unit of resources transferred from the
10
private to the public sector should be valued according to how much of it comes out of
11
consumption and how much out of investment⁴.
12
13
The general approach taken throughout this chapter is to calculate impacts on consumption,
14
and to find the appropriate discount factor for discounting those changes. We are, in effect,
15
taking consumption as our numeraire. This is convenient and natural, but there are other
16
ways of performing the calculations, using other numeraires. Using other numeraires,
17
relative prices over time (discount factors) will differ from those associated with the
18
consumption numeraire. The failure to recognize this is a great source of confusion in the
19
discounting literature. Thus, as suggested in section 4.7, if environmental goods had been
20
chosen as the numeraire, the discount rate could even have been negative. It is important to
21
remember, however, that nothing substantive depends on the choice of numeraire.
22
23
By the same token, if for example systematic relationships exist between the outputs and
24
inputs of a project and the total changes in consumption they induce, and if consumption
25
changes over time, then instead of discounting total consumption impacts as the SRTP, one
26
could calculate the direct impacts using another discount factor. The discussion above of
27
the Sandmo-Dreze formulation is a case in point. These alternatives do not provide
28
prescriptions, only alternative formulas for arriving at the same point.
29
30
The discrepancy between public evaluation of a marginal dollar to future generations, and
31
individuals' own intertemporal evaluations can arise even in the case of very simple social
32
welfare functions. Thus, assume that there is a utilitarian social welfare function, which
33
simply adds up the utility of successive generations, and for simplicity, assume each
34
generation lives for only two periods. The t generation's utility is represented by a utility
35
function of the form:
36
37
4A.5) U'(c',, c'₁+₁)
38
39
where the first argument refers to consumption the first period of the individual's life, the
40
second to consumption the second period. Then observed market rates of interest refer to
41
how individuals are willing to trade off consumption over their own life. These may or
42
may not bear a close correspondence to how society is willing to trade off consumption
43
across generations. The former corresponds to u'₂/u'₁, while the latter corresponds to
44
25
1
If the government has engaged n optimal intertemporal redistribution and does not face
2
constraints in imposing lump sum (i.e. nondistorting) taxes on each generation, then the two
3
will be the same, and equal to the marginal rate of transformation (in production, i.e. the
4
return to investment). But whenever either of these conditions is not satisfied, then market
5
rates of interest facing consumers (measuring their own marginal rates of substitution) need
6
bear no close relationship to society's marginal rate of substitution across generations.
7
Diamond and Mirrlees (1970, 1971) show that if the only reason for the discrepancy
8
between producer and consumer interest rates is optimally determined commodity taxes,
9
and there are no after-tax profits, possibly because there is a 100% pure profits tax, then the
10
government should use the producer rate of interest. Stiglitz and Dasgupta (1971) have
11
shown that this result does not hold if either of these assumptions is dropped.
12
13
Under certain circumstances (in particular the existence of optimal intergenerational lump
14
sum transfers), asymptotically the producer rate of interest will equal the pure rate of time
15
preference of society. More generally, when the government must resort to
16
distortionary taxes, not only is this not true, but the rates of discount employed may reflect
17
distributional considerations (see Stiglitz 1985).
26
1
Endnotes
2
3
&
.This approach ignores diversity of individuals within any generation, and is thus subject to Schelling's
6
criticism. Other simplifying assumptions here include the use of a utilitarian social welfare function, the
7
assumption of the same utility function for each generation, and no recognition of changes in leisure (or,
8
alternatively, by the assumption that those impacts can be folded into a single consumption equivalent
9
variable). The standard models further assume that the elasticity of marginal utility with respect to changes in
10
consumption (Θ) is constant. Each of these assumptions could be lifted, complicating the analysis
11
substantially. The major effect would be to make the variables p and Θ depend on other, possibly
12
endogenous, variables, rather than taking the simple form postulated here.
13
2. With a non-utilitarian social welfare function, social welfare may be written in the form, e.g.,
14
15
S = G(U(C)).
16
17
If G = U, then it is clear that the rate at which social marginal utility diminishes with increases in
18
consumption may differ from the rate at which private marginal utility diminishes, so that evidence about the
19
latter is only partially relevant for the former.
20
3. .There is a strong consensus within the economics profession that individual's marginal utility is convex.
21
The behavioral implications with respect to risk of the assumption that U" < 0 are consistently rejected by the
22
data (it would imply, in particular, that absolute risk aversion strongly increases with consumption).
23
4.For an open economy, the elasticity-adjusted rate on foreign loans also enters the calculus. However, for
24
analysis of a global issue, this extension is probably inappropriate, as globally the economy is closed.
27
1
References
2
3
Arrow, K.J. 1982. "The Rate of Discount on Public Investments with Imperfect Capital
4
Markets," in Lind 1982.
5
6
Arrow, K.J. and M. Kurz, 1970. Public Investment, the Rate of Return and Optimal Fiscal
7
Policy, Baltimore: Johns Hopkins University Press.
8
9
Atkinson, A.B., 1970. "On the Measurement of Inequality" in Journal of Economic Theory.
10
2: 244-63.
11
12
Bergson, A., 1938. "A Reformulation of Certain Aspects of Welfare Economics" in
13
Quarterly Journal of Economics. 52: 310-334.
14
15
Birdsall, N. and Steer, A. 1993. "Act Now on Global Warming -- But Don't Cook the
16
Books," Finance and Development, 30(1), pp. 6-8.
17
18
Bradford, D.F., 1975. "Constraints on Government Investment Opportunities and the
19
Choice of Discount Rate" in American Economic Review 65(50): 887-899.
20
21
Broome, J., 1992. Counting the Costs of Global Warming, Cambridge: White Horse Press.
22
23
Brown-Weiss, E., 1989 In Fairness to Future Generations: International Law. Common
24
Patrimony, and Intergenerational Equity, Tokyo: University Nations University.
25
26
Chakravarty, S., 1969. Capital and Development Planning. Cambridge MIT Press.
27
28
Chichilnisky, G., 1994. "Global Environmental Risks and Financial Instruments" in G. Pillet
29
and F. Gassmann (eds.). Toward a Decision Making Framework to Address Climate
30
Change. Report of the IPCC Working Group III Writing Team II Montreux Meeting, March
31
3-6. Switzerland: Wurenligen & Villigen. Paul Scherrer Institute.
32
33
Chichilnisky, Graciela, 1994. "What is Sustainable Development?" (New York: Columbia
34
Univ., mimeogr.)
35
36
Cline, W.R., 1992. The Economics of Global Warming. Washington DC: Institute for
37
International Economics.
38
39
Cline, W.R., 1993. "Modelling Economically Efficient Abatement of Greenhouse Gases".
40
Paper presented at the United Nations University Conference on "Global Environment,
41
Energy and Economic Development", September 1993. Tokyo.
42
43
Cowen, T. and D. Parfit. 1992. "Against the Social Discount Rate" in P. Laslett and J.
44
Fishkin (eds.), Philosophy, Politics and Society: Series VI, Future Generations. Yale
28
1
University Press.
2
3
Daly, H.E. and J.B. Cobb. 1989. For the Common Good. Boston MA: Beacon Press.
4
5
Dasgupta, P., S. A. Marglin, and A. K. Sen, 1972. Guidelines for Project Evaluation (New
6
York: UNIDO)
7
8
Dasgupta, P., and G.M. Heal, 1974, "The Optimal Depletion of Exhaustible Resources" in
9
Review of Economic Studies, Symposium.
10
11
Diamond, P.A., 1965, "The Evaluation of Infinite Utility Streams," in Econometrica 33:
12
170. [chk AER?: National Debt in a neoclassical growth model. AER. 55: 1126-50/]
13
14
Diamond, P., and J. Mirrlees, 1971. "Optimal Taxation and Public Production. I and II,"
15
American Economic Review, vol. 61 (March and June), pp. 8-27 and 261-278.
16
17
Feldstein, M.S., 1970. "Financing in the Evaluation of Public Expenditure". Discussion
18
Paper No. 132. Cambridge MA: Harvard Institute for Economic Research (August).
19
20
Fellner, W., 1967. "Operational Utility: The Theoretical Background and a Measurement" in
21
Ten Economic Studies in the Tradition of Irvina Fisher. New York: John Wiley and Sons.
22
23
Gramlich, E.M., 1990. A Guide to Cost-Benefit Analysis. Second Edition. Englewood Cliffs
24
NJ: Prentice Hall.
25
26
Harberger, Arnold C., 1973. Project Evaluation: Collected Essays (Chicago: Markham)
27
28
Harsanyi, J.C., 1955. "Cardinal Welfare, Individualistic Ethics, and Interpersonal
29
Comparisons of Utility" in Journal of Political Economy. 63: 309-321.
30
31
Hicks, J.R. 1938. Value and Capital, London: publisher.
32
33
Howarth, R. B., and R.B. Norgaard, 1992. "Environmental Valuation under Sustainable
34
Development," American Economic Review vol. 82: pp. 473-477.
35
36
Ibbotson, R.G., and Brinson, G.P. 1987. Investment Markets, New York: McGraw-Hill,
37
updated in Nordhaus 1994.
38
39
Intergovernmental Panel on Climate Change. 1990a. Climate Change, the IPCC Scientific
40
Assessment. Report from Working Group I. Cambridge: Cambridge University Press.
41
42
Koopmans, T.C., 1960. "Stationary Ordinal Utility and Impatience" in Econometrica 28:
43
287-309.
44
29
1
Koopmans, T.C., 1966. "On the Concept of Optimal Economic Growth," in Econometric
2
Approach to Development (Illinois: Rand MCNally)
3
4
Lind, R.C., ed., 1982. Discounting for Time and Risk in Energy Policy. Washington DC:
5
Resources for the Future.
6
7
Lind, R.C. 1994. "Intergenerational Equity, Discounting, and the Role of Cost-benefit
8
Analysis in Evaluating Global Climate Policy," in Nakicenovic, et al., eds.
9
10
Little, I. M. D., and J. A. Mirrlees, 1974. Project Appraisal and Planning in Developing
11
Countries (London: Heinemann)
12
13
Lucas, R. "On the Mechanics of Economic Development," Journal of Monetary Economics,
14
July, vol. 22, no. 1, pp. 3-42.
15
16
Manne, A.S., R. Mendelsohn, and R.G. Richels, 1993. "MERGE- A Model for Evaluating
17
Regional and Global Effects of Greenhouse Gas Reduction Policies". Mimeo. Electric
18
Power Research Institute, Palo Alto CA.
19
20
Manne, A.S., "The Rate of Time Preference: Implications for the Greenhouse Debate," in
21
Nakicenovic et al. 1994.
22
23
Mirrlees, J. A., 1967. "Optimum Growth when Technology is Changing" in Review of
24
Economic Studies 34: 95-124.
25
26
Mishan, E.J., 1975. Cost Benefit Analysis: An Informal Introduction. London: Allen &
27
Unwin.
28
29
Munasinghe, M., 1993. Environmental Economics and Sustainable Development.
30
Washington DC: World Bank.
31
32
Nakicenovic, N., W.D. Nordhaus, R. Richels, and F.L. Toth, eds. 1994. Integrative
33
Assessment of Mitigation, Impacts, and Adaptation to Climate Change: Proceedings of a
34
Workshop held on 13-15 October 1993. Laxenburg, Austria: International Institute of
35
Applied Systems Analysis.
36
37
Nordhaus, W.D., 1994. Managing the Global Commons: The Economics of Climate
38
Change. Mimeo. Cambridge, MA: MIT Press.
39
40
Nordhaus, W.D., 1993b. "Reflections on the Economics of Climate change" in Journal of
41
Economic Perspectives. 7(4): 11-25.
42
43
Nordhaus, W.D., 1993c. "Climate and Economic Development: Climates Past and Future"
44
in Annual Conference on Development Economics. Washington DC: World Bank.
30
1
Nordhaus, W.D., 1991. "To Slow or not to Slow: The Economics of the Greenhouse Effect"
2
in Economic Journal 101(407): 920-937.
3
4
Parfit, D., 1983. "Energy Policy and the Further Future: the Identity Problem, " in D.
5
MacLean and P.G. Brown, eds., 1983. Energy and the Future (Totowa, New Jersey:
6
Rowman and Littlefield, PP. 38-58)
7
8
Pearce, D.W., and D. Ulph, 1994. "Estimating a Social Discount Rate for the United
9
Kingdom", Mimeo. Centre for Social and Economic Research on the Global Environment,
10
University College London and University of East Anglia.
11
12
Peck, S.C. and T.J. Teisberg, 1992. "CETA: A Model for Carbon Emissions Trajectory
13
Assessment" in Energy Journal 13(1): 55-77.
14
15
Pestieau, P. M., 1974. "Optimal Taxation and the Discount Rate for Public Investment in a
16
Growth Setting," Journal of Public Economics vol. 3, pp. 217-35.
17
18
Pohl, G., and D. Mihaljek (1989), "Project Evaluation in Practice," unpublished paper,
19
World Bank (December), quoted in Richard N. Cooper, Environmental and Resource
20
Policies for the World Economy, Washington, D.C.: Brookings Institution, 1994.
21
22
Pscharopoulos, George (1985), "Returns to Education: A Further International Update and Implications,"
23
Journal of Human Resources 20 (fall), 583-604, cited in Nordhaus (1994).
24
25
Ramsey, F.P., 1928. "A Mathematical Theory of Saving". in Economic Journal 138(152): 543-559.
26
27
Rawls, J., 1971. A Theory of Justice. Cambridge MA: Harvard University Press.
28
29
Rothenberg, J., 1993. "Economic Perspectives on Time Comparisons" in C. Nazli., ed., Global Accord:
30
Environmental Challenges and International Responses. Cambridge MA: MIT Press.
31
32
Rothschild, M., and J.E. Stiglitz, 1973. "Some Further Results in the Measurement of Inequality" in Journal of
33
Economic Theory. 6: 188-204.
34
35
Sandmo, A., and J.H. Dreze, 1971, "Discount Rates for Public Investment Under Uncertainty" in International
36
Economic Review 2: 169-208.
37
38
Schelling, T.C., 1993. "Intergenerational Discounting". Mimeo. University of Maryland, College Park
39
(November).
40
41
Schmalensee, R., 1993. "Symposium on Global Climate Change" in Journal of Economic Persrectives. 7(4):
42
3-10.
43
44
Scott, M.F., 1989. A New View of Economic Growth. Oxford: Clarendon Press.
45
46
Sen, A., 1982. "Approaches to the Choice of Discount Rates for Social Benefit-Cost Analysis" in R. Lind et
47
al., eds., Discounting for Time and Risk in Energy Policy, 325-53. Washington DC: Resources for the
48
Future.
31
1
Sen, A., 1967. "Isolation, Assurance and the Social Rate of Discount" in Quarterly Journal of Economics 81.
2
no pages given
3
4
Solow, R., 1974a. "Intergenerational Equity and Exhaustible Resources" in Review of Economic Studies,
5
Symposium.
6
7
Solow, R., 1974b. "The Economics of Resources or the Resources of Economics" in American Economic
8
Association: Parers and Proceedings 64(2): 1-14.
9
10
Solow, R., 1992. "An Almost Practical Step Toward Sustainability," Washington: Resources for the Future.
11
12
Stiglitz, J.E., 1988. "Economic Organization, Information and Development" in H. Chenery and T.N.
13
Srinivasan, eds., Handbook of Development Economics. 94-160. Elsevier Science Publishers.
14
15
Stiglitz, J.E., 1985. "Inequality and Capital Taxation." Mimeo. IMSSS Technical Report #457. Stanford
16
University. (July).
17
18
Stiglitz, J.E., 1982. "The Rate of Discount for Cost-Benefit Analysis and the Theory of the Second Best" in
19
R. Lind et al., eds., Discounting for Time and Risk in Energy Policy. 151-204. Washington DC: Resources for
20
the Future.
21
22
Stiglitz, J.E., 1974. "Growth with Exhaustible Resources: Efficient and Optimal Paths" in Review of
23
Economic Studies, Symposium.
24
25
Stiglitz, J.E., and P. Dasgupta, 1971. "Differential Taxation, Public Goods, and Economic Efficiency" in The
26
Review of Economic Studies 37(2): 151-174.
27
28
Sundquist, E.T., 1990. "Long-Term Aspects of Future Atmospheric CO₂ and Sea-Level Changes". in R.R.
29
Revelle et al., Sea-Level Change. National Research Council, Washington DC: National Academy Press.
30
31
Tobin, James, 1970. "On Limiting the Domain of Inequality," Journal of Law and Economics, vol. 13, no. 2,
32
October, pp. 263-77.
33
34
United Nations, 1987. Our Common Future. Commission on Environmental Development. (Brundtland
35
Report).
36
37
Weitzman, M.L. 1993. "On the Environmental Discount Rate," Journal of Environmental Economics and
38
Management, V. 26, no. 2, pp. 200-209.
39
40
Weizsacker, C.C. von, 1965. "Existence of Optimal Programs of Accumulation for an Infinite Time Horizon"
41
in Review of Economic Studies 32: 85-104.
42
43
Wetherill, G.W., and E.M. Shoemaker, 1982. "Collision of Astronomically Observable Bodies with the Earth"
44
in Leon Silver et al., Geological Implications of Impacts of Large Asteroids and Comets on the Earth,
45
Boulder, CO: Geological Society of America.
46
47
To be completed:
48
49
Phelps, E. Golden Rule article
50
51
Meade, J. Golden Rule article
32
1
Robinson, J. Golden Rule article
2
3
Beltratti, Chichilnitsky, and Heal, ... Green golden rule article
4
5
Thaler and Lowenstein 1989.
33
Chapter 1
Introduction: Scope of the Assessment
Lead Authors:
J. Goldemberg, R. Squitieri, J. Stiglitz, A. Amano, X. Shaoxiong, R. Saha
January, 1995
1
INTRODUCTION: SCOPE OF THE ASSESSMENT
2
3
SUMMARY
1
4
1.1
INTRODUCTION
3
5
6
1.2
FEATURES OF CLIMATE CHANGE
4
7
8
1.3
CONTRIBUTION OF ECONOMICS
7
9
1.3.1 Risk
7
10
1.3.1.1
Portfolio Theory
8
11
1.3.1.2
Insurance
8
12
1.3.1.3
Precautionary investments
10
13
1.3.2 Sequential Decision Making
11
14
1.3.2.1
Value of information
11
15
1.3.2.2.
Option value
11
16
1.3.3
Dynamics
12
17
1.3.3.1
Kaya Identity
12
18
1.3.3.2
Non-renewable resources, backstop technologies, and
19
emission reduction strategies
13
20
1.3.4
International Public Goods
14
21
1.3.4.1
Property rights
15
22
1.3.4.2
Paying For An International Public Good: Principles
23
and Approaches
15
24
1.3.4.4
Enforcement
17
25
1.3.4.5
Knowledge
18
26
1.3.5
Efficiency
18
27
1.3.5.1
Bankable permits
19
28
1.3.5.2
Exchange/risk efficiency.
19
29
1.3.5.3
Comprehensiveness
19
30
1.3.6 General Equilibrium
20
31
1.3.6.1
Intertemporal Substitution
20
32
33
1.4
EQUITY
21
34
1.4.1
General Issues
21
35
1.4.1
Intergenerational Equity
24
36
1.4.2
Within-country Equity
25
37
38
1.5
ECONOMICS OF POLICY ACTIONS
25
39
1.5.1
Zero-cost Options
26
40
1.5.2
Government Reform
26
41
1.5.2.1
Removing Energy Subsidies
26
42
1.5.2.2
Property Rights Reform
26
43
1.5.2.3
Administrative Reforms
27
44
1.5.2.4
Regulating Non-greenhouse Externalities
27
45
1.5.2.5
Special Problems of Economies in Transition
27
46
1.5.2.6
Examples of Efficiency-Increasing Policies
28
1
1.5.3 Market Failures and Government Responses
29
2
1.5.3.1
Revising national accounts
31
3
1.5.4
Innovation
32
4
1.5.5
Carbon Taxes and Tradable Permits.
33
5
1.5.5.1
A Double Dividend?
35
6
1.5.5.2
Energy Taxes
35
7
1.5.5.3
Tradeable Permit Markets.
36
8
1.5.5.4
Combining taxes with tradeable permits
36
9
1.5.5.5
Intertemporal patterns of taxation
36
10
1.5.6 Regulatory Approaches
37
11
12
1.6
SUSTAINABLE DEVELOPMENT
38
13
1.6.1 The Economic Concept of Sustainable Development
38
14
1.6.2 Implications of Sustainable Development for Developing
15
Countries
39
16
17
1.7
CONCLUSIONS
40
18
19
1.8
ENDNOTES
42
20
57
21
22
23
1.9
REFERENCES
58
24
25
26
1
SUMMARY
2
3
Climate change presents the analyst with a set of formidable complications: large
4
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
5
long time lags between emissions and effects, an irreducibly global problem, wide regional
6
variation, and multiple greenhouse gases of concern. The risks of climate change are
7
highly asymmetrical, with a large probability of a small loss, and a small probability of a
8
large loss. Even in the presence of significant uncertainty, this asymmetry, plus the
9
principles of risk aversion and portfolio balancing provide the rationale for going beyond
10
no-regrets policies to those that incur net costs.
11
12
The atmosphere is an international public good, in that all countries benefit from each
13
country's reduction in greenhouse emissions; greenhouse gases are an international
14
externality, in that emissions by one country affect all other countries to the same extent.
15
16
Both public goods and externalities require a legal framework within which the
17
problems they pose can be addressed. No such legal framework now exists for global
18
climate change. Mechanisms for control of international public goods may include the
19
definition of property rights, the definition of limits to emissions and a consensus for
20
distributing the same in a fair and equitable manner. If, on the other hand, each agent acts
21
in its individual interest, the result will be too little of the public good and too much of the
22
externality.
23
24
Climate change demands a decision process that is sequential, can respond to new
25
information with mid-course corrections, and can include insurance, hedging, and the option
26
value of alternative courses of action. The challenge today is to identify short-term
27
strategies in the face of long-term uncertainty. The question is not, what is the best course
28
over the next 100 years, but rather, what is the best course for the next few years, knowing
29
that a prudent hedging strategy sill allow time to learn and change course.
30
31
Policy measures to reduce risks to future generations include 1) immediate reductions
32
in emissions; 2) R&D on new supply and conservation technologies; 3) continued research
33
on how much change is likely and what its effects will be; and 4) investments to assist in
34
adaptation if significant climate change occurs. A well-chosen portfolio of policies will
35
yield greater benefits for a given cost than any one option undertaken by itself. Striking the
36
appropriate balance requires taking into account costs, benefits, and risks.
37
38
In an interrelated global economic system, an attempt to reduce greenhouse gas
39
emissions in one region or one sector of the economy may be offset by increases in other
40
regions or sectors. This may occur through a) the loss of comparative advantage in the
41
carbon-intensive sectors of the regions that limit emissions; b) the relocation of industries;
42
or c) changes in world energy prices and the resulting shift in consumption. Any control
43
strategy must account for these global effects.
44
1
1
Issues of efficiency and equity can largely be separated. The Framework Convention on
2
Climate Change calls on all parties to implement cost-effective measures for abatement,
3
enhancement of sinks, and adaptation. The Framework Convention also explicitly requires
4
an equitable sharing of the burdens of response, recognizing the common but differentiated
5
responsibilities of the parties. Different countries will be affected differently by climate
6
change and by policy responses to it. The South is more likely to be adversely affected
7
than the North; moreover, developing countries often lack the financial and technical
8
resources to respond. The Framework Convention does not, however, include a formula for
9
sharing the costs of addressing climate change.
10
11
Efficiency requires that emission reductions occur where their cost is lowest,
12
irrespective of who bears the financial responsibility. Efficiency calls for removing energy
13
subsidies, reforming and clarifying property rights that affect energy use and carbon
14
storage, and reducing non-greenhouse externalities that have the side benefit of reducing
15
greenhouse emissions. Efficiency may also be promoted, and greenhouse emissions
16
reduced, by better information dissemination and by addressing capital market imperfections
17
that inhibit the adoption of energy-efficient technology. Dynamic analysis indicates large
18
potential gains from flexibility in timing of greenhouse reductions to allow for the
19
economical turnover of capital stock, and to allow time for the development of low-cost
20
substitutes. Policies that promote efficiency by requiring nations to face the full costs of
21
their actions will also address equity concerns.
22
23
Efficiency also calls for international mechanisms such as joint implementation and
24
coordinated economic instruments. Coordinated carbon taxes and tradable carbon emission
25
permits can correct the market failure associated with greenhouse emissions.
2
1
1.1 INTRODUCTION
2
3
In recent years, atmospheric emissions of greenhouse gases have risen significantly.
4
Concentrations are currently about 25 percent greater than at the beginning of the Industrial
5
Revolution. If current trends continue, concentrations will double from pre-industrial levels
6
before the end of the next century and, if unchecked, continue to rise thereafter (IPCC
7
1990).
8
9
The scientific community has noted the potentially serious effects of increased
10
concentrations. These climatic effects could, in turn, have further effects on the biosphere,
11
including an increase in mean global temperature, an increase in sea level, changes in
12
agricultural yields, forest cover and water resources, and a possible increase in storm
13
damage.
14
15
Increased concentrations of greenhouse gases are the result of fossil fuel burning,
16
livestock raising, and other human activities. Concerted action on the part of individuals
17
and governments will be required to slow the increase in concentrations. Changes in
18
greenhouse gases concentrations and the analysis of the climatic and other physical
19
consequences of those changes lie within the purview of the physical sciences. The role of
20
human activity in generating greenhouse gases, the consequences of those changes for
21
humans, and possible responses, lie within the purview of the social sciences.
22
23
Climate change impacts are likely to vary dramatically from country to country. A
24
warmer climate could benefit sectors of the economies of some mid- and high-latitude
25
countries. At the same time, a rising sea level and the possibility of increased storm surges
26
could threaten the survival of some small island states and coastal areas, and could increase
27
the risk of midcontinent drought and desertification for inland areas on the periphery of
28
deserts.
29
30
Within the past decade, a consensus has emerged on some key issues in the economics
31
of climate change. This report describes areas of consensus, as well as areas of
32
disagreement, the sources of disagreement, and further research that could narrow the range
33
of disagreement. This chapter frames the issue of climate change largely from the point of
34
view of economics but also from that of other social sciences, introducing the more detailed
35
discussions in the chapters to follow.
36
37
The commitment of resources to mitigate climate change may rest on one of two
38
arguments. The first arises from fundamental values, the second from decision analysis:
39
40
1. We have only one planet. Some changes are largely irreversible, and may occur
41
rapidly. Prudence calls for avoiding a large-scale experiment with the planet. Thus,
42
avoiding climate change lies beyond normal economic calculus.
43
44
2. The potential exists for sudden, largely irreversible non-linearities with major effects
3
1
on the economy, particularly the economy of certain countries or regions.
2
3
Even if the first view is adopted, economics has much to contribute to the discussion, for
4
the question of cost-effective emissions reductions must still be addressed. If the second
5
view is adopted, economics and cost-benefit analysis will clearly be relevant, both in
6
deciding how much mitigation to undertake, and in designing the measures.
7
8
This chapter, and others in this Assessment Report, draw on the findings of Working
9
Groups I and II, and follow the guidelines provided by the Framework Convention. That
10
Convention makes clear that important questions remain to be addressed by subsequent
11
negotiations, including the adequacy of national commitments. This chapter takes the
12
Framework convention as a political document, recognizing that particular terms of the
13
agreements reached by international negotiation may or may not accord with accepted
14
scientific research; that for instance, the agreements may or may not have provided the
15
basis of a cost-effective approach to mitigating climate change. It is hoped that the findings
16
of this chapter and the assessment report more broadly will enhance an understanding of the
17
costs and consequences of alternative actions, and will provide the scientific basis for
18
ongoing negotiations.
19
20
1.2 FEATURES OF CLIMATE CHANGE
21
22
Climate change could impose a variety of impacts on society. IPCC Working Group II
23
analyses these impacts in detail. They include effects on agriculture¹, forests², water
24
resources³, the costs of heating and cooling⁴, the impact of a rising sea level rise on small
25
island states and low-lying coastal areas⁵, and a possible increase in storm damage.
26
Although most attention to date has focused on negative impacts, some impacts will be
27
positive. Beyond these tangible impacts are a variety of intangible impacts⁶, including
28
damages to existing ecosystems and the threat of species losses⁷.
29
30
Climate change presents the analyst with a set of formidable complications: large
31
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
32
long time lags between emissions and effects, an irreducibly global problem, wide regional
33
variations, and multiple greenhouse gases of concern.
34
35
Large uncertainties. While natural scientists agree that greenhouse gas emissions are
36
rising, they do not agree about the mechanism linking concentrations, and temperature.
37
Further, although natural scientists agree that some warming will occur, they do not agree
38
on the speed of change, or the ultimate amount of change (IPCC 1992, 1994). In addition,
39
social scientists do not agree on the size of the behavioral responses or economic effects
40
that would follow, or on the effect of these changes on well-being.
41
42
Nonlinearities and irreversibilities. Nonlinearities occur when changes in one variable
43
cause a more than proportionate impact on another variable. For example, some have
44
suggested that even a modest increase in atmospheric greenhouse gas concentrations could,
4
1
beyond a certain point, trigger a substantial increase in temperature. Alternatively, even a
2
modest increase in average temperature might significantly increase weather-related
3
agricultural losses because, for many crops, extra days of extreme heat severely limit yields.
4
In Figure 1, if the threshold temperature for crop damage is Tₒ, then even a small increase
5
in the mean temperature, from ml to m2, may greatly increase the number of days above
6
the threshold, represented by the area under the curves to the right of T₀⁸. Irreversibilities
7
are changes that once set in motion cannot be reversed, at least on human time scales.
8
"Runaway warming" that some have hypothesized is an example. Some oceanographers
9
have expressed concern that warming might disrupt the deep water formation and the North
10
Atlantic Ocean circulation, responsible for much of the oceans' carbon uptake. Once
11
disrupted, this current might not revert to its original position for centuries⁹.
12
13
Long planning horizon. Greenhouse gas concentration changes occur over a long
14
period of time, with the full consequences of actions taken over the coming decades being
15
felt in future centuries. The truly long term nature of the problem is one of the distinctive
16
aspects of greenhouse gas warming. Seldom has the world consciously faced a set of
17
decisions likely to affect our descendants one, two, or three centuries from now¹⁰.
18
Because the costs of taking action today are born by the current generation, while the
19
benefits accrue possibly hundreds of years in the future, the world community is now faced
20
with issues of intergenerational equity on an unprecedented scale. While society has
21
addressed similar problems concerning trade-offs over periods of fifty or even a hundred
22
years, the long planning horizon for climate change puts the analytic questions at issue in a
23
new light. The length of time involved has one further implication: changes in technology,
24
as well as population and consumption patterns, become of paramount importance.
25
26
Long life of capital stock. Every country has made large capital investments in its
27
cities, farms, ports, etc. Some of this investment cannot be changed without large costs:
28
low-lying port cities, for example, cannot easily be rebuilt. For other investments, the cost
29
of change will be small; for diversified agricultural economies, the cost of switching from
30
one annual crop to another will be small if the temperature increase is on the order of
31
0.2°C. to 0.3°C. per decade, as climate scientists now expect (IPCC 1994). For less
32
diversified economies, the costs may be larger.
33
34
Inertia in the climate system. Atmospheric concentrations, rather than emissions,
35
determine the amount of warming that occurs. Concentrations change much more slowly
36
than emissions, meaning that affected nations might not have enough time to prevent
37
warming damage or to mitigate economic impacts after the effects of climate change
38
become evident. In this respect, the risks of climate change are unlike those of earthquakes
39
or floods. The combination of time lags in the climate system and uncertainty about effects
40
requires taking action before unambiguous information emerges about the timing and
41
magnitude of climate effects.
42
43
Global scope. Climate change is a global challenge, which cannot be answered by a
44
single country acting by itself. Mitigation must be coordinated globally. In an interlinked
5
1
world economy, not only are the actions of a single country, or group of countries, not
2
likely to be sufficient to address the problem, they are likely to be largely offset by actions
3
of other countries. If, for instance, one group of countries reduces timber cutting to
4
increase carbon absorption, then the price of lumber will rise, which may induce other
5
countries to increase cutting in their forests.
6
7
Moreover, whereas economic analysis generally takes the point of view of a single
8
decision maker or government, the important decisions on climate change will of necessity
9
be made by many sovereign governments. Economic and decision sciences are not yet able
10
to predict the outcome of bargaining problems of this type¹¹.
11
12
Regional variation. Impacts are likely to vary greatly both within and among countries.
13
Some countries and regions will suffer from warming; others will benefit, at least in some
14
sectors. Some cold countries will benefit from a reduction in heating costs and an increase
15
in the length of growing seasons; some warm countries will see a drop in yields from
16
agriculture and forestry; low-lying states are likely to suffer from increased storm surges
17
and flooding. For the world as a whole, the net effect at any time will be the sum of local
18
effects at many points on the globe, some positive and some negative; analysts have no way
19
to estimate this sum without detailed local calculations (summarized in Chapter 6).
20
21
Aggregation. On a priori grounds, there is no reason to believe that greenhouse
22
warming should have in the aggregate, either positive or negative long-run effects (although
23
adjustment implies costs in the short run). Indeed, studies twenty years ago focused on the
24
dire consequences of global cooling. It seems unlikely that the earth is somehow endowed
25
at this moment with the optimal temperature; the Earth has nonetheless adjusted to its
26
current climate, and readjustments in either direction may be costly.
27
28
Multiple gases of interest. Increases in radiative forcing (heat trapping ability) depend
29
on the concentration of all greenhouse gases, not just CO2, even though most economic
30
modeling to date has limited itself to CO2. Because greenhouse gases differ in radiative
31
forcing and atmospheric lifetime, analysts have devised measures of global warming
32
potential (IPCC 1990), making the different gases commensurable¹².
33
34
Importance of net emissions. Because greenhouse gas concentrations depend on net
35
rather than gross emissions, changes in forests and other greenhouse gas sinks must be
36
taken into account¹.
37
38
Efficiency vs. Equity. How much to reduce emissions is a matter of efficiency. Who
39
pays is a matter of equity. Economics has much to say about the former, but much less
40
about the latter. Nonetheless, equity considerations will drive many of the policy decisions
41
to be made under the Framework Convention on Climate Change.
42
43
6
1
1.3 CONTRIBUTION OF ECONOMICS
2
3
Economics and the social sciences offer perspectives on climate change not provided by
4
the physical sciences. In the classic definition, economics is the study of the allocation of
5
scarce resources that have alternative uses. Economics emphasizes the importance of
6
tradeoffs between different uses of resources, and the foregone value of other uses of a
7
resource, called the opportunity cost. In the context of climate change, this means that 1)
8
costs and benefits matter; 2) resources are not free; and 3) resources used for one purpose
9
are no longer available for other purposes.
10
11
This chapter sets out the logic of cost-benefit analysis as applied to climate change.
12
Standard cost-benefit analysis requires 1) a valuing of costs and benefits over time, using
13
willingness to pay as a measure of value and 2) a criterion for accepting or rejecting
14
proposals¹⁴. The standard criterion is the compensation principle (Kaldor date, Hicks
15
date), which says that if the project yields positive net benefits, then those made better off
16
could compensate those made worse off with something extra left over. The result is an
17
unambiguous gain in welfare, without the necessity of weighing effects on different
18
individuals.
19
20
Climate change raises difficulties with both requirements. Valuation is difficult because
21
of the difficulty in valuing environmental amenities, which are generally not traded in the
22
market. And the compensation principle will not apply if mechanisms for affecting
23
transfers do not exist, either between countries or regions in one generation, or -- especially
24
-- between generations. If transfers are not feasible, then the analysis must assign weights
25
to different individuals (for example, the utilitarianism gives equal weight to each person).
26
Only then can conclusions be drawn about net benefits for society as a whole. This issue
27
will arise again in the discussion of equity in section 1.4 below.
28
29
Beyond these fundamental concepts are ideas, originally from other areas of economics,
30
that may be applied directly to the study of climate change; this includes work on risk,
31
dynamics, sequential decision making, public goods and externalities, taxation, and general
32
equilibrium.
33
34
1.3.1 Risk
35
36
The uncertainties associated with climate change are large. Moreover, results available
37
to date from integrated assessment models indicate that climate change risks are highly
38
asymmetrical, with a high probability of a small loss and a low probability of a large loss
39
(Manne and Richels 1992, Peck and Teisberg 1993, Nordhaus 1994). In the past thirty
40
years, much new economic research has focused on rational responses to risk¹⁵, including
41
three areas important to a systematic examination and rational response to climate change:
42
portfolio theory, insurance, and decision analysis.
43
44
7
1
1.3.1.1 Portfolio Theory
2
Risk and Uncertainty
3
A portfolio manager attempts to get the
4
best return for a given level of risk. One
Uncertainty arises when a decision can lead
5
important approach is to buy several types
to a range of outcomes.
6
of assets, whose returns are not correlated
Expected return or expected value of a
7
or are negatively correlated (that is, whose
decision is the mean of the distribution of
8
prices move either independently or in
returns, the amount a person would on
9
opposite directions). In this respect, climate
average receive as a consequence of the
10
change policy decisions can be compared
decision.
11
to investment portfolio decisions.
Risk aversion measures an individual's
12
unwillingness to take risks.
13
Several responses can reduce the impact
14
of climate change: insurance and
Risk premium is the amount an individual
15
precautionary investments 16 Precautionary
would pay to replace the uncertain
16
investments may take the form of
distribution of outcomes with the expected
17
mitigation or in actions that enhance the
value.
18
ability of future generations to adapt.
Certainty equivalent is the amount that
19
Other choices include agreements for one
makes an individual indifferent between it
20
country to fund emissions reductions in
and a risky proposition; for a risk-averse
21
another country (joint implementation),
person, the certainty equivalent is higher
22
technology transfer and other forms of
than the expected return; the difference is
23
international cooperation; and phasing out
the risk premium.
24
existing policies, such as subsidies to fossil
25
fuels, that reduce welfare and directly or
26
indirectly increase greenhouse gas emissions.
27
28
A well-chosen portfolio of climate change investments will yield greater benefit for
29
a given cost than any one option undertaken by itself. For an individual country, the
30
issue is how to choose the portfolio of policy measures best suited to its circumstances, and
31
to adjust the portfolio over time in response to new developments. Governments will be
32
making climate change decisions for several decades at least. This means that they will
33
have many opportunities to adjust the size (total resources) and mix (choice of measures)
34
of their portfolio of responses.
35
36
1.3.1.2
Insurance
37
38
Several concepts from insurance have important applications to climate change: risk
39
aversion, risk premium, and certainty equivalent.
40
41
That individuals and societies are risk averse means that average utility is increased by
42
pooling risks, or equivalently that people are willing to pay to reduce the risks they face. If
43
society as a whole is risk averse, then some investments with a negative expected return,
44
for example in climate change mitigation, should be undertaken if they reduce the
8
1
probability of a loss, or the costs of future adaptation.
2
3
The magnitude of those expenditures depends on (a) society's degree of risk aversion;
4
and (b) the magnitude of the risk. The risk premium--the extra amount that society is
5
willing to pay to reduce a risk--is small if the stakes (say, the maximum loss) are small, and
6
large if the stakes are large. An investment of a dollar is justified if it reduces the loss of
7
expected utility by more than a dollar, and not if it reduces the loss by less than a dollar.
8
Thus, results reported below focusing on the expected loss of GDP from climate change do
9
not directly address the risk premium. If a possible outcome is a loss of 10%, even though
10
the expected loss is only 3%, then the certainty equivalent loss will exceed 3%. A dollar
11
investment that reduces this certainty equivalent loss by more than a dollar should be
12
undertaken; such an investment could either reduce the average loss, e.g. by reducing the
13
probability of the loss occurring (through mitigation actions) or reduce the variance of the
14
loss. For example, some actions the reduce extreme losses will have more than
15
proportionate returns.
16
17
The insurance expenditures associated with mitigation actions and investments are, in a
18
sense, only the differences between the actual expenditures and the no-regrets benefits (the
19
benefits other than those associated with greenhouse gas emissions). Thus, investments in
20
fuel efficient cars may have a direct benefit in reducing the cost of running a car. The
21
mitigation investment is only the additional investment for climate purposes.
22
23
The economic principles of risk aversion and portfolio theory provide the rationale
24
for going beyond no-regrets policies. Naturally, actions that reduce greenhouse gas
25
emissions at zero or negative costs should be undertaken. But a prudent response to the risk
26
of greenhouse gas warming goes beyond no-regrets policies to some policies incurring net
27
costs.
28
29
Traditional insurance involves pooling a large number of diversifiable risks. Insurance
30
exist when people differ in their ability to bear a given risk, or would be affected
31
differently by the risk. Insurance markets transfer risks from those who are less able to bear
32
the risk to those who are more able to bear the risk, and by spreading the risk among a
33
large number of individuals, reduce the aggregate impact of the risk¹⁷. By transferring
34
risk, traditional insurance can address one aspect of the risk from climate change: the large
35
differences in expected regional impacts. This holds irrespective of who pays the cost,
36
because wealthier individuals and countries are better able to bear risk¹⁸, and because
37
many countries likely to be most adversely affected will be developing countries of the
38
South, while many of the countries least affected (or positively affected) will be the
39
industrialized countries of the North, who could provide insurance for effects of climate
40
change that might fall harder on the South¹⁹.
41
42
Traditional insurance will not, however, address all risks from climate change.
43
Shipowners can buy insurance against storm loss because shipping risks are diversified: one
44
storm is unlikely to sink all the ships covered by the insurer. But because climate change
9
1
is global by nature, and therefore undiversifiable, traditional (intragenerational) insurance
2
cannot fully insure against climate change risks. Rather, some form of intergenerational
3
insurance is likely to arise as a way to redistribute more of the risk.
4
5
Losses associated with climate change are likely to be both correlated and large,
6
compared to losses absorbed in a single year by the commercial insurance industry. The
7
long term nature of climate change insurance also raises the problem of contract
8
enforcement: will contracts signed today be enforceable tomorrow? Will the insurers be
9
around to pay claims 50 or 100 years from now? (Even in the industrialized countries,
10
private markets may be inadequate to insure against losses from a major national disaster
11
today.) These considerations suggest that private markets may not be able to provide
12
insurance against climate change, and argue for international action to establish insurance
13
markets, perhaps with government reinsurance. Should such an insurance market be
14
established, careful attention will have to be given to ensuring that the insured parties
15
engage in appropriate adaptation actions, mitigating the losses that might be associated with
16
any greenhouse gas warming.²⁰
17
18
19
1.3.1.3 Precautionary investments
20
21
A business makes precautionary investments to reduce the total risk of its portfolio.
22
Numerous policy measures are available to reduce risks to future generations from climate
23
change. Four have dominated discussions in recent years: 1) immediate reductions in
24
emissions to slow climate change²¹; 2) R&D on new supply and conservation technology
25
to reduce future abatement costs; 3) continued research to reduce uncertainties about how
26
much change will occur and what effects it will have; and 4) investments in actions to
27
assist human and natural systems to adapt to climate change if it occurs. Other choices
28
include agreements for one country to fund emissions reductions in another country (joint
29
implementation), technology transfer and other forms of international cooperation; and
30
phasing out existing policies, such as subsidies to fossil fuels, that reduce welfare and
31
directly or indirectly increase greenhouse gas emissions.
32
33
Precautionary investments may also enhance the ability of future generations to react.
34
An important reason that people establish savings accounts is to reduce the impact of
35
unfavorable events in the future. Similarly, a society may elect to accumulate capital
36
against the possibility of a large loss from climate change. This is one thread of the debate
37
over discount rates in Chapter 3. Those who argue for a discount rate close to the
38
opportunity cost of capital point out that society may choose between immediate
39
greenhouse gas mitigation, at a cost, and delayed mitigation, with some of the money saved
40
put aside as a savings account for our grandchildren in the event of large climate-induced
41
damages.
42
43
44
10
1
1.3.2 Sequential Decision Making
2
3
As a policy question, global climate change is often posed as a choice between a) doing
4
nothing at all, or b) committing to all-out effort. Given the large current uncertainties about
5
costs and benefits of greenhouse mitigation, this is the wrong way to frame the issue, as it
6
obscures the choices that should be evaluated. Moreover, because option b) may be
7
perceived as too expensive to get political support, policy paralysis often results.
8
9
A more useful formulation is: "Given current knowledge and concerns, what actions
10
should we take over the next one or two decades to position ourselves to act on new
11
information that will become available?" (Lind 1993) For example, decisionmakers would
12
like to know if the possibility of irreversible damages²² justifies immediately undertaking
13
an aggressive abatement program²³.
14
15
Climate change demands a decision process that is sequential and can incorporate new
16
information. Timing will be a key element, and the date of resolution of uncertainty an
17
important element of the analysis. Figure 1 shows schematically the progression from a
18
simple decision to a sequence of linked decisions. The simple decision is sometimes
19
referred to as the "learn, then act" model, the second as "act, then learn"; or, in the case of
20
the decision depicted, "act, then learn, then act."
21
22
A sequential decision making strategy aims to identify short-term strategies in the face
23
of long-term uncertainty. The next several decades will offer opportunities for learning and
24
mid-course corrections. The relevant question is not: what is the best course for the next
25
100 years, but rather: what is the best course for the next few years, knowing that a prudent
26
hedging strategy will allow time to learn and change course.
27
28
For example, the choices might be 1) immediate investment in new plant and equipment
29
2) aggressive R&D on greenhouse abatement technology or 3) deferring large investment
30
for 10 years, until the nature and size of the threat is better understood, and when costs will
31
presumably have dropped and the job can be done more efficiently.
32
33
Because of the high cost of being wrong in either direction, the value of information
34
about climate change is likely to be great. In particular, the value of information about the
35
sensitivity of temperature to CO2 increases, the temperature damage function, the gdp
36
growth rate, and the rate of energy efficiency improvement is likely to be high (Chao 1991,
37
Peck and Teisberg 1991, Manne and Richels 1992, Nordhaus 1994)²⁴.
38
39
The presence of uncertainty along a dynamic path creates an option value, the value to
40
preserving choices for the future. In climate change, the term has been used in two
41
different ways. One stresses the irreversibilities of climate change; mitigation expenditures
42
now preserve the option of avoiding adaptation expenditure later. The other stresses
43
irreversibilities in investment, and the cost of premature turnover of capital.
44
11
1
1.3.3 Dynamics
2
3
The problem of greenhouse gas warming involves additions to concentrations resulting from
4
net emissions over extended periods of time. Thus the analysis must focus on dynamics.
5
Dynamic analysis involves three stages: the dynamic processes involved, the trade-offs,
6
and judgments concerning those trade-offs.
7
8
Dynamic analyses have lead to important insights. For example, atmospheric
9
concentrations and therefore temperatures depend on the total amount of greenhouse gases
10
emitted over a period of years. A given concentration target can be achieved by a variety
11
of emissions time paths. Time paths that provide for the economical turnover of existing
12
capital stock and time to develop low-cost substitutes are likely to be less costly. This
13
suggests large potential gains from flexibility in timing of emission reductions²⁵.
14
15
16
1.3.3.1
Kaya Identity
17
18
The driving forces in emissions of any greenhouse gas can be seen in the following identity
19
for carbon dioxide emissions (Kaya 1989):
20
21
CO2
=
CO2/E
E/Q
Q/L
L
22
carbon dioxide
carbon dioxide
energy
output population
23
emissions
emissions per
per unit
per
24
unit energy
output
capita
25
26
or, expressed in rates of change:
27
28
d In CO2/dt = d InCO2/E /dt + d InE/Q /dt + d InQ/L /dt + d InL /dt
29
30
that is, the percentage rate of change in carbon dioxide emissions is equal to the rate of
31
change in carbon dioxide emissions per unit energy plus the rate of change in energy
32
requirements per unit output plus the rate of change in output per capita plus the rate of
33
change in population²⁶.
34
35
This identity clarifies different approaches to reducing emissions. For a developed
36
country with a stable or slowly growing population, as long as increases in emissions/output
37
efficiency (emissions per unit energy times energy per unit GDP) keep pace with labor
38
productivity, CO2 emissions will not increase. Because of the substantial potential for
39
energy efficiency improvements, this seems feasible for most developed countries. Many
40
opportunities exist for increasing end-use efficiency, represented by the second term on the
41
right-hand side, for example from cars to public transportation, from less to more fuel-
42
efficient cars and homes²⁷. Fuel switching (for example, from coal-based electricity to oil,
43
gas, hydroelectric, wind, and geothermal), represented by the first term, also offers the
44
potential to limit CO2 emissions in many countries.
12
I
For many developing countries, emissions will increase unless energy efficiency and
2
greenhouse gas emissions per unit of energy change to offset growth in per capita output
3
and population. For many developing countries with rapidly growing populations, pressures
4
for economic development will make it difficult to direct capital from investments with
5
higher greenhouse gas emissions to those with lower greenhouse gas emissions²⁸.
6
On two essential issues evidence is currently limited: (a) to what extent will
7
improvements in energy efficiency require net increases in investment beyond the resources
8
saved from reduced energy usage, i.e., how much does aggressive emission reduction
9
depress economic growth? Order-of-magnitude calculations suggest the presence of only
10
limited trade offs, at least for the near term; (b) do developing countries have the
11
institutional capacity to achieve the desired increases in emissions efficiency?
12
13
1.3.3.2
Non-renewable resources, backstop technologies, and emission reduction
14
strategies.
15
16
Though in principle the atmosphere is a renewable natural resource, the longevity of the
17
greenhouse gases and the relationships between stocks and flows mean that for practical
18
purposes, it may be better treated as an exhaustible natural resource, but one where welfare
19
depends not just on the flow out of the stock, but on the stock itself²⁹.
20
21
The central problem with natural resources, whether renewable or not, is timing. Many
22
renewable resources possess a maximum sustainable rate of exploitation. The harvest
23
cannot long exceed this maximum rate without depleting the stock, and ultimately reducing
24
the harvest. Corresponding to the sustainable flow rate is a steady-state stock. If, initially,
25
the actual stock exceeds the steady-state stock, then, for a while, the flow can exceed the
26
maximum sustainable flow. The question is how to distribute this excess over time. Even
27
when the stock is at the sustainable level, in times of emergency it may be desirable to
28
exceed the maximum sustainable flow. For a renewable resource, this can be done, though
29
only at the expense of decreased flows later³⁰.
30
31
Timing of reductions in greenhouse gas emissions should reflect differences in costs,
32
discounting (to evaluate those costs), and risk. If technological change will make future
33
emissions reductions much less costly, some reductions should be postponed³¹
34
Conversely, research on "learning effects" shows that if actions taken today will lower costs
35
faced tomorrow, then these dynamic benefits should be included in the calculus (Arrow
36
1962, Atkinson and Stiglitz 1969). In the context of climate change, if emissions
37
constraints stimulate technical or other developments that help to lower the costs of
38
continuing or additional emission abatement, then reductions should be accelerated (Grubb
39
1993, 1995). If discount rates are high, the costs borne by future generations will carry less
40
weight than if discount rates are low. In the presence of risk, non-linearities, or
41
irreversibilities, the principle of risk aversion suggests a strategy of early mitigation.
42
43
The theory of non-renewable resources has a second set of lessons for climate change
44
policy. Among the four energy categories -- coal, gas and oil, biomass and non-carbon
13
1
sources --gas and oil are exhaustible; this limits total emissions from gas and oil. With
2
exhaustible natural resources, the question is not how much will be consumed, only
3
when³².
4
5
Figure 2 depicts fuel use in an economy, in three phases. In the first phase, the
6
economy relies on non-renewable resources; in the second, on coal; in the third, on
7
backstop technologies-- e.g., biomass combined with non-carbon sources. The switch
8
points depend on rising energy prices and improving technology, which lowers the costs of
9
the backstops. The figure shows alternative long run scenarios. In panel A, the price of
10
the backstop technology falls sufficiently slowly that for a time, the economy relies on coal.
11
In panel B, the price of the backstop technology falls fast enough to eliminate the
12
intervening stage of primary reliance on coal.
13
14
Since, to a first-order approximation, the total carbon load from oil and gas is fixed (and
15
limited), the total carbon load on the atmosphere is primarily related to coal usage. From
16
this perspective, an important uncertainty is the pace at which the cost of the backstop
17
decreases. If it decreases fast enough, the intermediate stage of coal dependence will be
18
short, and the total carbon load low, while if the price decreases slowly, the carbon load
19
can be much larger.
20
21
From this perspective, conservation of gas and oil is an important part of a risk strategy,
22
for it provides insurance against the possibility of delayed arrival of backstop energy
23
sources. 33 This, in turn, has important implications for the "leakage" debate, which asks
24
whether, in the event that the North imposes carbon taxes but the South does not, the
25
South's response will largely offset emission reductions made in the North. It is also
26
possible that lower prices for gas and oil will induce coal-rich countries to decrease their
27
reliance on coal, thus possibly producing negative leakages. But to the extent that occurs,
28
the insurance provided by greater conservation of oil and gas is eliminated, and it is this
29
insurance which should be an essential part of the dynamic strategy. If the backstop arrival
30
is delayed, then the earlier fuel switching by China and India will yield no long-run
31
benefits; while if the backstop arrival is early, the whole issue is largely moot. Leakage
32
needs to be looked at not from a static perspective, of what occurs in a single year, but
33
from a dynamic perspective, corresponding to the long run nature of the global climate
34
problem.
35
36
37
1.3.4 International Public Goods
38
39
The atmosphere is an international public good in that atmospheric concentrations are
40
the result of combined actions by all countries. A pure public good (Samuelson, 1954) has
41
two properties: the marginal cost of an additional individual using the good is zero
42
(non-rivalry), and the marginal cost of exclusion--of stopping an individual from enjoying
43
the good--is prohibitive (non-excludability.)* The atmosphere has both characteristics.
44
One country's greenhouse gas reductions affect all other countries as well. If this thwarts
14
1
climate change, all countries benefit³⁵.
2
Public Goods and Externalities
3
Different countries will be affected
4
differently, so that the benefits of avoiding
Externality: An externality, or spillover, arises
5
greenhouse gas warming will differ from
when the private costs or benefits of
production differ from the social costs or
6
country to country. Further, some actions
benefits. Because the social costs or benefits
7
will affect simultaneously local
are external to the private costs that firms face,
8
atmospheric conditions (providing local
private markets alone the economy will tend to
9
public goods) and greenhouse
produce too little of a public good (like
10
concentrations. For example, actions that
education) and too much of a public bad (like
11
reduce urban driving improve local air
pollution).
12
quality and at the same time reduce
Public good: A public good has two
13
greenhouse gas emissions³⁶.
properties: the marginal cost of an additional
14
individual using the good is zero, and the
15
1.3.4.1
Property rights
marginal cost of exclusion--of stopping an
16
individual from enjoying the good--is
17
An important strand of economic
prohibitive. Lighthouses are public goods:
thought associates externalities with a
when lighthouse services are provided to one
18
person, others may enjoy the same services
19
failure to assign property rights (Coase
without cost.
20
1990). Assigning property rights to the
21
atmosphere is particularly difficult, since
Market Failure: Private markets may
22
this would require the agreement of many
sometimes fail to provide a good at the most
23
sovereign states. Tradeable greenhouse
desirable level: private markets alone will likely
emission permits, discussed below, can be
provide too few lighthouses and too much
24
pollution.
25
thought of as an attempt to resolve the
26
problem by explicitly assigning property
27
rights to greenhouse emissions.
28
29
1.3.4.2
Paying For An International Public Good
30
31
Who should pay for a global public good? every country faces this question internally
32
in determining who should pay for the public goods it provides. Who should, for instance,
33
pay for pollution control within a country?
34
35
Economists generally agree on several principles:
36
37
First, concerns about equity should be separated from concerns about efficiency, though
38
ultimately they are intertwined. An implication of this principle is that because pollution is
39
a social cost of production (and consumption), everyone should be made to pay the full
40
social costs of the pollution they generate. Thus, if there is a social cost to a unit of
41
emissions of greenhouse gases, that cost is the same no matter who produces the emissions.
42
All should pay the full social costs of their actions, whether rich or poor. In this
43
perspective, corrective (Pigouvian) taxes should be imposed uniformly.
44
15
I
Second, it is inappropriate to redress all equity issues through climate change initiatives,
2
although climate change should not aggravate disparities between one region and another.
3
4
Economists agree less about the framework for deciding the burden of financing
5
mitigation and adaptation. At least four approaches have been proposed to determine how
6
the burdens of taxation should be shared. One approach looks at benefits: just as those who
7
benefit from private goods must pay for them, those who benefit from a public good should
8
be made to pay for them. The principle has some force when large differences in
9
preferences exist within any income class; providing a particular public good benefits some
10
of those individuals more than others, creating inequalities in the absence of benefit taxes.
11
A major problem frequently encountered, particularly for pure public goods, is that it may
12
be difficult to determine who benefits. It is in general possible to ascertain the economic
13
benefits of mitigation, and these are likely to be quite unequally distributed. But this
14
principle, by itself, does not fully determine who should bear the costs: appropriately
15
designed mitigation strategies will produce a surplus of benefits over costs, which must
16
somehow be divided.
17
18
A second approach looks at ability to pay. It is often held that richer countries (or
19
individuals) should pay more than poorer ones. This approach sometimes rests on the claim
20
that all people are entitled to a certain minimum consumption (Dasgupta 1982). But this
21
principle does not answer the question of how much extra the richer countries should pay.
22
23
A third approach is based on contribution to the problem. Because the North has
24
contributed more than two-thirds of the stock of greenhouse gases in the atmosphere today,
25
this approach seems to suggest a large responsibility for the North. On the other hand, by
26
the time greenhouse gas concentrations double from preindustrial levels, the developing
27
countries are projected to be contributing more than half of annual emissions, and roughly
28
half of the total stock in the atmosphere (IPCC, 1990; Cline 1992). Thus under this
29
criterion, the developing countries might eventually pay far more of the mitigation costs
30
than under the other principles described earlier.
31
32
Economists have turned to a fourth approach -- the social welfare function -- to answer
33
the question of how much extra different parties should pay, as well as the question of how
34
to distribute the surplus. The discussion of equity below differentiates between the
35
Rawlsian and utilitarian approaches. In this case, both approaches yield similar results:
36
In the absence of incentive problems, both imply that all of the surplus should be allocated
37
to the poorer countries, or that the burden of effort be borne by the richer countries³⁷.
38
39
Yet a different approach holds that social scientists as such have nothing to say about
40
these ethical issues. Coase [1960], for instance, approaches the problem of externalities by
41
emphasizing (a) the importance of assigning property rights, so that, in the absence of
42
bargaining costs, an efficient solution can be obtained³ and (b) that the outcome is
43
independent of how property rights are assigned³⁹ Coase also emphasizes the importance
44
of transactions costs, which will often influence the choice of policies.
16
1
A simple approach that yields efficiency but does not require redistribution (and is thus
2
consistent with the two principles enunciated above) requires coordinated tax rates so that
3
all countries face the same energy prices. This approach makes the cost of emitting an extra
4
ton of carbon equal across all countries, and each country retaining the revenues thus
5
generated⁴⁰. The net cost of such a tax (ignoring the benefits from reduced greenhouse
6
gas emissions) will, in general, be smaller for poorer countries, as a percentage of their
7
national output: the burden of the tax is progressive, even though the tax is levied at the
8
same rate in all countries⁴¹.
9
10
Accounting for past damages. The Framework Convention on Climate Change directs
11
the Annex I countries to take the lead in responding to the threat of climate change. Some
12
have argued in addition that because the North has been the major contributor to current
13
levels of greenhouse gases, it should bear most of the costs of mitigation. This view says
14
that costs should be borne not in proportion to benefits expected, but in proportion to
15
contribution to pollution. This argument, however, is not based on efficiency, which
16
requires that incentives be prospective (forward looking), not retrospective⁴. No incentive
17
effects attach to imposing charges based on past actions. Whether to charge nations that
18
contributed CO₂ to the atmosphere is an issue of ethics, not efficiency. Moreover, many
19
have challenged the ethical basis for assigning responsibility based on past damages⁴³.
20
21
The controversial issue of population growth, while central to the question of economic
22
development, bears on climate change largely through its effect on emissions. This chapter
23
will limit its treatment of population to this one effect⁴⁴.
24
25
1.3.4.3 Enforcement
26
27
Both externalities and public goods need a legal framework within which the problems
28
they pose can be addressed. Without compulsory taxation, there is an incentive for each
29
individual to be a free rider, though there is some empirical evidence that the free rider
30
effect may not be as significant as economists have previously assumed (Bohm [date]). In
31
the absence of compulsory taxation, externalities can only be addressed with well defined
32
property rights (Coase 1990)⁴⁵ and a legal system that enforces compensation for
33
externalities.
34
35
Though in many areas, great strides have been made in the development of an
36
international rule of law, problems are likely to face any system for enforcing climate
37
change agreements. For example, the incentive for free riding is great, whether through
38
delay, nonagreement, or noncompliance. Several alternatives have been proposed, each
39
raising political or economic concerns. No consensus now exists about the relative
40
desirability of these alternatives, though it is clear that some sort of international judicial
41
process will be required to determine and enforce compliance⁴⁶.
42
43
44
17
1
1.3.4.4
Knowledge
2
3
A key element in addressing the problem of climate change is an increase in
4
knowledge--knowledge about climate science as well as about the economic and social
5
aspects of impacts, mitigation, and adaptation. Much of this knowledge is in the nature of
6
an international public good. Developing ways of increasing energy efficiency will benefit
7
all countries. While in some cases, those engaged in the research will be able to
8
appropriate for themselves a significant fraction of the private benefits (mostly in reduced
9
energy costs), they will not be able to appropriate the broader social benefits, unless there is
10
a sufficiently high energy tax (or permit fee) that the price fully internalizes the emissions
11
externality. But even then, the social benefits of innovation tend to far exceed the private
12
benefits.
13
14
This suggests the need for an international agreement to fund basic research and
15
subsidize applied research, particularly in energy-related technologies for the developing
16
countries. There need not be a central funding agency, nor a central directorate
17
determining which research should be undertaken. But some mechanism, possibly
18
including joint implementation, must exist for sharing research plans results, and for
19
ensuring that the fruits of this research are made freely available.
20
21
1.3.5 Efficiency
22
23
With some exceptions noted below, issues of efficiency and equity can be separated⁴
24
Analysts agree that any actions responding to climate change should be cost-effective: no
25
matter who bears the cost of emission reductions, reductions should occur where their cost
26
is lowest⁴ Because of the low energy efficiency in many developing countries, many
27
have proposed that the major effort at the emission reductions be concentrated there.
28
29
Mechanisms for reducing emissions equitably and efficiently, including joint
30
implementation, tradeable permits, and coordinated tax policies, are discussed below and in
31
chapter 11. All of these approaches, however, attempt to confront all individuals and
32
producers in all countries with the same cost of emissions. Emission control is an
33
international public good in that greenhouse emission reductions have the same effect,
34
wherever they occur. Just as efficiency in the production of steel or any other commodity
35
requires that all consumers and producers face the same price, so too with emissions.
36
Uniform prices can be achieved either through coordinated energy taxes or through
37
tradeable permit requirements; but unless the rules are applied in a systematic way to both
38
developed and developing countries, emission reductions will be inefficient.
39
40
Partial participation in an international emissions reduction program, however, will
41
significantly reduce its effectiveness. The growth of international trade has resulted in
42
important links between the developed and developing countries, and the total effects of any
43
policy undertaken in the North can only be evaluated in terms taking into account the
44
systemic responses, including responses from the South, as discussed under "General
18
1
Equilibrium," below⁴⁹.
2
3
1.3.5.1
Bankable permits
4
5
Efficiency imposes several requirements. One, just described, is that at any moment, the
6
costs of reducing emissions should be minimized. The second is intertemporal efficiency:
7
the marginal cost of reducing emissions at two points in time must be the same. If it will
8
cost less to reduce emissions at some future date, adjusting for time discounting, option
9
values (risk), and impacts on atmospheric concentrations, then the reduction schedule
10
should be adjusted accordingly.
11
12
Intertemporal efficiency would be promoted by allowing banking of permits (allowing a
13
source to use fewer permits in one year and more in another), and by the development of
14
futures and options markets. A bankable permit system would address some North-South
15
equity issues, including the concern in the South that delays in mitigation now by the North
16
would leave a greater burden for the future.
17
18
1.3.5.2
Exchange/risk efficiency.
19
20
When different parties to an agreement hold widely differing views about risk and the
21
probability of loss, significant efficiency gains can result from transferring risk among
22
parties. An earlier section discussed the importance of establishing an international
23
insurance market for those facing the threat of losses under climate change, and noted the
24
advantages of establishing a market within which countries who are less concerned about
25
the economic risks associated with climate change can assume more of the insurance
26
burden. Any efficient international system for addressing the problems of global climate
27
change must include both mitigation and insurance obligations. Governments that believe
28
they have a comparative advantage in assuming climate risks can assume a larger share of
29
those risks, trading off other obligations, and substantially reducing overall costs of
30
responding to climate change.
31
32
1.3.5.3
Comprehensiveness
33
34
Efficiency also requires that the cost of reducing all greenhouse emissions be minimized.
35
This principle implies that any mitigation program must include all greenhouse gases
36
(taking into consideration their heat-trapping potentials and atmospheric lifetimes⁵⁰), and
37
must include carbon sinks.
38
39
Finally, and perhaps most controversially, it implies that mitigation strategies focus on
40
all elements of the Kaya identity above, ensuring that the marginal cost of reductions be the
41
same for each of the possible strategies. Thus, population control may be an element in a
42
long term mitigation strategy no less than a shift in the composition of production or an
43
increase in the energy efficiency of the economy.
44
19
1
1.3.6 General Equilibrium
2
3
General equilibrium theory, a cornerstone of economic research over the past century,
4
demonstrates the advantage of looking beyond first-stage effects. It offers two important
5
insights for climate change analysis. First, the various parts of an economic system are
6
interrelated; perturbations to one part have ramifications for other parts, which may be quite
7
distant. Second, when all of the reverberations are taken into account, the net effect of an
8
action may be markedly different from the initial (and intended) effect.
9
10
One implication of general equilibrium theory has already been noted: taxes imposed on
11
one part of the global economy may have little if any effect on global emissions; they may
12
simply result in a relocation of economic activity. Increasingly, the world's economic
13
system must be viewed from a global perspective. Location of economic activities is
14
determined primarily by relative factor prices, taking into account certain specific locational
15
advantages and specialized competencies. If, for example, the OECD countries impose
16
carbon taxes on energy-intensive industries, those industries may relocate outside the
17
OECD. Further, if greenhouse mitigation puts an economic drag on the developed
18
countries, developing countries would be affected through trade.
19
20
If different countries have different obligations to reduce greenhouse emissions, different
21
implicit tax rates will result. This will interfere with world economic efficiency--decreasing
22
world real output--possibly with little effect on total greenhouse gas emissions. For
23
example, the most energy intensive activities--such as aluminum production--may well
24
relocate to developing countries⁵¹.
25
26
While many of the energy-economy-carbon models described in subsequent chapters
27
attempt to estimate the magnitude of carbon leaks, most are based on standard international
28
trade models, in which the location of production of various goods and machines is fixed
29
(Whalley and Wigle, 1992)⁵². Thus, estimates of carbon "leakages" are based only on
30
commodity substitution⁵³. But allowing for relocation of industries -- as would occur in
31
the very long run, the time span of interest for an analysis of climate change-- leakages
32
may well be higher. No consensus now exists on the magnitude of long-run leakages.
33
34
1.3.6.1
Intertemporal Substitution
35
36
General equilibrium issues also arise when production can be shifted from one period to
37
another. For example, in a partial equilibrium analysis, a tax on gas or oil raises the price
38
of the fuel taxed, thereby reducing its consumption, and thus associated emissions. But an
39
exhaustible natural resource like gas or oil has, over the long run, by definition an inelastic
40
supply (ignoring for the moment extraction costs, which, in the case of gas and oil, are
41
small relative to the price). The general theory of incidence argues that when a commodity
42
is in inelastic supply, a tax affects the price, but not the level of consumption. That is,
43
when all countries impose a tax on gas or oil, the price of oil falls, by an amount just equal
44
to the tax. The full incidence falls on producers; as a first approximation, the level of
20
1
consumption--and thus the level of emissions--remains unchanged. This provides a first
2
approximation. Obviously, if the tax is large enough, the price will fall below the cost of
3
extraction, and supply will be reduced. Moreover, in the case of exhaustible natural
4
resources, taxes may affect the timing of consumption of the oil and gas (Stiglitz, 1977,
5
1979,
,
1993)⁵⁴.
6
7
8
1.4
EQUITY
9
10
Which nations will be allowed to increase their greenhouse emissions, and who pays for
11
greenhouse gas abatement and adaptation are among the most contentious issues in climate
12
change. This has immediate implications for policy as well, because the initial allocation of
13
emission rights and emission constraints will largely determine the distribution of costs.
14
The Framework Convention on Climate Change explicitly directs the parties to consider
15
equity:
16
17
The Parties should protect the climate system for the benefit of present and future
18
generations of humankind, on the basis of equity and in accordance with their common
19
but differentiated responsibilities and respective capabilities. Accordingly, the developed
20
country Parties should take the lead in combating climate change and the adverse effects
21
thereof
22
23
[Account must be taken of] the differences in their starting points and approaches,
24
economic structures and resource bases, the need to maintain strong and sustainable
25
economic growth, available technologies and other individual circumstances, as well as
26
the need for equitable and appropriate contributions by each of these Parties to the
27
global effort regarding [the Convention's] objective⁵⁵
28
29
30
1.4.1 General Issues
31
32
Within the limits of cost-benefit analysis, equity arises because of the principle of
33
compensation, discussed in section 1.3, above. For example, suppose it could be shown
34
that a business-as-usual path produced higher total benefits than a path with lower
35
greenhouse emissions. We could not from this evidence conclude from this that the world
36
as a whole would be better off. Suppose that costs of warming fall predominantly on one
37
group or one generation, while benefits accrue to another. (For example, some have
38
speculated that the costs of damages from warming would fall largely on the South, without
39
a compensating increase in benefits (Parikh date).) Unless the gainers actually compensate
40
the losers for their losses, cost-benefit analysis cannot conclude that the change has been on
41
balance good for society. Compensation is particularly difficult if future generations would
42
bear most of the costs, because no "Fund for Future Greenhouse Victims" exists (Steer
43
1992). Thus, some have argued that in the absence of mechanisms to make these transfers,
44
we should not rely on possible future transfers from gainers to losers, but should instead
21
I
insist that the gainers pay the costs up front. An alternative explanation addresses these
2
equity concerns by assigning different weights, perhaps based on economic status, to
3
changes in consumption of different individuals (Atkinson date).
4
5
No consensus exists among either economists or philosophers about the appropriate
6
ethical responses to the changes that would come with climate change. Should, for
7
instance, owners of resources be compensated for the losses that they incur as a result of
8
mitigation actions and the consequent change in prices or values? Economists have often
9
argued no, for several reasons. First, some wealth is not a reward for productive activity,
10
but merely an accident. It is not because country X did something that $200 billion worth
11
of minerals or oil was discovered to lie beneath its territory. Many would say that these
12
random allocations of wealth have actually contributed to world inequality, and that
13
eliminating a part of these windfall gains would, from the perspective of an egalitarian
14
social welfare function, be welfare-increasing. Another view denies that government
15
policies would be taking away value from assets that by right should be there; until now,
16
according to this view, these resources have simply been overpriced, not fully reflecting the
17
social costs imposed by their use. In this view, actions to discourage overuse would simply
18
rectify a previous mistake. 56
19
20
Whether investors or countries should be compensated for the adverse effects on their
21
market values remains controversial. There is, however, reasonable consensus on three
22
general principles: First, workers in adversely affected sectors may need assistance to
23
switch occupations (The market failure here is that workers cannot purchase insurance to
24
insure themselves against these kinds of adverse shocks.) Second, gradual transitions may
25
significantly lower the absolute cost of the transition. For instance, workers leave jobs
26
through natural attrition; if those leaving are not replaced, the industry will be scaled down,
27
with no transition cost to any individual worker. Third, the magnitudes of the
28
uncompensated redistributions associated with any change in policy are often correlated
29
with the magnitude of the political opposition.
30
31
Most policy changes produce winners as well as losers. If the relative price of natural
32
gas increases, owners of natural gas deposits may actually be better off. In economies with
33
progressive taxation, either of capital gains or consumption, then implicitly some part of
34
those gains are shared more broadly.
35
36
Virtually all policies discussed below also have different effects on different groups.
37
Residents of very hot and very cold climates consume more energy for heating and cooling,
38
and thus would be worse off, relative to those in more moderate climates. In some
39
countries, city dwellers can choose less energy-intensive modes of transport than those in
40
the countryside.
41
42
Many of these impacts will be reflected in capital costs. Thus, the value of land in
43
temperate climates is likely to rise, and the value in extreme climates to fall. (Similar
44
points can be raised, of course, about the costs imposed by climate change itself.) This
22
I
capitalization effect has both favorable and unfavorable implications. In the long run,
2
residents of colder climates are likely to consume less heating fuel, but perhaps at a higher
3
price, which would leave them on balance about where they started. Energy prices are
4
likely to rise, and land rents to fall, in an almost offsetting way. On the other hand, current
5
owners of land would bear the full brunt of the present discounted value of all future
6
increases in taxes (or the tax-equivalent cost of regulations designed to reduce energy
7
utilization.) As a result, unless policy changes are introduced gradually, dramatic changes
8
in land values may occur, with possibly large effects on financial institutions and the
9
economy as a whole. Anticipation of these policies would partially offset these dramatic
10
changes in land values⁵⁷.
11
12
National Security. Climate change itself may affect the national security of many
13
countries. At the same time, policies to reduce greenhouse emissions may affect export
14
earnings and therefore national security of the energy exporting countries.
15
16
While countries may be willing (or forced) to accept changes in national wealth as a
17
result of changes in world prices induced by the response to the threat of climate change,
18
countries are less likely to be willing or able to accept what they may perceive as implied
19
threats to national security over which they have some control. Thus, a country with a
20
large endowment of coal becomes more vulnerable if it comes to rely on imported oil or
21
gas--supplies of which could be cut off in times of war. These countries may feel it
22
imprudent to switch, even if private economic gains were to be had; and they are
23
particularly unlikely to switch if the benefits take the form of an international public good.
24
25
Increasing world political stability would clearly address these concerns. But even were
26
that successful, it would not suffice. To increase national security, policies will need to
27
focus on increasing energy efficiency and reducing energy demand. This is an example of
28
the necessity in designing an international structure for greenhouse gas emissions for
29
allowing sufficient flexibility that the particular circumstances of each country can be
30
appropriately taken into account.
31
32
Benchmarks. The earlier discussion of equitable distribution of the burdens of
33
responding to greenhouse gases, employing generally accepted principles of public finance,
34
avoided the concept of benchmarks, which have played an important role in international
35
negotiations. Indeed, the "obligations" assumed under the Framework Convention obligates
36
the developed countries to return emissions to 1990 levels by the year 2000. One
37
justification for this form of obligation is that it gives some sense of equity. The difficulty
38
of the task for each country was gauged by its 1990 emissions; requiring each to keep
39
within that target appeared to impose a proportionate burden on each. It was a rough sense
40
of equity, because no attention was paid to the relative incomes of developed countries and
41
therefore no attention was paid to the implied tax rate.
42
43
But in a more fundamental sense this criterion can be criticized as inequitable. It pays
44
no attention to past efforts at achieving energy efficiency, nor to other circumstances that
23
1
might affect the implied tax rate. For instance, a country which, during the preceding ten
2
years, had made every effort to increase energy efficiency and switch consumption to less
3
energy-intensive commodities would face the burden of reducing its emissions still further.
4
Because the marginal cost curve for emissions reductions rises steeply beyond a certain
5
point, the implied tax rate would be considerably higher than for a country that had
6
previously encouraged overconsumption of energy, for example by energy subsidies. For
7
the second country, achieving the emission targets might require only eliminating the
8
energy subsidies, a policy with an implied negative tax rate. For a similar reason, countries
9
with large endowments of hydroelectric power may find it relatively difficult to meet an
10
emission target of this form.
11
12
Public finance theory has focused extensively on "second-best" policies, recognizing the
13
difficulty of achieving first-best objectives of either economic efficiency or distributive
14
justice. In this context, the central issue is whether there exist alternatives to
15
benchmarking, setting target emission reductions in relation to past emissions. When the
16
U.S. government recently issued tradeable emission permits for sulfur dioxide, it took
17
account of emission reductions already achieved. Benchmarking reflects information that
18
would not be reflected in a simple criteria, such as a particular emission level per unit
19
population or GDP. Thus, a strong case can be made for including benchmarking, if not in
20
the final allocation of permits (obligations), at least in the transition rules.
21
22
1.4.1 Intergenerational Equity
23
24
Efforts to control greenhouse emissions will provide benefits primarily for our
25
grandchildren and their descendants. We face a difficult task in estimating and judging
26
what aspects of climate and environment they will value and how best to preserve those
27
aspects for them. If we take aggressive action to limit climate change, they may regret that
28
we did not use the funds to push ahead development in Africa, to better protect the species
29
against the next retrovirus, or to dispose of nuclear materials safely. Chapter 3 addresses
30
directly the most important issue in intergenerational equity: choice of an appropriate
31
discount rate.
32
33
This argument also applies to actions with differential impacts on different regions. If
34
greenhouse warming turns out to be a major threat to countries of the South, and if the
35
North fails to reduce emissions aggressively now, countries of the South might suffer
36
additional damage later. Alternatively, if the North chooses to embark on an aggressive
37
control regime now, and if this cuts into Northern growth rates, the result would shrink
38
exports markets for the South, and thus reduce growth there; in addition, if the North views
39
its greenhouse efforts as, in effect, development aid for the South, it might cut back on
40
other programs (sanitation, water, education for women, etc.) with a greater impact on life
41
expectancy, health, and well-being.
42
43
44
24
1
1.4.2 Within-country Equity
2
3
While most discussions of equity and climate change have so far focused on North/South
4
issues or on issues between one country and another, issues of equity within a country are
5
also important, and indeed play a central role in the political debates about appropriate
6
responses to climate change. Most policy recommendations involve large within-country
7
losses for certain groups. For instance, any policy leading to less use of coal, and lower
8
producer prices for coal, will lead to large losses for coal mine owners and workers⁵.
9
10
The net efficiency gains (in reduced emissions) relative to the distributive effects may
11
differ markedly across resources. Thus, if elasticity of world oil supply is small, a tax on
12
oil will be reflected in the prices received by producers, and have little effect on cumulative
13
consumption of oil, though it may result in some short-run substitution against coal.
14
Policies aimed at increasing the date of arrival of substitutes for fossil fuels could lead to an
15
increase in current emissions, though long-run effects on atmospheric concentrations would
16
be positive.
17
18
19
1.5
ECONOMICS OF POLICY ACTIONS
20
21
Earlier sections set forth a basic framework for analyzing policies related to global
22
climate change, including a combination of mitigation, adaptation, and possibly climate
23
engineering. Striking the appropriate balance requires taking into account the costs,
24
benefits, and risks associated with each strategy. For instance, if risk were irrelevant,
25
governments should reduce emissions to the point at which a dollar of extra spending
26
would yield a dollar of expected savings in damages imposed by climate change, or save an
27
extra dollar of expected costs of adaptation. Adding risk and sequential decision making
28
complicates the analysis but leaves unchanged the basic principles. Because of the lasting
29
impact of climate change, and the magnitude of the resulting economic uncertainties, most
30
policy analysis has focused on a narrower set of questions:
31
32
1. What actions would improve economic efficiency (including the social costs of
33
implementing the policy) and reduce net greenhouse gas emissions? How much
34
emission reductions could be achieved by these means?
35
36
2. Beyond these zero-cost options, what are the least-cost methods of reducing
37
greenhouse gas emissions? What does the cost curve look like⁵⁹? What are the
38
alternative policy measures, and how do they compare?
39
40
3. What are the essential ingredients of an adaptation strategy, and to what extent will
41
market forces, on their own, provide the appropriate adaptive responses?
42
43
44
25
1
1.5.1 Zero-cost Options
2
3
A variety of inefficiencies in the energy sector--many of them government
4
induced--would, if eliminated, increase economic efficiency and reduce greenhouse gas
5
emissions at the same time. How large is the reservoir of conservation opportunities?
6
Proponents of the two major approaches to the question have debated this point for more
7
than a decade. Top-down models extrapolate observed behavior into the future. Bottom-up
8
models, combine engineering analyses to estimate costs with economic models of individual
9
choice. Top-down models generally show significant costs to reducing greenhouse
10
emissions in the future⁶⁰. Bottom-up, or technology-specified models, have been used to
11
show the existence of significant reductions in the cost of energy as new low emissions
12
technologies are adopted. Some proponents of bottom-up models argue that emission
13
reductions can be achieved at essentially no cost, seemingly repealing the long standing
14
economists' presumption that there is no free lunch⁶¹.
15
16
Much of the disagreement turns on empirical estimates. Economists have catalogued the
17
unintended consequences of government regulation. Many have also identified important
18
market failures that could give rise to inefficiencies within the private sector itself. The
19
next two sections will examine each of these effects.
20
21
1.5.2 Policy Reform
22
23
A variety of government reforms could enhance energy efficiency, including removing
24
energy subsidies, reforming or clarifying property rights, reducing non-greenhouse gas
25
externalities, and administrative reforms.
26
27
1.5.2.1
Removing Energy Subsidies
28
29
Energy subsidies induce inefficient energy use, reducing the total output of the economy
30
as well as increasing CO2 emissions. These subsidies are especially prevalent in
31
developing countries⁶. Shah and Larsen (1991) estimated world energy subsidies in 1990
32
to have been $230 billion, calculating that their elimination would reduce global carbon
33
emissions by 9 1/2% in addition to improving allocative efficiency thereby generating a
34
welfare gain in subsidizing countries. Burniaux et al. (1992) obtained similar results using
35
the GREEN model, concluding that the elimination of all existing distortions on energy
36
markets would yield an increase in world real income of 0.7% per year in addition to
37
cutting world emissions by 18% in 2050. (Dean 1994) Likewise, Unterwurzacher and Wirl
38
(1991) estimated that raising energy prices to world levels in Poland, Hungary, and
39
Czechoslovakia would reduce carbon emissions from those countries by 30%. Agricultural
40
subsidies also distort the outcome, especially affecting the size of forests.
41
42
1.5.2.2
Property Rights Reform
43
44
One responsibility of governments is to define property rights and enforce contracts.
26
1
Ill-defined property rights encourage overconsumption of resources. A clearer definition of
2
property rights could be particularly important in helping to decrease deforestation, for
3
example, while improving economic efficiency. Uncertainties about future property rights
4
may also contribute to economic inefficiency. Thus, for example, in developing countries
5
where large forests are owned by a few large landowners, excessive deforestation may
6
result from landowners' fear that their tenure will be limited.
7
8
1.5.2.3 Administrative Reforms
9
10
Defining property rights and eliminating energy subsidies are two important actions
11
governments can take to reduce greenhouse emissions. At the same time, many less
12
sweeping reforms can improve economic efficiency and simultaneously reduce greenhouse
13
gas emissions. For example:
14
15
Pricing of government produced electricity. Many governments price electricity not at the
16
market price but at the cost of production. Economists generally recommend that
17
electricity, like any good, be priced not at its cost of production, but at the competitive
18
price. In countries with a mix of plants, this means that electricity from all sources should
19
be priced the same--at the highest marginal cost of production⁶³
20
21
Land use and other regulation. Changes in land use policy can also reduce energy
22
consumption (especially for transportation and space conditioning), and thus greenhouse gas
23
emissions⁶⁴
24
25
1.5.2.4
Regulating Non-greenhouse Externalities
26
27
Many activities producing greenhouse emissions also generate pollution of other types.
28
For example, fossil fuel combustion releases conventional air pollutants; rush-hour auto use
29
contributes to road congestion. In the presence of these spillover effects or externalities,
30
market solutions will not properly reflect the externalities generated, leading to the
31
overconsumption of environmental resources. Energy taxes, congestion pricing, or tradable
32
permits can correct these market signals, resulting in lower emissions of both greenhouse
33
gases and other pollutants. Some reforms, such as congestion pricing, also reduce the need
34
for roads and other physical capacity.
35
36
1.5.2.5
Special Problems of Economies in Transition
37
38
The economies in transition provide special opportunities for mitigating greenhouse
39
emissions. In the old Soviet bloc, high energy subsidies and other price distortions affected
40
energy usage directly, as well as indirectly through the composition of output. Spotty
41
environmental regulation meant that Eastern Bloc nations lacked the environmental controls
42
common in the OECD. The capital shortage of the past decade has contributed to the
43
problem, through a general deterioration of physical capital stock.
44
27
1
While these problems are largely of governments' making, the remedy is likely to rely
2
on a combination of public and private actions: effective environmental regulation,
3
elimination of government-caused price distortions, and an economic environment in which
4
foreign and domestic investment can enhance the efficiency (including energy efficiency) of
5
the economy. For example, many analysts believe that cutting methane leakage from gas
6
pipelines will yield both high economic and cost-effective reductions in greenhouse
7
emissions.
8
9
1.5.2.6
Examples of Efficiency-Increasing Policies
10
11
Policies that cause individuals not to take into account the full social costs of their
12
actions often result in greater energy use and greenhouse emissions. The National Action
13
Plans of many countries have revealed examples of such policies, and have suggested
14
remedies, including:
15
16
Unit pricing of waste disposal to encourage recycling: The life-cycle social cost of
17
consuming a good includes its costs of production plus disposal. Most consumers, and
18
many businesses, pay a flat fee for trash disposal; with a flat fee, the marginal cost of
19
throwing away an extra pound of trash is zero. By moving from flat fees to unit pricing,
20
the actual price consumers pay to buy and dispose of a good will more closely match its
21
full life-cycle social cost.
22
23
Pay-at-the-pump automobile insurance: In most countries, drivers pay automobile
24
insurance yearly or monthly. Once the premium is paid, the marginal insurance cost of
25
driving an extra mile is zero, even though driving more does increase the chance of being
26
in an accident. As an alternative, drivers could be required to pay a portion of their
27
insurance bill at refueling. With pay-at-the-pump insurance, a tax would be levied on
28
gasoline and earmarked to pay for insurance premiums. This would raise the cost of
29
gasoline at the pump, but lower auto insurance premiums.
30
31
Subsidized parking. In some industrialized countries, many employers provide parking to
32
employees at no cost or lower-than-market cost, thus lowering the relative price of
33
commuting by car relative to public transportation. A distortion arises when governments
34
tax income spent on public transportation, but not income implicit in the parking subsidy.
35
36
Subsidized housing. In some industrialized countries, home mortgage interest is
37
tax-deductible. In all but a few countries, the implicit income on owner-occupied housing
38
is not taxed. These tax provisions encourage individuals to consume more housing than
39
they otherwise would. In cold or hot climates, where more housing space requires more
40
energy, this tax treatment increases CO2 emissions.
41
42
Subsidized trucking. Studies suggest that virtually all road damage is caused by heavy
43
trucks, which pay only a portion of the expense of building and maintaining the road
44
system. Many countries thus subsidize trucking compared with rail or barge transport,
28
1
probably increasing greenhouse emissions⁶.
2
3
1.5.3 Market Failures and Government Responses
4
5
Economists agree that governments can adopt policies to increase economic efficiency at
6
the same time that they reduce greenhouse emissions. For example, in some countries, fuel
7
prices do not reflect the full social cost of fuel burning; taxes can correct this market
8
failure. There is less agreement about whether, given market prices, firms fail to take
9
advantage of all of the energy efficiency opportunities available to them. This controversy
10
underlies the bottom-up versus top-down controversy treated at greater length in Chapter 8.
11
Engineers have identified a host of seemingly profitable actions that would also save
12
energy. Many economists, however, view this as evidence that the engineering analysis has
13
omitted characteristics important to consumers.
14
15
The substantial differences in practices both within and between countries suggests scope
16
for substantially increasing energy efficiency. Moreover, even best practices within a
17
country may not put it at the technological frontier. In deciding whether to adopt a new
18
production process, businessmen look only at the private costs and benefits. Many
19
technologists, however, conclude that even considering private costs only, firms should be
20
undertaking many energy-efficiency improvements. This section attempts to reconcile the
21
different schools of thought by reference to information-based market imperfections, as well
22
as by the criteria by which businesses make decisions.
23
24
Information dissemination. Acquiring information is costly; providing and disseminating
25
information has many features of a public good (Stiglitz 1988, Romer 19 , Arrow ). In
26
the absence of government intervention, there will be too little production and dissemination
27
of information. This is particularly true for information with widely dispersed impacts, as
28
opposed to information about, for example, the production of certain chemicals, which is
29
primarily of value to a few companies⁶.
30
Moreover, both theory and evidence support the view that markets, on their own, do not
31
provide an efficient level of disclosure of information⁶. Indeed, some evidence indicates
32
that markets may try to obfuscate relevant information.66 This provides the rationale for
33
government provision of information, or laws that in many countries require disclosure of
34
interest rates and other consumer-relevant information, including appliance energy
35
consumption.
36
37
Bureaucratic structure, limited scope of attention. In recent years, economic and
38
organizational theory⁶⁹ has emphasized that large organizations are not, in general, run by
39
owners; that the managers, even with the best designed incentives, do not in general
40
maximize the firm's market value; and that one of the principal scarce factors within an
41
organization is time and attention⁷⁰. How managers direct their attention has much to do
42
with what the firm does.⁷¹ The information disclosures noted in the previous subsection,
43
as well as a number of other government programs focusing on energy efficiency in
44
consumer products, electric lights, and motors, help focus management attention on energy
29
1
efficiency. The marginal managerial time required to make efficient energy decisions may
2
be small and focusing attention on this issue--when information is being freely provided
3
through government and other sources--may thus yield private returns well beyond these
4
slight marginal costs.
5
6
Returns to Scale and System Effects (network externalities): Some technologies might be
7
economically attractive at a large scale of production, but not on the much smaller scale on
8
which they might initially be adopted. Other technologies exhibit dynamic scale
9
economies: unit cost falls over time as a function of the cumulative output of firms or
10
industries. Technology "networks" may also affect diffusion rates. For example, cars and
11
trucks fueled with electricity, natural gas, methanol, etc., require a refueling infrastructure,
12
which itself competes for resources with the conventional fuel infrastructure already in
13
place⁷².
14
15
Building codes can be justified both in terms of these network externalities effects and in
16
terms of information failures. Consumers often have limited information concerning the
17
construction of their houses, and obtaining the information after the house is completed is
18
often difficult. Even were they to be provided with construction details, they would have
19
difficulty interpreting the implications.
20
21
Capital market imperfections. A major explanation of the difference between best-practice
22
and actual-practice technology is that bottom-up models often compute cost-effectiveness
23
using a discount rate substantially lower than the cost of capital calculated by firms.73
24
Studies of implicit discount rates consistently show that households and firms use discount
25
rates substantially above market rate for long-term government bonds.
26
27
Two explanations have been offered: 1) Risk: Interest rates facing firms and households
28
reflect the risk premium that lenders require to compensate them for the probability of
29
default. Firms often use discount rates that include a risk premium to reflect the riskiness
30
of projects; 2) Capital constraints: Individuals and firms often face rationing in capital
31
markets, both for credit and equity. Recent research has provided a rationale for this
32
rationing based on the fact that information is imperfect and costly.⁷⁴
33
34
These capital market problems have one important implication: Models analyzing
35
best-practice cost-effective technologies using discount rates lower than those typically
36
employed by firms will overestimate the rate of dissemination of these technologies and
37
underestimate the perceived costs (to the firms and households adopting these technologies)
38
of mitigation strategies.
39
40
But these capital market problems raise three other questions: (1) Are firms rational in
41
using such high discount rates? (2) Does the use of such high discount rates imply a
42
market failure? (3) If so, will government intervention improve on the market outcome?
43
44
Economists emphasize that an analysis of the costs and benefits of a project must
30
1
separate four issues: timing, risk, capital constraints, and information. Discount rates are
2
only to be used for timing. Risk should be treated by converting costs and benefits into
3
certainty equivalents, then discounting costs and benefits for each year at the relevant
4
discount rate. Higher risks should not result in higher discount rates.⁷⁶ Similarly,
5
capital constraints should be reflected in the shadow price of capital, not in the discount
6
rate.⁷⁷ Because of limited information (and a version of the "winners' curse"78) firms
7
often require threshold rates of return significantly greater than the market rate of interest.
8
In doing so, they may confuse time and information risk: that is, the rules of thumb firms
9
use to evaluate investments may sometimes lead to market inefficiencies, including perhaps
10
in the area of energy-efficient technologies.
11
12
Even were firms to follow the economist's guidelines, in the presence of capital
13
constraints, market outcomes would not, in general, be socially efficient (i.e., they are not
14
constrained Pareto optimal.) There may be significant discrepancies between social and
15
private returns to investment (even apart from the externalities associated with greenhouse
16
gases or technological diffusion.) This provides part of the rationale for possible
17
government interventions in capital markets. Though these capital market imperfections
18
imply that there is no presumption that market allocations are efficient, there is no
19
consensus that they lead to significant underinvestment in energy-efficient technologies in
20
particular⁸⁰.
21
22
1.5.3.1 Revising national accounts
23
24
Some have suggested revising the conventional systems of national accounts to
25
incorporate full social pricing of resources. An early contribution suggested a new
26
measure of economic welfare based on consumption that increases quality of life (Tobin
27
and Nordhaus, 1972). These authors and others recognized that national income
28
accounting, widely adopted after World War II, measures aggregate income and expenditure
29
flows, but does not incorporate environmental costs and benefits.
30
31
Conventional national income accounting does not fully report three categories of
32
resource expenditures: a) defensive expenditures, either for pollution prevention before the
33
fact or for cleanup after the fact (although these expenditures are not separately reported,
34
they are counted in GDP); b) consumption of environmental goods (such as exhaustible
35
resources)81; and c) conflicting uses of environmental services (such as the atmosphere,
36
used by producers as an input into production, and by households as a consumption good).
37
38
One proposal would include in GDP the effect of changes in quality of the environment.
39
In Eastern Europe and the former Soviet Union, steady increases in reported post-war GDP
40
masked the effects of decades of environmental degradation; for part of that period,
41
environment-adjusted GDP almost certainly declined⁸².
42
43
However, important conceptual problems in defining levels and changes of
44
environmental assets, complicate the task of modifying national accounts. First, the stock
31
1
of natural resources has no obvious definition. While most geologists would agree on the
2
size of coal stocks - their location is known and their in situ value can be estimated this
3
cannot be said for oil or minerals⁸³. Second, environmental assets--such as air quality--
4
present another set of problems, because no market prices exist to value the asset.
5
6
Four approaches are commonly used to calculate changes in the natural environment
7
(Peskin and Lutz, 1990):
8
9
a) The environmental expenditure approach, used recently in the United States, subtracts
10
pollution abatement expenditures from GDP;
11
b) The physical accounting approach, used in Norway and France, establishes satellite
12
accounts using physical units of measurement to account for flows and stocks of
13
resources;
14
c) The depreciation approach adjusts gross and net product by subtracting out the value of
15
natural resource depletion (Repetto 1989; El Serafy ***); and
16
d) The comprehensive approach uses both physical measures and value (Statistical Office of
17
the United Nations, 1992).
18
19
Another measure of broad-based welfare, although not including environmental amenities, is
20
the UN's human development index or HDI (UNDP, 1990). The HDI gives a composite
21
measure of human development by combining three key indicators: longevity (measured by
22
life expectancy at birth), education (measured by adult literacy and mean years of
23
schooling), and income (real GDP per capita adjusted for purchasing power). Although the
24
HDI is not directly related to global environmental issues, both climate change and
25
abatement policies may affect it⁸⁴.
26
27
1.5.4 Innovation
28
29
Standard competitive analysis argues that given all required information and technology
30
market economies produce efficient outcomes. But recent economic analyses have shown
31
that, in general, market economies need not result in the efficient allocation of resources to
32
information production and dissemination, and to innovation. The first of these issues was
33
discussed earlier. The second is more complex.
34
35
In the absence of intellectual property rights, firms would have less incentive to
36
innovate. With standard patent terms, firms are not able to appropriate all the returns to
37
their innovative activity. Setting the optimal patent life involves balancing off the
38
inefficiencies resulting from the exercise of monopoly power during the duration of the
39
patent (static inefficiencies) with the increased incentives for innovation85. Largely
40
because innovators seldom appropriate all of the returns to their innovation, there is a
41
general consensus that markets provide insufficient incentives for R & D; and the greater
42
the spill-overs, the greater the undersupply.86 Since the spill-overs are likely to be greater
43
the more basic the research, this suggests a role for government in subsidizing basic and
44
near-basic research. In the same way, the high cost of establishing intellectual property
32
1
rights impedes the transfer of technology to developing countries.
2
3
Still, there is a general consensus among economists that the patent system provides a
4
better basis of financing applied research than do government grants, largely because of
5
difficulties government has in picking those most likely to produce high returns. The
6
question is, is there any reason to believe that any market failures, in terms of insufficient
7
level of innovation, are worse in this area than in other areas, i.e. are there any special
8
grounds for arguing for government R & D subsidies, provided the government has
9
corrected energy prices to reflect the externalities generated? Obviously, in the absence of
10
such corrections, market incentives to provide energy saving innovations will be distorted,
11
just as market incentives to adopt energy saving technologies are reduced⁸⁷. (Tradeable
12
permits have similar effects to corrective taxes, since firms will value reductions in
13
emissions from new technologies, since such reductions will require them to purchase fewer
14
permits.)
15
16
Innovation is important, because it provides perhaps the best opportunity for low-cost
17
methods of reducing emissions. Several studies have confirmed the impact of accelerated
18
deployment of advanced energy technologies on the future rate and timing of anthropogenic
19
climate change⁸⁸
20
21
1.5.5 Carbon Taxes and Tradable Permits.
22
23
Economic efficiency requires all agents in the economy to pay the full marginal social
24
costs of their actions. But firms and households are not charged for the additional warming
25
potential they add to the atmosphere, and so do not pay the full social costs they impose.
26
Two economic instruments can correct this market failure: carbon taxes and tradeable
27
permits.
28
29
A tradeable permit scheme involves a determination of the total level of permits and a
30
distribution of the initial allocation, with emission levels for any firm limited to the number
31
of permits held. The initial distribution may be made by an auction or allocation
32
according to benchmarks (e.g. per capita as of a given date), or by historical emission levels
33
("grandfathering"); alternatively, emission rights could be grandfathered in at current levels,
34
and gradually shifted over to a per capita allocation as of a given date.
35
36
Once permits are distributed among the regulated entities, a market is set up allowing
37
companies to buy and sell permits according to their plants' planned emissions. The cost of
38
production then includes not only the costs of conventional inputs, but also the costs of
39
additional permits to offset additional emissions. Plants whose cost of mitigation is low
40
will find it relatively easier to abate pollution rather than to buy permits. Plants with higher
41
costs of mitigation will have a greater preference for buying permits than for abating
42
pollution. The price of the permits, which are artificially created scarce resources, is
43
determined by the market. With the use of tradable permits, companies have an incentive
44
to improve the efficiency of their production, thereby reducing their emissions level, as they
33
1
can sell excess permits on the market and generate revenue. 89
2
3
While permits thus create a marginal cost of production related to the marginal
4
emissions, carbon taxes impose a tax directly on the marginal emissions. Both systems thus
5
force producers and households to face the true social costs of their actions⁹⁰. Permits are
6
better for regulating large sources; carbon taxes are better for small sources. In principle,
7
both could be adjusted to achieve the desired level of emissions desired, although
8
adjustments of this sort are likely to be difficult in practice.
9
10
The initial allocation of permits will largely determine the distribution of costs of
11
abatement (Chapter 3 discusses these issues of equity at greater length), and at the same
12
time influence the growth path of participants' economies. For example, an allocation
13
based on population at a given date would provide an incentive for population control⁹¹.
14
15
Imposing carbon taxes can have large distributive consequences. While a system of
16
grants can largely offset these distributive consequences, such offsetting grants might well
17
not be made. Providing tradeable permits equal to existing levels of emissions seemingly
18
makes no firm or household a loser. But granting permits in that way represents effectively
19
a grant of money (such permits have monetary value) in a way which may not well accord
20
with standard ethical principles. For instance, by embarking on an ambitious program to
21
reduce emissions, a firm may qualify for fewer permits than it would otherwise. Not only
22
does this violate ordinary notions of fairness, but anticipation of granting permits in this
23
way would, accordingly, have strong adverse effects on emissions⁹².
24
25
That a tax has large distributive consequences, while presenting a political impediment to
26
its introduction, is not necessarily an argument against it. Some argue that those who failed
27
to pay the full social costs of their actions earlier are not therefore entitled to special
28
allotments now⁹³.
29
30
Once one recognizes that the distribution of permits across countries will, inevitably, be
31
decided by some principle other than current levels of emissions, then it becomes clear that
32
both taxes and tradeable permits will have distributive consequences. An agreement among
33
countries to impose uniform corrective taxes, with each country retaining its own revenue,
34
would have little redistributive consequences across countries and the burden of the tax
35
would, as noted earlier, likely be progressive.
36
37
Governments in the North might decide to use some of the revenues so generated to
38
encourage activities (such as R & D directed at technology appropriate for developing
39
countries) that benefit the South, or to provide other forms of assistance. Decisions about
40
this could be made bilaterally, or collectively. By contrast, decisions about how tradeable
41
permits would be allocated across countries would have to be made multilaterally. Arriving
42
at a formula for distributing these property rights may be far more difficult than arriving at
43
a tax rate and a procedure for its revision, as any such formula may entail substantial
44
redistribution.
34
1
1.5.5.1
A Double Dividend?
2
3
Revenues from carbon taxes may allow a reduction in distortionary taxes elsewhere in
4
the economy. If the (compensated) elasticity of demand of labor is relatively high, and the
5
revenues from the carbon tax are used to reduce taxes on labor income, then there would be
6
a double dividend from the carbon tax in the reduced dead-weight loss from the labor tax,
7
which would otherwise be significant.
8
9
At least three objections have been raised to this idea. First, conceptually: rationalizing
10
the tax system by reducing the most distortionary taxes is certainly a worthy goal, but
11
should not be confused with imposing a carbon tax; neither implies the other. Second,
12
empirically: if the (compensated) labor supply elasticity is relatively low, then the dead-
13
weight loss form the labor tax is low, and the commensurate welfare gain is reduced.
14
Third, politically: if carbon tax revenues are used to offset the existing deficit, rather than
15
to reduce taxes on labor and capital, then the carbon tax acts more like an ordinary tax
16
increase, increasing distortions from taxation to pay for budget items with a lower return
17
than the extra burden imposed. The gains to total welfare (reductions in dead-weight loss)
18
depend on the welfare losses associated with these other distortionary taxes, as well as the
19
cross-elasticities of demand between carbon and other taxed commodities⁹⁴.
20
21
Even though carbon taxes may have a positive effect on economic welfare, they can at
22
the same time have a negative effect on measured economic growth, since those measures
23
typically do not include the value of environmental degradation. Researchers differ on the
24
size of the loss. The wide spread in the numerical results, however, should not obscure
25
agreement among researchers on a number of important points. All models used in the
26
major comparison studies to date have projected: first, that intervention would be required
27
to achieve the emissions targets; second, that the size of the required tax increases with the
28
stringency of the carbon limit; and third, that the size of the appropriate carbon tax varies
29
over time, even for the same emissions or concentration target9⁵.
30
31
1.5.5.2
Energy Taxes
32
33
Energy taxes as a means of controlling greenhouse emissions must be viewed as "second
34
best" taxes in that they do not directly tax the externality, greenhouse emissions. While
35
carbon taxes directly penalize the externality-generating activity, less targeted alternatives,
36
such as energy taxes, may be politically more acceptable. Carbon taxes reduce emissions
37
first, directly, by moving up the demand curve; second, indirectly, by encouraging
38
consumers to switch to less carbon-intensive energy sources. On the other hand, energy
39
taxes work through the first path, by reducing total energy consumption. But to the extent
40
that certain kinds of energy, like hydroelectric, have, at least in the short run, a relatively
41
inelastic supply, a major impact will be on oil, gas, and coal; and to the extent that oil and
42
gas supplies are best described by a model of an exhaustible natural resource, with
43
relatively low extraction costs, most of the supply reduction will occur in coal. Thus,
44
indirectly, there will be a considerable amount of switching, through the indirect effects.
35
1
1.5.5.3
Tradeable Permit Markets.
2
3
In order for systems of emission permits to achieve reductions in emissions efficiently,
4
there needs to be a market for emissions across international boundaries. There is some
5
debate about the role of government or international organizations in establishing a market
6
for such emissions. While some believe that there are private incentives for the
7
establishment of markets, others contend that government can play a key market facilitation
8
role through establishing centralized clearinghouses for information or even provide for
9
permit banking (storage) or brokerage (trading) to facilitate trades between private parties.
10
These services would prove especially useful in the more complex international context.
11
12
1.5.5.4
Combining taxes with tradeable permits
13
14
While carbon taxes and tradeable permits are typically presented as alternatives, policy
15
makers may prefer to combine them. The major disadvantage is the additional
16
administrative cost. The advantage is more subtle: the market value of tradeable permits is
17
reduced as taxes increase. With an optimal carbon tax, and with a tradeable permit supply
18
set equal to the optimally chosen level of emissions, the price of a permit should be zero.
19
More generally, the greater the tax, the less the value of a permit (for a fixed supply of
20
permits); and thus, the less the distributive consequences of alternative rules for allocating
21
the initial endowments of permits. Another possible combination is permits for large
22
sources and a tax (set to equal the permit price) on small sources.
23
24
1.5.5.5
Intertemporal patterns of taxation
25
26
If the target is the long-run atmospheric concentrations of greenhouse gases, then climate
27
change damages will be approximately the same for emissions in any particular year,
28
although the optimal carbon tax must be adjusted for differences in costs, discounting, and
29
risk⁹⁶. The focus on concentrations also implies that early reductions are more valuable
30
than later reductions.
31
32
For exhaustible natural resources, such as oil, economic efficiency requires that those
33
deposits with the lowest cost of extraction be extracted first. Hotelling (1931) argued that
34
competitive equilibrium implies that rents (price minus costs of extraction) must rise at the
35
rate of interest. The price of the backstop technology (an energy source assumed to be
36
available in unlimited quantities at a certain price after a certain date, such as electricity
37
from solar photovoltaic cells) determines the set of resources to be ultimately exploited, i.e.
38
all resources for which the cost of extraction is less than the price implied by the backstop
39
technology's price. Thus, it is the tax on oil or gas at the date of switching to the backstop
40
technology which determines the ultimate amount of oil and gas that will be extracted, and
41
thus the total burden of CO2 placed on the atmosphere by oil and gas. If that were the
42
only matter of concern, one could imply announce a commitment to impose such a tax
43
sometime in the future, when relevant backstop technologies become available and
44
competitive. That announcement would, if believed, have an immediate effect on current
36
I
prices.
2
3
Figure 3 shows the dynamics of exhaustible resources. S denotes the total available
4
stock of gas and oil. S. represents the amount of stock left in the ground, a function of the
5
"backstop" price. Imposing a uniform emission tax lowers the cut-off price, and flattens the
6
price curve. The effect on current price depends on how close the current stock is to
7
economic exhaustion (S₀). Given the currently high levels of rent, it is probable that the net
8
current effect on producers is to raise the producer price, not lower it. Thus, short-run
9
leakage effects will be the opposite of that predicted in many models.
10
11
1.5.6 Regulatory Approaches
12
13
Regulation of greenhouse emissions may take many forms, including fuel restrictions,
14
technology standards, and various economic incentives; Chapter 11 discusses these options
15
in detail. Economists have long argued for the use of economic incentives for
16
environmental management, although governments have so far relied on traditional
17
regulations almost exclusively, as traditional approaches have been more acceptable to the
18
public and industry. For example, Corporate Average Fuel Economy (CAFE) standards in
19
the U.S.A. require automobiles to meet certain mileage standards.
20
21
Proponents of the traditional approaches often claim that these approaches "force"
22
technology, with less redistribution than forcing technology through taxes. Thus, if
23
automobile makers are required to attain a certain mileage standard, they will meet the
24
standard; on the other hand, gasoline taxes might have to rise significantly to reduce fuel
25
consumption by the same amount. Evidence for the claim of technology forcing, however,
26
is equivocal. In several cases in which industry failed to meet the applicable standards,
27
regulators withdrew the standard in the face of unacceptably high economic costs. The
28
apparent advantage of technology forcing -- one large club instead of the subtle and
29
continuous incentives provided by market forces -- is often in fact a disadvantage.
30
31
Other disadvantages to the traditional approach are: First, traditional regulations do not
32
in general result in economic efficiency, since those in one sector face implicit or explicit
33
incentives at the margin that differ from those in other sectors. Second, traditional
34
regulations fail to account for offsetting private responses that may neutralize the
35
regulation's intended effects and even cause environmental harm. Third, traditional
36
regulations provide no incentives for exceeding the given target, even when doing so might
37
result in little additional cost⁹⁷.
38
39
Traditional regulations that focus on inputs and technology rather than outputs have the
40
further disadvantage of not directing research toward meeting performance objectives at
41
least cost. For instance, when stack gas scrubbers are required, research will be directed at
42
producing scrubbers at least cost, rather than reducing emissions at least cost. Hence, a
43
dynamic inefficiency is added to the obvious static inefficiencies. Finally, because of the
44
nature of the regulatory process, traditional regulatory designs are more likely to be
37
1
captured by special interest groups⁹⁸.
2
3
4
1.6
SUSTAINABLE DEVELOPMENT
5
6
The concept of sustainable development was formulated about 1980 as a response to the
7
apparent conflict between environmental concerns and the need for economic growth,
8
especially in developing countries. At the time, preserving biodiversity and maintenance of
9
environmental quality seemed incompatible with a 5- or 10-fold increase in world output, as
10
would be necessary if per capita incomes of the South were eventually to approach those of
11
the North now. The sustainable development debate rekindled interest in the question of
12
resource scarcity, originally addressed in the economics literature by Malthus (1798), and
13.
revived in the policy arena with the publication of Limits to Growth (1972). A variety of
14
definitions of sustainable development have been proposed. The Bruntland Commission
15
offered this interpretation:
16
17
Sustainable development is development that meets the needs of the present without
18
compromising the ability of future generations to meet their own needs99.
19
20
Although the Commission clearly had in mind environmental considerations, its report did
21
not spell out exactly what sustainable development included. The Interamerican
22
Development Bank explicitly included environmental concerns in its formulation:
23
24
[Sustainable] development distributes the benefits of economic progress more equitably,
25
protects both local and global environments for future generations, and truly improves
26
the quality of life.
27
28
1.6.1 The Economic Concept of Sustainable Development
29
30
Although sustainable development began as an ethical principle, it is at the same time
31
an economic concept, focusing on two issues: 1) intertemporal equity and 2) capital
32
accumulation and substitutability.
33
34
Intertemporal equity. Robert Solow's definition (Solow 1992), which focuses on
35
intertemporal equity, has enjoyed wide currency among economists: sustainable
36
development requires that future generations be able to be at least as well off as current
37
generations. The central implication is that any environmental degradation should be offset
38
by increases in capital stock sufficient to ensure future generations at least the same
39
standard of living. Sustainable development does not preclude the use of exhaustible
40
natural resources, but requires that any use be appropriately offset.
41
42
In practice, sustainability as defined by Solow provides few constraints on growth paths
43
for the developed countries, so long as steady increases in productivity continue. Technical
44
change alone, without further capital accumulation, may well sustain future living standards
38
1
and offset any effects of environmental degradation. To see this with a numerical example,
2
note that even if estimates of adaptation costs are taken to be 1 to 3 percent of GDP should
3
significant warming occur, and if even moderate rates of technical progress of 1 to 1.5%
4
per annum continue to occur, then future generations 45 to 70 years from now will have
5
twice the income of the current generation. Even with no discounting, it would be hard on
6
this account alone to justify the sacrifice of further consumption by this generation in order
7
to enhance the standard of living of the future generation.
8
9
Capital accumulation and substitutability. To what extent can technology, skills, and
10
capital equipment substitute for a decline in exhaustible resource stocks or a decline in per
11
capita environmental amenities? Solow's definition, and much of economic theory to date,
12
implicitly assumes that substitutes exist or could be found for all resources. As formulated
13
by Pearce (1988), if substitution possibilities are high, as most evidence from economic
14
history indicates, then no single resource is indispensable, and intertemporal equity stands
15
as the only crucial issue. If on the other hand, human and natural capital are complements
16
or only partial substitutes (e.g., if because of the irreversibility of extinction¹⁰⁰, capital
17
accumulation is only a partial substitute for biodiversity) then different classes of assets
18
must be treated differently, and some assets are to be preserved at all costs.
19
20
Pearce (1992) distinguished between strong and weak sustainability. Weak
21
sustainability requires that any depletion of natural capital be offset by increases in human-
22
produced capital -- the Solow criterion -- or by the substitution of other forms of natural
23
capital, such as renewable assets in place of nonrenewable assets. Strong sustainability
24
requires that some natural capital, being irreplaceable, must be preserved It has been
25
argued that there are no close substitutes for the atmosphere and the climate it produces,
26
implying no substitution possibilities and hence the need to preserve the atmosphere.
27
28
1.6.2 Implications of Sustainable Development for Developing Countries
29
30
In many developing countries, Solow's definition would not be viewed as
31
acceptable, since it seems to place no weight on their aspirations for growth and
32
development. Developing countries have also implicitly criticized the debate over
33
substitutability for the same reason: if some natural assets must be preserved at any cost,
34
then there may be no trade off with development. A leading spokesman for the G-77 has
35
emphasized the importance of economic growth in achieving sustainable development:
36
37
None of these linked [development] issues can be resolved unless and until there is
38
broad-based development in the South. Only such broad-based development can
39
provide the foundation of international security. The Northern approach is to attack the
40
symptoms, with a residual emphasis on poverty eradication. But the international
41
community must insist on addressing the underlying causes for concern. Development,
42
environmental protection, peace, and security are indivisible¹⁰².
43
44
Similarly, the G-77 and China emphasized the need for economic growth in the
39
1
following statement on sustainable development and the environment introduced during the
2
INC negotiations in 1991:
3
4
Protection of the global climate against human-induced change should proceed in an
5
integrated manner with economic development in light of the specific conditions of each
6
country, without prejudice to the socio-economic development of developing countries.
7
Measures to guard against climate change should be integrated into national
8
development programmes taking into account that environmental standards valid for
9
developed countries may have inappropriate and unwarranted social and economic costs
10
in developing countries 103
11
12
13
1.7
CONCLUSIONS
14
15
Climate change presents the analyst with a set of formidable complications: large
16
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
17
long time lags between emissions and effects, an irreducibly global problem, wide regional
18
variation, and multiple greenhouse gases of concern. The risks of climate change are
19
highly asymmetrical, with a large probability of a small loss, and a small probability of a
20
large loss. Even in the presence of significant uncertainty, this asymmetry, plus the
21
principles of risk aversion and portfolio balancing provide the rationale for going beyond
22
no-regrets policies to those that incur net costs.
23
24
The atmosphere is an international public good, in that all countries benefit from each
25
country's reduction in greenhouse emissions; greenhouse gases are an international
26
externality, in that emissions by one country affect all other countries to the same extent.
27
28
Both public goods and externalities require a legal framework within which the
29
problems they pose can be addressed. No such legal framework now exists for global
30
climate change. Mechanisms for control of international public goods may include the
31
definition of property rights, the definition of limits to emissions and a consensus for
32
distributing the same in a fair and equitable manner. If, on the other hand, each agent acts
33
in its individual interest, the result will be too little of the public good and too much of the
34
externality.
35
36
Climate change demands a decision process that is sequential, can respond to new
37
information with mid-course corrections, and can include insurance, hedging, and the option
38
value of alternative courses of action. The challenge today is to identify short-term
39
strategies in the face of long-term uncertainty. The question is not, what is the best course
40
over the next 100 years, but rather, what is the best course for the next few years, knowing
41
that a prudent hedging strategy sill allow time to learn and change course.
42
43
Policy measures to reduce risks to future generations include 1) immediate reductions
44
in emissions; 2) R&D on new supply and conservation technologies; 3) continued research
40
1
on how much change is likely and what its effects will be; and 4) investments to assist in
2
adaptation if significant climate change occurs. A well-chosen portfolio of policies will
3
yield greater benefits for a given cost than any one option undertaken by itself. Striking the
4
appropriate balance requires taking into account costs, benefits, and risks.
5
6
In an interrelated global economic system, an attempt to reduce greenhouse gas
7
emissions in one region or one sector of the economy may be offset by increases in other
8
regions or sectors. This may occur through a) the loss of comparative advantage in the
9
carbon-intensive sectors of the regions that limit emissions; b) the relocation of industries;
10
or c) changes in world energy prices and the resulting shift in consumption. Any control
11
strategy must account for these global effects.
12
13
Issues of efficiency and equity can largely be separated. The Framework Convention on
14
Climate Change calls on all parties to implement cost-effective measures for abatement,
15
enhancement of sinks, and adaptation. The Framework Convention also explicitly requires
16
an equitable sharing of the burdens of response, recognizing the common but differentiated
17
responsibilities of the parties. Different countries will be affected differently by climate
18
change and by policy responses to it. The South is more likely to be adversely affected
19
than the North; moreover, developing countries often lack the financial and technical
20
resources to respond. The Framework Convention does not, however, include a formula for
21
sharing the costs of addressing climate change.
22
23
Efficiency requires that emission reductions occur where their cost is lowest,
24
irrespective of who bears the financial responsibility. Efficiency calls for removing energy
25
subsidies, reforming and clarifying property rights that affect energy use and carbon
26
storage, and reducing non-greenhouse externalities that have the side benefit of reducing
27
greenhouse emissions. Efficiency may also be promoted, and greenhouse emissions
28
reduced, by better information dissemination and by addressing capital market imperfections
29
that inhibit the adoption of energy-efficient technology. Dynamic analysis indicates large
30
potential gains from flexibility in timing of greenhouse reductions to allow for the
31
economical turnover of capital stock, and to allow time for the development of low-cost
32
substitutes. Policies that promote efficiency by requiring nations to face the full costs of
33
their actions will also address equity concerns.
34
35
Efficiency also calls for international mechanisms such as joint implementation and
36
coordinated economic instruments. Coordinated carbon taxes and tradable carbon emission
37
permits can correct the market failure associated with greenhouse emissions.
41
1
2
1.8
ENDNOTES
3
4
1.A warmer climate would directly affect the temperature-sensitive sectors of the economy: agriculture, forests,
6
and fisheries; and construction. Because of the risk of drought, arid and semi-arid regions are likely to be most
7
vulnerable to warming. A warmer climate may also encourage insect populations; this, in turn, is likely to
8
decrease agricultural yields in some regions. On the other hand, increased CO2 concentrations increase
9
photosynthesis and the effect of water consumption in controlled settings, and may also do so in farmers' fields,
10
increasing crop yields.
11
12
A warmer climate would hurt agriculture in some regions and help in others. Recent global studies find that
13
Northern temperate regions could benefit, particularly for small increases in temperature, but low latitude areas
14
could lose. The global studies show significant losses from climate change alone that are significantly offset by
15
the direct beneficial effects of higher ambient levels of carbon dioxide on plant growth. Reilly, et al. (1993)
16
calculate that global agricultural production will increase for global mean warming of up to 2°C.
17
Most analysts expect that agriculture can adapt without significant added costs for the world as a whole (although
18
some regions will suffer greater-than-average losses), in part because of crop shifting and other adaptation by
19
farmers. Trade in international commodity markets (Ausubel, 1990; Rosenzweig and Parry, 1994; Reilly,
20
Hohmann, and Kane, forthcoming; Kane, Reilly, and Tobey, 1992) can ensure that agricultural prices change only
21
moderately, though particular regions may gain or lose significantly.
22
2.A warmer climate implies both a poleward movement of forests and changes in forest composition. The result
23
would be to increase boreal forests and, by a smaller percentage, to decrease tropical forests. On balance, doubling
24
CO2 concentrations would modestly reduce both standing biomass and forest area (Sedjo and Solomon, 1989).
25
Some analysts also calculate an additional temporary decline in forest cover during the transition to a warmer
26
climate (EPA, 1989); this decline could last for several hundred years, as more forest death occurs on southern
27
edges than additional growth occurs on northern edges. Higher CO2 concentrations might also increase tree
28
growth, partially offsetting direct losses from warming.
29
.Some analysts conclude that a warmer climate would significantly reduce water supply (Gleick, 1987). Most
30
climate models now predict hotter and drier weather for the mid continents (IPCC, 1990a). If this occurs, the
31
downward pressures on water supply will be intensified by upward pressures on water demand.
32
33
Because of the key role of local conditions, further generalizations are difficult. River basin runoff is the
34
difference between precipitation and evaporation plus absorption (IPCC, 1990a). Hence small variations in these
35
three terms--each of which depends on the accuracy of regional climate forecasts--can make a large difference
36
in the runoff.
37
38
4. A warmer climate implies reduced heating costs and greater cooling costs. At least for countries in
39
temperate and cool climates, this would move the population-weighted average temperature closer to the ideal
40
indoor temperature. Nordhaus (1991) finds that warming of approximately 3.0°C in a 1981-sized U.S.
41
economy will increase electricity costs by $1.65 billion and decrease heating costs by $1.16 billion (1981
42
dollars), resulting in net costs of $0.49 billion. Rosenthal, Gruenspecht, and Moran (1994) find that global
43
warming lowers the cost of heating and cooling in the U.S.. Cline (1992) finds net costs of $9.9 billion.
44
45
5.Global mean sea level appears to have risen 10-20 cm over the past century (Warrick, Oerlemans et al., 1990).
46
The IPCC has projected a warming-induced rise in mean sea level of 21 to 71 cm by 2070 (IPCC, 1990a), sharply
47
reduced from estimates of 3 to 8 m. made in the early to mid-1980s (Schneider and Chen, 1980; Hoffman et al.,
48
1983; Hoffman et al., 1986), which assumed a disintegration of the West Antarctic Ice Sheet. The current "best
42
I
estimate" for mean global sea level rise by 2030 under "business as usual" is 18 cm (Warrick, Oerlmans et al.
2
1990; Raper et al., 1990).
3
4
A rising sea level could inundate some small island nations, flood some low-lying coastal cities (e.g. in
5
Bangladesh), damage coastal farmland, and contaminate water supplies (IPCC, 1990a). More severe storms
6
would accelerate coastal erosion and affect aquifers, intensifying problems existing today. Warmer oceans might
7
damage coral reefs, a natural defence for some coastal areas. To account for the costs of protective measures,
8
the costs of defending the world's developed coastlines against a 1 m rise over the next 100 years have been
9
estimated to range from 0.01% for the former USSR to 0.74% for Indian Ocean small islands, with a world
10
average of 0.028% of GNP per annum (IPCC, 1992).
11
6. Some analysts believe that the justification for costly and more restrictive actions rests on intangible costs
12
(Nordhaus, 1993). Intangible refers to the difficultly to measuring, and include migration, comfort, health,
13
leisure activities, urban infrastructure, and air pollution (Fankhauser 1992; Cline 1992). A warmer climate
14
would improve human comfort in cold areas, and in the winter generally, while decreasing comfort in warm
15
areas. Mearns et al. (1984) calculate a threefold increase in heat waves for a 1.7 degree C. rise in U.S. mean
16
temperature. It is not yet clear whether net comfort will rise or fall for a given rise in temperature. Chapter
17
6 covers these issues in more detail.
18
7.Both the natural rate and human contribution to the process of species loss are difficult to estimate (EPA, 1989).
19
Predicting the effect of climate change on species distribution is more difficult still. The magnitude and even the
20
sign of these intangibles remains in dispute, for uncertainty about the duration and type of environmental changes
21
that would be caused by climate changes make the long-term projection of species change highly complex.
22
Moreover, while some species may have difficulty adapting to climatic changes, opportunities for other species
23
that otherwise may have become extinct may open up. Population pressures have added to pressure on
24
ecosystems, particularly in the third world. Climate change may exacerbate these damages, particularly in Africa,
25
where environmental degradation has been particularly pronounced during the last 15 years (UN, 1989).
26
8. A warmer climate might also increase climate variability, though climatologists cannot say with assurance
27
whether climate will become more or less variable daily and seasonally. The normal variability of existing
28
climate also makes it difficult to detect any warming that might be occurring. The "signal to noise" problem
29
makes it possible for observers to mistakenly consider a normal extreme event as evidence of a trend, or to
30
fail to see a trend in a noisy data series. Because of the signal-to-noise problem, the scientific community is
31
unable to indicate confidently whether extremely warm years of the 1980's are evidence of climate change or
32
not (IPCC 1992, A. Solow 1992).
33
9.Recent ice core data from Greenland points to earlier temperature rises of several degrees occurring within a
34
few decades (IPCC 1994); reasons for these sudden changes are still not understood, but might have come from
35
changes in deep ocean currents. The other side of this debate holds that on the whole the biosphere is
36
homeostatic or self-correcting; this "Gaia hypothesis" compares the biosphere to a living being: once moved away
37
from equilibrium, self-correcting forces naturally move it back to equilibrium (Lovelock, 1979). The two
38
hypotheses are not necessarily inconsistent. Within a range of variation, homeostatic properties could dominate,
39
even if stability was not guaranteed outside that range.
40
Countries make century-long choices implicitly, as when they choose population policies, policies affecting
41
long-term capital formation and productivity growth, or protection of environmental assets.
42
11. Although bargaining theory has contributed some basic principles, such as the importance of threat points.
43
1
12. Scientists have proposed several methods to measure the warming potential of different greenhouse gases
2
(IPCC Special Report 1994). Although global warming potential (GWP), endorsed by the IPCC, is a useful
3
concept in formulating comprehensive approaches to greenhouse mitigation policies (Stewart and Wiener,
4
1990), some analysts have recently criticized the concept on the grounds that GWP implicitly sets the
5
opportunity costs of an increment in radiative forcing equal for all periods in the future (Schmalensee, 1993).
6
If all greenhouse gases had the same rate of decay, then this problem would not arise.
7
13. Although forests cannot be expanded indefinitely, and thus increased carbon sequestration is not a permanent
8
solution to increasing greenhouse gas emissions.
9
10
4.Advances is cost-benefit analysis have allowed the introduction of risk and equity issues in a systematic way.
11
15. Frank Knight's often-quoted distinction separates risk, for which the probabilities of different outcomes are
12
known, from uncertainty, in which either they are unknown, or some potential outcomes are not specified. Ref.
13
Knight, Frank, 1922, Risk, Uncertaintv, and Profit.
14
16. This division of responses is from Nordhaus (1994).
15
17. While most insurance markets do not reduce overall risk, there are a few exceptions: fire insurance companies,
16
by requiring commercial buildings to install sprinklers, do reduce the overall level of losses.
17
18. This assumes decreasing absolute risk aversion (about which there is a general consensus). Even if northern
18
countries were adversely affected, so long as the adverse effects were proportionately smaller, then, assuming
19
decreasing relative risk aversion (about which there is less consensus), the north would be in a position to insure
20
the south.
21
22
Insurance contracts may also be created because of differences in judgments concerning the probability of the
23
insured event occurring. This suggests the potential of an important principle to be invoked in future international
24
negotiations/agreements: countries that believe that the risks of climate change are low, and are therefore
25
seemingly unwilling to take strong actions to mitigate these risks, ought to be willing to provide insurance against
26
climate change at low cost (since it has, from their perspective, an actuarially low value). (See below for a
27
discussion of problems of enforce.) (Heal and Chichilnisky **date, reference**; earlier, Stiglitz ** date ** on
28
social risk bearing.)
29
30
Insurance markets, if appropriately designed, would have one further advantage: some of the losses associated
31
with climate change are easily avoidable, e.g. construction of durable ocean-front houses. Insurance firms work
32
hard to mitigate the losses that are incurred. Either the insurance should be based on exogenous events (e.g. not
33
on the dollar losses incurred, but on the rise in the sea water), or the insurance companies should be given broad
34
discretion to require the insured to undertake actions to mitigate losses.
35
Chapter 6 provides estimates of expected regional damage.
36
20.In the absence of such recuirements, moral hazard problems arise. It may be desirable to focus government
37
intervention not on the primary insurance market but on the reinsurance market. See recent U.S. government
38
analysis of the market failures and the design of appropriate responses to those market failures in insurance
39
markets for natural disasters.
40
21. Computer modelers participating in the Energy Modeling Forum examined an "accelerated R&D"
41
scenario in which the cost of nonelectric backstop falls from $100 to 50$ per barrel of oil equivalent, and the
42
cost of the electric backstop falls from 75 to 50 mills per Kwh. The four models used were remarkably
43
consistent in their estimates of economy-wide costs, reporting GDP losses falling by 65 per cent for the 20
44
1
percent emissions reduction scenario (EMF 1993).
2
22. For example, low-lying states may suffer permanent damage.
3
23. Weitzman et al. (1981), cited in Lind (1993), make these points in formulating a sequential decision strategy
4
for developing synthetic fuels.
5
24.Manne and Richels, Nordhaus, and Peck and Teisberg all report a high value of information on *** and ***
6
in their computer simulations. In addition, *** also reports a high value to information on ***
7
5.Richels and Edmonds (1993) provide a demonstration of this proposition; they calculate relatively low costs
8
for stabilizing CO2 concentrations if flexibility in timing is allowed, compared to capping and stabilizing
9
emissions to achieve the same atmospheric concentration.
10
26. Alternatively, the Kaya identity may be written
11
12
Growth rate
=
growth rate
-
decline in energy
-
emissions
13
of CO2 emissions
of output
per unit output
per unit energy use
14
15
That is, CO2 emissions will not rise as long as output grows no faster than the combined decline in energy
16
intensity per unit of production and CO2 emissions per unit of energy use. This formulation applies most usefully
17
to the developed countries.
18
27. Chapter 8, "Evaluating the Costs of Mitigation," treats the important issue of inertia and technology.
19
28.Note that energy efficient development paths for developing countries have been proposed (Goldemberg and
20
Reddy 1988).
21
22
29. That is, for conventional exhaustible resources, there is a stock, S. Welfare depends on flows out of the stock
23
each year:
24
U(S, - So, S₂- S₁+₁ - S₁,......), where S, is the stock at the end of period t.
25
26
Here, welfare depends directly only on the stocks, though indirectly, through the effects on consumption, on
27
emissions, which affect the change in stock:
28
29
U(S,,C,(S, - S₀),S₁,C₂(S₂- S₁),........S₁, C,(S,+₁ - S₁),......).
30
30. Even when the flow exceeds the long-run sustainable level, it will not be optimal to reduce the flow
31
instantaneously, unless or even when - there are zero costs of adjustment.
32
33
For the atmosphere, a sustainable stock of greenhouse gases means stable concentrations. Current emissions are
34
estimated to be about twice emissions consistent with stable concentrations.
35
31. Postponing action may lead to some irreversible damages, for example the flooding of low-lying states.
36
32. This is not quite correct, since price affects incentives for exploration, and some marginal wells would not
37
be drilled if oil prices fall too low.
38
33. For analyses of market and optimal responses to uncertainty about the arrival of backstop technologies, see
39
Dasgupta, Gilbert, and Stiglitz [ ] and Dasgupta and Stiglitz [ ]1.
45
1
34.Although for counter examples to the received wisdom see Coase (1990).
2
35.Formally, if A measures the quality of the atmosphere, then each individual's or country's welfare, U, is a
3
function of its own consumption, C, and the shared public good, A: U' (C', A). This does not mean that all
4
individual's (country's) value changes in A the same; that is may differ in magnitude, and
5
even in sign.
6
36.Although there may also be trade-offs: reducing those gases that contribute most to local pollution may
7
sometimes be at the expense of increased emissions of greenhouse gases.
8
37.This kind of optimization problem was first studied by Edgeworth, Mathematical Psychics ( ). The
9
importance of incentive effects for the analysis of distributional issues was first emphasized by Mirrlees (1971)
10
and Fair (1970). There are, obviously, important incentive effects: if the LDCs were able to classify any
11
expenditure that had some effect on mitigation as a mitigation expenditure, with the cost borne by the developed
12
countries, they would have an incentive to undertake excess expenditures of this type. The GEF (Global
13
Environmental Facility) directly addresses this issue, by only providing funds for incremental costs, that is those
14
costs that go beyond what would have been the efficient level of expenditures ignoring the public good benefits
15
of greenhouse gas mitigation.
16
38. The importance of the assumptions of the absence of transactions costs, including the presence of perfect
17
information, has only gradually come to be recognized. See Farrell [ ], or Stiglitz [1988] for an elementary
18
textbook treatment.
19
39. That is, for instance, it makes no difference whether smokers or non-smokers are given the property rights to
20
air. Whether smokers value smoking more or less than non-smokers value clean air will determine whether
21
smoking occurs. How property rights are assigned does make an important difference for the distribution of
22
welfare. Coase's result that outcomes are unrelated to the initial assignment of property rights obviously ignore
23
potentially important income effects.
24
25
A slight extension of this perspective says that social scientists should simply describe the outcome of the
26
bargaining process by which property rights are assigned. Beginning with the important work of Nash [ ], a
27
variety of bargaining theories have been developed, most of which emphasize the importance of "threat
28
points"--the outcomes which arise in the absence of a bargaining agreement-- to the determination of the eventual
29
outcome. In this case, the fact that the net losses of many developed countries may be limited relative to those
30
of many of the less developed countries suggests a bargaining solution in which much more of the costs of
31
mitigation are borne by the less developed countries than under the "social welfare function" allocations described
32
earlier.
33
40.For small taxes, these are "compensated" taxes, and have no welfare effect, though they have a substitution
34
effect, and therefore do reduce pollution.
35
41. The loss in welfare (ignoring the benefits from reduced greenhouse gas warming) are the Harberger triangles,
36
and can thus be shown to be proportional to the product of the elasticity of demand for energy and the share of
37
energy in national output. Since poorer countries are likely to have less access to alternatives which increase the
38
elasticity of demand, and since the share of energy is larger in richer countries, the burden of the tax is
39
progressive.
40
42. The polluter pays principle endorsed by the OECD is exclusively prospective.
46
1
43.Using either an egalitarian social welfare function approach, or a Rawlsian analysis "behind the veil of
2
ignorance" (Rawls 1971) leads to the rejection of the "polluter pays principle." Since at the time the relevant
3
actions are taken, the polluter is not cognizant of the effects, such fees have no incentive effects, but rather appear
4
as random taxes, lowering each person's expected utility, and in particular the expected utility of the worst-off
5
individual.
6
7
There is a further ethical issue: ascertaining who the true beneficiary of escaping paying for the pollution is
8
generally difficult. It need not be the individual, firm, or country actually engaging in the externality generating
9
activity; in competitive markets, when firms are not charged the full social costs of production, product prices
10
will fall, giving consumers a substantial fraction of the benefits.
11
44. Manne and Richels (1992) show that, under the IPCC emissions scenarios, even the most drastic controls on
12
emissions from developed countries would be insufficient to stabilize greenhouse concentrations without some
13
means of controlling emissions from developing countries.
14
45.Public goods exist when property rights are not or cannot be clearly assigned. The atmosphere is an
15
international public good because assigning property rights to the atmosphere is difficult for one nation acting
16
alone, and particularly difficult when many sovereign states must agree among themselves. Tradable greenhouse
17
emission permits, discussed below, attempt to resolve the problem by explicitly assigning property rights to
18
greenhouse emissions.
19
20
46.See Abreu [ ].
21
47. In some cases, equity considerations may prevent Coase (1960), in his discussion of externalities, emphasized
22
the separability between efficiency and equity issues. Though there have been several important qualifications
23
to Coase's conjecture, emphasizing the importance of public goods, imperfect information, and transaction costs
24
(see Farrell , Stiglitz [1988]), still, the basic insight remains applicable here.
25
48. Chapter 7 discusses this issue at greater length.
26
49. These concerns are not just theoretical possibilities, as the following two examples illustrate. Assume that
27
the North imposes high energy taxes, but the South fails to do so. Energy intensive industries, such as aluminum,
28
migrate from North to South. But energy efficiency in the south is much less than in the north, so that the total
29
energy used to produce a ton of aluminum could increase substantially. While economic efficiency would call
30
for locating energy intensive industries where energy efficiency is greatest, a system of partial controls would
31
results in energy intensive industries being located where energy efficiency is lowest. Similarly, the reduced
32
energy consumption by the North will result in lower producer prices of oil and gas, leading to increased
33
consumption of energy in the South, partially offsetting any energy conservation induced in the North.
34
50. The precise manner in which this should be done is a technical matter, treated in the literature on Global
35
Warming Potential (see, e.g., IPCC 1990). To the extent that there are large differences in atmospheric lifetimes,
36
then the relative weighting of different greenhouse gases should change over time, since the "shadow price"
37
associated with effects on relative concentrations at different dates will differ.
38
51. Similarly, if the developed countries restrict forest cutting, the price of lumber may rise, inducing the
39
developing countries to cut down more of their own trees. While total carbon sequestration may not increase,
40
environmental and economic efficiency will decrease if, as some researchers have concluded, hardwood forests
41
in the less developed countries may be the least desirable ones to cut down from an ecological or economic
42
perspective (Edmonds and Reilly, 1983).
47
1
52. Only a few models take into account international capital flows. Thus, most models do not address issues
2
of industry relocation (McKibben and Wilcoxen, 1992). Chapter 11 provides a more complete discussion on
3
leakages.
4
53. In the case of production of highly substitutable commodities, carbon leakage will, of course, be much greater.
5
6
54. Whether taxes in fact have this effect depends in part on the shape of the demand curves. With intertemporal
7
separability in demand curves and constant elasticity, with no backstop technology, a constant ad valorem tax has
8
no effect on the pattern of consumption.
9
10
Coal presents markedly different issues, not so much because of its greater emissions per unit energy, but because
11
of it higher cost of extraction to price ratio. Lowering producer prices may result in less coal being consumed,
12
provided alternative energy sources become available. Thus, taxes on coal are likely to have significant general
13
equilibrium as well as partial equilibrium effects; the increase in the price of coal will lead to a substitution of
14
gas and oil. If alternative energy sources are not available, such policies will only affect the intertemporal timing
15
of coal consumption (given the much more limited resources of gas and oil). But even that might be of some
16
value in reducing long run greenhouse gas emissions, as the ability to extract energy from coal may increase
17
significantly over time. Analyzing the optimal intertemporal structure of taxes, to minimize long run ambient
18
levels of greenhouse gases, taking into account both intertemporal substitution and substitution across energy
19
sources, is a complicated technical issue, that to date has not been adequately analyzed.
20
55. United Nations Conference on Environment and Development, Framework Convention on Climate Change,
21
May 9, 1992.
22
56. The most difficult problem is posed by those investors who invested in these resources, on the basis of one
23
regime (where these resources were not taxed). Do they have any special claim to compensation for a "change
24
in regime." Changes in demands and supplies occur for virtually all resources, and are an inevitable part of the
25
risks in investing. While most economists would argue that arbitrary and capricious changes in policies contribute
26
to business uncertainty, and therefore have an adverse effect on economic growth, reasoned changes in policies
27
in response to changes in information are an inevitable part of the business risk.
28
57. For example, some U.S.A. electric utilities are already making decisions in anticipation of some future policies
29
to limit greenhouse emissions.
30
58. These issues also arise among countries: countries with large coal deposits will find the value of their natural
31
wealth eroded, and quite naturally will be less enthusiastic about international agreements having that
32
consequence.
33
59. The studies show a marked variation in GDP losses across models. For example, stabilizing emissions at their
34
1990 levels is estimated to reduce U.S. GDP by .2% to .8% in the year 2010 - roughly a $20 billion to $80 billion
35
loss for that year. Estimate of the costs of reducing emissions by 20% below 1990 levels in the year 2010 range
36
from .9% to 1.7% of GDP. Aggregated models (top-down) have generally reported higher costs, while
37
disaggregated models (bottom-up) have shown lower costs. Chapter 9 contains a more complete discussion.
38
39
These GDP losses occur when the carbon taxes lead to investments that are more expensive than those that would
40
take place in the absence of the taxes. The higher the carbon taxes, the greater the investment in price-induced
41
conservation and the more fuel switching toward the less carbon-intensive substitutes.
42
43
The overall impact of a carbon tax will depend not only on the size of the tax but also on the uses to which the
44
revenues are put. In the standard EMF scenarios, it was assumed that tax revenues will be redistributed in a
45
neutral manner (i.e., without affecting the marginal tax rates). There are, of course, numerous ways in which tax
48
1
revenues can be used. These include: reducing budget deficits; reducing marginal rates of income, payroll,
2
corporate or other taxes; granting tax incentives to preferred activities; or increasing the level of government
3
expenditures. The costs of the tax will vary widely depending on how the revenues are recycled.
4
.Top-down models estimate that for developing countries there exist low-cost options to reduce emissions in
5
the near term, but eventually costs would exceed 1 to 2% of GDP (EMF 1993). For economies in transition,
6
because of historical inefficiencies and energy subsidies, there exist large opportunities to reduce emissions at
7
little or no cost. For developing countries, problems of informal economies make hard estimates difficult, but
8
the cost of stabilizing emissions would likely be large enough to cut into economic growth.
9
Recent comparisons indicate that the most important differences between top-down and bottom-up models arise
10
from differences in input parameters, rather than from differences in model structure.
11
Government institutions and regulations often hinder the efficient use of energy. The developing countries
12
are least able to absorb the costs of these inefficiencies. Thus, while some developing countries argue that they
13
cannot afford to reduce greenhouse emissions , the same countries have the most to gain from reforming
14
government-caused inefficiencies. At least in the short run, international agreements committing countries to
15
eliminate at least the most egregious of these practices might go a long way to addressing the problem of
16
emission reductions,
17
18
This may also be a problem with electricity generated by the private sector, as regulation has historically set
19
price equal to average cost, rather than allowing it to match the competitive price. In many countries, the increase
20
of competitive pressures has moved electricity prices closer to the marginal cost of production.
21
64.Full utilization of non-fossil fuel energy sources, taking account of other environmental impacts.) For example,
22
when hydroelectric power generation, which does not increase greenhouse emissions, can be cost-effectively
23
expanded without other environmental effects, it should be done.
24
25
Eliminating Regulations Impeding Efficient Energy Utilization. Many, perhaps most, countries have a host of
26
regulations which increase energy use as they impede economic efficiency. For instance, the United States has
27
had a policy of restricting oil exports to Alaska. Whatever the merits of that policy, it has forced Japan to import
28
oil from Indonesia and Saudi Arabia. World oil transportation costs have thus been greatly increased, at the
29
expense of the American economy. Another example of government reform, included in the US' Action Plan
30
(1993), encourages efforts to expand and improve natural gas markets through continued regulatory reform. These
31
reform efforts include guidelines to allow greater natural gas use in the summer in coal- and oil- fired power
32
plants.
33
34
Other regulation. Unintended effects of many tax, expenditure and other policies have contributing further to
35
inefficiencies in land use. Among the unfortunate effects of the U.S. Superfund program for the management of
36
hazardous wastes has been the creation of large unoccupied holes in the center of major cities.
37
.Consider the following thought experiment: compare an optimally designed road system which only carried
38
cars; and contrast that with an optimally design road system which also carries trucks. The incremental cost of
39
carrying trucks is, in most countries, much larger than the proportionate share of the cost they bear in gasoline
40
taxes and other fees.
41
66.For example, in many countries, governments have taken an active role in the dissemination of information
42
to the agriculture sector. These programs are in some measure responsible for the large increase in agricultural
43
productivity in countries with agricultural extension services.
44
49
1
67.This issue has been addressed in several papers. See, in particular, Stiglitz [1975, 1984] and Grossman [1981].
2
This is both because those who would, under "true" disclosure, be at a competitive disadvantage have an
3
incentive to add "noise" and because there are strong market forces for product differentiation; in markets with
4
homogeneous commodities, profits, even with a limited number of suppliers, will be driven to zero (in Bertrand
5
competition.) For a discussion of these and related issues, see Salop [1977], Salop and Stiglitz [1977, 1982],
6
Stiglitz [1977, 1988].
7
69. The standard reference in the organizational literature is March and Simon Economic theories emphasizing
8
the non-value maximizing behavior of managers include the works of Baumol [ ], Marris [ ], and Leibenstein
9
[ ]. The principal agent literature [Ross, 1973, Stiglitz, 1974] provided the informational micro-foundations for
10
understanding the divergences of interests. See Stiglitz [1988]. A more recent overview is provided by Stiglitz
11
[ ] and the symposium in the Journal of Economic Perspectives.
12
70.This key point, while noted in March and Simon's [1958] original work, was elaborated upon by Radner [
13
] and Hannaway [ ].
14
71. The facts that time is a scarce commodity and that decision making in large organizations is decentralization
15
do not in themselves constitute a market failure; they do not prove that resources are not efficiently allocated
16
given the real constraints facing society, which include time. However, Greenwald and Stiglitz [1986, 1988] have
17
established a very general theorem showing that when information is imperfect and costly, market equilibrium
18
is, in general, not (constrained) Pareto efficient. Thus, there is no presumption concerning the efficiency of the
19
market economy, even in the absence of the kinds of externality and public goods problems that are associated
20
with greenhouse gases. For a more extended discussion, see Stiglitz [1994].
21
22
One of the main insights of recent advances in the economics of information is to provide a sound
23
micro-foundations for these theories of the firm. And indeed, the importance of the limitations on the availability
24
of information, and the consequent importance of attention directing efforts applies to individuals as well as to
25
organizations. Some studies have suggested that the limited success of the special tax provisions in the United
26
States designed to encourage savings (IRA accounts) was primarily due to the competitive efforts of banks to
27
recruit these accounts, and the attention which savings got as a result.
28
29
2.Network externalities are manifested in other ways: builders fail to install energy efficient light bulbs, because
30
customers dislike them, because stores do not carry replacements; and stores do not carry them because the
31
demand for them is too low.
32
33
When there are important network externalities, market equilibria are frequently inefficient. The economy
34
might, for instance, get "stuck" in the wrong equilibrium. Government action can, in these instances, "force" the
35
economy to move from one equilibrium to another.
36
37
73.This is not the only explanation of differences between bottom-up and top-down models. There are several
38
other features of market behavior that bottom-up models often ignore.
39
40
(a) Hidden Costs: Consumers value a range of attributes difficult to include in an engineering model. For
41
example, auto buyers value not only initial costs and fuel economy (which computer models can easily calculate),
42
but also performance, safety, and durability, which they typically do not.
43
44
(b) Divergence between laboratory and in-use performance: Especially for new technologies, actual energy use
45
often differs significantly from energy use calculated in the laboratory. It is the latter upon which purchasers
50
1
focus.
2
3
(c) Variation across individual consumers: Engineering models generally assume an average consumer; actual
4
consumers may display a wide range of characteristics and usage patterns. Except when demand functions are
5
linear in the relevant variables, the consumption of the "average" individual is not equal to the average
6
consumption; and what is optimal for the average person may not be optimal for a significant fraction for the
7
population.
8
74.For a survey, see Jaffee and Stiglitz [ ]. The basic theory of credit rationing was developed in Stiglitz and
9
Weiss [1981] and the theory of credit rationing is developed in Greenwald, Stiglitz, and Weiss [1984] and Myers
10
and Maljuf [1984].
11
75. This generally accepted methodology is, for instance, reflected in the guidelines issued by the Office of
12
Management and Budget in the United States for the evaluation of projects and regulations. The applied literature
13
does not address the question of whether this procedure is appropriate in the presence of certain types of time and
14
risk non-separabilities.
15
76. Though if the variance of the net benefits is increasing over time in a particular manner, the differences in the
16
two methodologies may not be large.
17
77.Again, under certain restrictive conditions, where the shadow value of a capital constraint is changing
18
systematically over time, the differences in the two methodologies may not be great.
19
78.See Wilson Though the original discussion of winners curse focused around bidding in auctions, it has
20
subsequently come to be applied to a range of other market phenomena. See, e.g. Stiglitz [ ].
21
79.For a discussion of the role of the state in capital markets, see Stiglitz [1994].
22
80.In some industrialized countries, energy efficient home mortgage lending may help correct the problem.
23
Lenders generally set criteria for the maximum loan amount based on the borrowers' ability to repay, which in
24
turn depends on income and wealth. The fact that a particular expenditure which would enhance efficiency and
25
reduce utility bills is not given special attention. Energy efficient mortgages provide funds to households to
26
make energy efficiency enhancing investments intended to pay for themselves, i.e. result in reduced utility bills
27
equal to or greater than the interest payments. With capital constraints, builders may have an incentive to trade
28
off initial capital costs for higher maintenance costs (lower energy efficiency). Building codes specifying minimal
29
levels of energy efficiency and full disclosure of expected life cycle energy costs may help address these market
30
distortions.
31
32
33
81. A country that rapidly depletes its natural resources may show a high rate of growth under conventional
34
income accounting, but a lower rate of growth when resource depletion is taken into account. Repetto (1991,
35
1992) calculated resource-adjusted GDP for several countries rapidly harvesting their stocks of hardwoods and
36
other resources, arguing that conventional measures sharply overstated GDP.
37
82.Cobb and Daley (1989) have even claimed that U.S. per capita GDP, when adjusted for environmental damage,
38
was stagnant between 1950 and 1986. This assertion is hard to reconcile with the steady improvement in most
39
measures of environmental quality since 1970, when measurements standards were established.
51
1
83. Analysts now use two methods to estimate stocks. The first assumes a fixed stock of a natural resources such
2
as oil. Consumption of oil then depletes the stock by the amount of consumption. The second begins by treating
3
discovered reserves as the asset. Thus, additions to reserves increase the asset, while consumption reduces it.
4
If in any given year, new discoveries match resource utilization, then according to this method no net depletion
5
has occurred.
6
84. Many researchers have noted deficiencies in standard national income accounts. First, national income
7
accounts do not, in general, provide an adequate measure of welfare; second, they do not provide the correct
8
information relevant for making policies relevant to sustainable development. Sustainable development is
9
concerned with society's resources; an economy is growing when its resource base (capital stock combined with
10
natural resources) is growing. GDP does not, and is not intended to be, a measure of resource availability. Firms
11
have two sets of accounts--cash flow (income) statements and balance sheet statements. GDP is a statement of
12
the former type.
13
14
Standard accounting procedures require that firms, in an attempt to present an accurate account of "true income,"
15
take account of depreciation. GDP measures gross output; it does not take into account depreciation, either of
16
natural or physical capital stocks. The reason is simply that it is hard to get accurate measures of depreciation.
17
18
Net national product does take into account depreciation, the change in capital stock. And it is this account which
19
should be most subject to criticism, since it accounts for changes in the physical capital stock, but not in other
20
capital assets, in particular, environmental assets and natural resources.
21
22
A number of difficult conceptual problems face the analyst defining levels and changes in levels of these assets.
23
24
25
First, how should the "stock" of natural resources be defined? Coal poses perhaps the easiest situation. The
26
location of coal reserves is known. Costs of extraction are high, so the rents (the value of coal in situ) is low.
27
The depletion can be measured not by the coal used times the market price, but the coal used times the in situ
28
value. But for oil and other minerals, information about where reserves are located is vital. Two models have
29
been proposed. One sees the world as having a fixed stock of natural resources (say oil). When one uses oil,
30
one is depleting this stock. Thus, to calculate the value of depletion, one does not need to know the entire stock;
31
the flow (the amount of oil consumed) provides an accurate measure of the change in stock.
32
33
The alternative model looks at the size of discovered reserves. Reserves are treated as the asset. Additions to
34
reserves thus are viewed as increasing the resource base. If in any given year, new discoveries match resource
35
utilization, then there is no net depletion. This is the approach being taken by the U.S. Department of Commerce.
36
This accounting framework would be correct if there were an infinite supply of the resource (reflected in zero
37
rents); the essential "capital" good is information about where the resource is located.
38
39
Environmental assets--such as air quality--present another set of problems, because there are not market prices
40
to value the asset. Dynamic optimization problems of the kind described earlier can be used to calculate shadow
41
prices. How sensitive these shadow prices are to specific assumptions remains to be investigated.
42
43
Accounting systems do not, however, have to aggregate all information together. Just as information about
44
longevity and other indicators of well being (see below) serve to complement information from national income
45
accounts concerning standards of living, so too can information about physical environmental measures be used
46
to complement information from the extended national income accounts.
47
85. There is some concern that excessively broad and long patents may actually impede innovation. When
48
technological progress occurs by building on previous innovations, later innovator require permission of earlier
49
innovators to realize the returns on their innovation. While advocates of broad patent coverage argue that the
52
I
parties always arrive at efficient bargaining solutions, critics point out that the outcomes of bargaining models
2
with incomplete information often entail large inefficiencies.
3
86. Matters are more complicated, since the patent does not reward the innovator with his marginal contribution--
4
the increase in the present discounted value of benefits as a result of the innovation occurring earlier than it
5
otherwise would have occurred. For a fuller discussion, see Stiglitz [1994], Dasgupta and Stiglitz [1980a, 1980b],
6
Barzel [ ].
7
87.
If less developed countries fail to implement fully a set of corrective taxes or tradeable permits, or if less
8
developed countries fail to adopt and enforce effectively intellectual property rights, there will be insufficient
9
incentives to produce energy and emission savings innovations, particularly those appropriate for the level of
10
technological knowledge, human capital, and factor prices in those countries. If less developed
11
countries do take these actions, there is concern that they will result in higher prices of innovations, and thus the
12
pace of adoption will be retarded. An effective form of aid, targeted to reducing greenhouse gas emissions, may
13
be subsidies directed at producing appropriate energy saving and emission reducing technologies for LDCs.
14
15
88. Edmonds et al. (1994) has studied the importance of available advanced energy technologies such as those
16
proposed by Johannsen (1993). Edmonds et al. use the Edmonds-Reilly-Barnes model for energy related
17
greenhouse gas emissions; the MAGICC model for atmospheric composition, climate response, and sea level rise;
18
the IPCC scenario IS92a (IPCC, 1992) as the reference base case and five alternate energy scenarios that are far
19
more advanced over today's energy supply and transformation technologies. The five energy scenarios are:
20
21
a. advanced fossil fuel technologies
22
b. advanced liquified hydrogen fuel cells
23
C. advanced hydrogen fuel cells without liquified hydrogen
24
d. low cost biomass
25
e. accelerated rate of exogenous end-use energy intensity improvement
26
27
Combined, the energy technologies reduce annual emissions from fossil fuel use to levels that stabilize
28
atmospheric concentrations below 550 ppmv (i.e. double the concentration prior to the Industrial Revolution).
29
The tax rate, assumed to apply globally, used was the marginal cost of stabilizing fossil fuel carbon emissions
30
in the reference case. With values reflected for only carbon dioxide emission reductions, the cost of global
31.
emission reductions grow from approximately $35 (US) Billion in 2005 to $230 (US) billion per year in the year
32
2095. With advanced fossil fuels, low cost solar electric power, low cost fuel cell vehicles, the present discounted
33
value of adding low cost biomass fuel to the energy technology bundle is almost half a trillion dollars (US). The
34
present discounted value of the advanced energy technologies taken together is $1.8 trillion (US).
35
36
The introduction of advanced biomass energy production technology was found to play a key role in reducing
37
emissions. Biomass energy at $2.00/GJ growing to become the core energy supply technology by 2050 could
38
significantly reduce emissions highlighting the potential role of technology development and deployment relative
39
to that of fiscal and regulatory intervention.
40
41
These results should be viewed as illustrative rather than predictive. In this analysis, the gains from introduction
42
and deployment of advanced energy technologies is dependent on the order of technologies evaluated in the study.
43
44
45
89. The literature has identified three types of permit systems. The ambient permit system (APS) works on the
46
basis of permits defined according to exposure at the receptor points. Each polluter, then, may face quite
47
complex markets - different permit markets according to different receptor points, and hence different prices. The
53
1
simpler emissions permit system (EPS) issues permits on the basis of source emissions and ignores what effects
2
those emissions have on the receptor points. Within a given region or zone, the polluter would have only one
3
market to deal with and one price. Finally, there is the pollution offset (PO) system wherein the permits are
4
defined in terms of emissions, trade takes place within a defined zone. However, the standard has to be met at
5
all receptor points. The exchange value of the permits is then determined by the effects of the pollutants at the
6
receptor points. The PO system thus combines characteristics of the EPS and the APS. (Pearce and Turner
7
1990) These distinctions are of limited relevance for greenhouse gases, where what is of concern is global
8
emission levels. The specific location of the emissions is of no concern.
9
10
90. The choice between taxes and tradeable permits depends on the objectives of the policy maker and nature of
11
the uncertainty about the marginal cost and marginal benefit curves for carbon emission reductions (Weitzman,
12
1974). Theory tells us that if the nature of the curves is known with very little certainty, but the marginal cost
13
curve is known to be relatively steeper (i.e. a change in the level of pollution allowed brings about a greater
14
change in the marginal costs of mitigation compared to the marginal benefits) then taxes should be the policy of
15
choice. This is because, in this case, an erroneous estimation of the optimal tax rate will lead to a relatively small
16
deviation from the optimal pollution level. On the other hand, an erroneous estimation of the optimal level of
17
total emissions in a permit scheme will lead to a relatively large deviation from the optimal cost of the permits.
18
19
If the marginal benefit curve is known to be relatively steeper than the marginal cost curve, however, tradable
20
permits are the better option. Here, an erroneous estimation of the optimal tax rate will lead to a relatively large
21
deviation from the optimal level of emissions while an erroneous estimation of the optimal level of emissions in
22
a permit strategy will lead to a relatively small deviation from the optimal cost of the permits.
23
24
In the case of greenhouse gas emissions, the time horizon for adjustment is sufficiently long that many of these
25
uncertainties become less important. If the tax rate initially chosen yields too high a level of emissions in one
26
year, it can be increased, and the net impact of the erroneous initial estimate on global warming (or the total cost
27
of achieving a given level of atmospheric concentration) will be negligible. In any case, as the earlier discussion
28
on sequential decision making has emphasized, there is likely to have to be continued revisions in either tax rates
29
or permit levels.
30
31
32
Still, there is some argument that the required adjustments under a permit scheme may be less burdensome.
33
(Tietenberg 1992) For instance, if the authority feels that the old standard needs some tightening they may enter
34
the market themselves and buy some of the permits, holding them out of the market.
35
36
There must be effective, competitive markets in tradeable permits, if such schemes are to achieve efficient
37
outcomes. There are transactions costs of running such schemes, just as there are transactions costs associated
38
with collecting tax revenues. Whether transactions costs gives one system a decided advantage over the other
39
is not clear. There seems to be no compelling reason to believe that good markets in tradeable permits would
40
not develop.
41
42
43
91. It is possible to design allocations of trading permits which (i) on average, impose no net burden on developing
44
countries (thus conforming to the ability to pay principle); (ii) provide those economies which are growing faster
45
per capita with commensurately greater permits, thus imposing no net drag on economic growth, provided the
46
economy exhibits an increase in fuel efficiency at least equal to the average of fast growing LDCs; and (iii)
47
rewards those economies which are able to reduce greenhouse gas emissions faster than benchmark, either through
48
greater control of population growth, through larger increases in energy efficiency, or through switching from
49
higher to lower carbon fuels.
50
54
1
The extent to which individual circumstances of countries should be taken into account in setting benchmarks
2
remains a question for international negotiations; to the extent that high emissions is due to natural endowments
3
(e.g. the availability of coal rather than natural gas as a source of energy), a persuasive case can be made for
4
benchmarks to reflect initial emission levels. To the extent that high emissions are due to inappropriate energy
5
pricing policies, the case that benchmarks should reflect initial emission levels is far more tenuous.
6
92. Similarly, some developing countries have asked, should the North be given higher levels of permits, simply
7
because it has, in the past, been the chief source of greenhouse gases?
8
93. For a discussion of the polluter pays principle, see above.
9
94. Standard tradeable permit schemes essentially take the revenue from a carbon tax, and distributes it to current
10
user emitters, rather than using the revenue to reduce other taxes. An alternative to these standard schemes is
11
for the government to auction off the tradeable permits.
12
13
If taxing carbon leads to reduced labor supply or reduced savings, then government revenues from wage or capital
14
taxes may be reduced, more than offsetting the direct revenue gain from the carbon tax. Cross elasticities of this
15
magnitude are unlikely, though any such cross elasticity will reduce the net gain from the carbon tax. The
16
magnitude of the double dividend has been the subject of some dispute, with Goulder ] and taking opposite
17
views.
18
19
95. Two main studies provide insights into the root of the variance in estimates of the economic effects of carbon
20
taxes: The Energy Modeling Forum Study (12) and the OECD comparison project. In each case, sophisticated
21
sensitivity analyses are run by standardizing key economic assumptions along with use of common reference case
22
scenarios of reductions. The magnitude of the effect on economic growth will depend both on assumptions
23
concerning the effect of carbon taxes on savings and labor supply, and the induced investment to offset the higher
24
energy prices. If higher energy prices do not lead to much capital substitution and if the cross elasticity with
25
savings and labor are low then the likely effect on economic growth will be small.
26
27
28
OECD Project on Economy-Wide Cost Estimates of Carbon Taxes:
29
An OECD model comparisons project was conducted to compare economy wide estimates of the effects of carbon
30
taxes. Time horizons as well as the key economic assumptions on growth, population and resource prices, and
31
the reduction scenarios for six global models were standardized. The global models compared were the GREEN
32
model, the IEA model and four North American models [Edmonds-Reilly Model (ERM), Global 2100 of Alan
33
Manne and Rich Richels (MR), the Carbon Rights Trade Model (CRTM) of Tom Rutherford, and the Whalley-
34
Wigle model] (Dean 1994).
35
36
There is significant variation in tax rates and costs for the same amount of emissions reduction among models
37
due to differing assumptions regarding several key considerations. Several factors explain the differences between
38
model results. The most important factors are:
39
40
a. The degree of substitution between fuels the ease with which producers and consumers can switch from
41
high-carbon content fuels to low-carbon content fuels;
42
b. Expectations about future energy prices and taxes;
43
C. The speed of emissions reduction;
44
d. The way in which revenue is recycled;
45
e. The treatment of the removal of energy subsidies; and
46
f. Assumptions regarding backstop technology and a host of other technical and economic factors.
47
48
Because of the varying approach to these questions, the range of estimated tax rates and costs are quite wide.
55
I
For a 45% reduction in baseline emissions by 2020, the required tax would be in the range of $150-$325 per ton
2
of carbon and the cost might be in the range of 1.5%-2.9% of world GDP. A 70% reduction in baseline
3
emissions by 2050 could require a tax between $230 and $880 and a loss in world GDP of 2.4%-3.8% (Dean
4
1994).
5
6
The required carbon taxes and associated costs vary significantly across regions in all of the models. This
7
indicates that the same proportional reductions in emissions across all regions would give rise to very different
8
costs in different regions and would thus be globally inefficient - with great potential for savings in the global
9
cost of reducing emissions through the use of emission trading between countries or regions (see section 1.5.2.1)
10
or a global carbon tax.
11
12
Three insights emerging from the OECD study are (Dean, 1994):
13
a. Small amounts of emissions reduction can probably be achieved with low taxes;
14
b. Large reductions can only be achieved at high tax rates (i.e. marginal reduction costs rise with emissions
15
reductions).
16
C. Carbon-free backstop technologies are likely to slow the rise of the carbon tax, or halt it altogether, if they
17
are available at constant marginal cost.
18
19
Energy Modeling Forum Project 12 (Impact of Carbon Emission Control Strategies)
20
A recent study at Stanford, the Energy Modelling Forum Project - 12, examines the cost of reducing CO2
21
emissions (Energy Modeling Forum, 1993). A diverse group of economic models, employing common
22
assumptions for selected numerical inputs, were used to analyze a standardized set of emission reduction
23
scenarios. In all, 14 top-down models participated in the study.
24
25
The EMF model comparison provides the most comprehensive application of top-down methodologies to date.
26
The study addresses a wide range of policy questions. How large are emissions likely to grow in the absence
27
of controls? How much market intervention will be required to meet alternative targets? What will be the price
28
tag? In exploring economic costs, the modelers were asked to examine the impacts of timing, research and
29
development, and revenue recycling.
30
31
The EMF exercise provides a wealth of useful information for policy making. Although the focus was primarily
32
on the U.S., many of the insights are applicable to developed countries in general.
33
34
In selecting parameters for standardization, the EMF study focused on what were felt to be the most influential
35
determinants of mitigation costs. These included: GDP, population, the fossil-fuel resource base, and the cost
36
and availability of long-term supply options. in addition, although the EMF models differed considerably in their
37
technology representation, the study attempted to impose uniformity with regard to world oil prices, the oil and
38
gas resource base, and the cost of backstop technologies. For its reference case, EMF adopted the average of the
39
IPCC high and low economic growth cases (IPCC, 1990c). Also for consistency with the IPCC, the study
40
adopted the population growth projections based on Zachariah and Vu.
41
42
The modelers generally used taxes based on the carbon content of the fossil fuels in order to achieve a prescribed
43
emissions reduction. The magnitude of the tax provides a rough estimate of the degree of market intervention
44
that would be required to achieve the carbon emissions target. Estimates range from $20 to $140 per ton for the
45
carbon taxes required to hold emissions at 1990 levels in 2010. Estimates of the carbon taxes required to reduce
46
emissions by 20% below 1990 levels in 2010 range from $50 to $330 per ton.
47
48
Two parameters are particularly important in explaining the differences in tax projections: the price elasticity of
49
energy demand and the speed with which the capital stock adjusts to higher energy prices. Neither were
50
controlled in the EMF experiments. Those models using lower price elasticities required higher taxes to achieve
51
the same emissions goal. Those models which assumed greater malleability of capital required lower taxes.
56
1
2
96. Rapid capital stock retirement may add to the cost of immediate CO2 reductions. Also, reductions may be
3
cheaper later because of technical changes in the intervening years.
4
97. Indeed, the "ratchet effect," a commonly observed phenomena in the command and control economies
5
(Stiglitz, 1975, Weitzman, 19 ) leads firms to just satisfy the target.
6
98. Two examples from the United States Clean Air Act: 1) the requirement that coal-burning power plants
7
install stack-gas scrubbers even if they burn low-sulfur Western coal; 2) the mandates for ethanol in reformulated
8
gasoline.
9
99. World Commission on Environment and Development, Our Common Future, Oxford: Oxford University Press,
10
1987.
11
100.It has been argued that new species and varieties produced by genetic engineering may be able to more than
12
offset the loss in genetic variability from climate impacts, particularly because of the increasing capacity to direct
13
those mutations toward socially desirable objectives.
14
101. Other views represented in the debate on sustainable development but not generally accepted in the
15
economics profession include those of Daly (1977, 1980) and Daly and Cobb (1991), and those of the new field
16
of ecological economics (Costanza 1991).
17
102. Tariq Osman Hyder, "Climate Negotiations: the North/South Perspective," in Mintzer, ed. Hyder has headed
18
the Pakistani delegation at meetings of the INC, and GEF (Global Environment Facility).
19
103. "Principles of the G-77 and China on the Climate Convention," cited in Tariq Osman Hyder, "Climate
20
Negotiations, the North/South Perspective," in 1. Mintzer, ed., Confronting Climate Change: Risks, Implications,
21
and Responses, Cambridge: Cambridge University Press, 1992.
57
1
1.9
REFERENCES
2
3
Abreu (**date**)
4
5
Arrow (**date)
6
7
Ausubel, J.H. 1990, Conventional Wisdom About Impacts of Global Change. Rockefeller
8
University, New York. ?pp.
9
10
Baumol (**date**)
11
12
Barzel (**date**)
13
14
Bongaarts, J. 1994. "Population Policy Options in the Developing World," Science (263),
15
11 February 1994, pp. 771-776.
16
17
World Commission on Environment and Development 1987. Our Common Future.
18
London: Oxford University Press.
19
20
Burniaux, J.M., Martin, J.P. and Oliviera Martins, J. 1992. 'The effects of existing
21
distortions in energy markets on the costs of policies to reduce CO2 emissions: evidence
22
from GREEN.' OECD Economic Studies No. 19
23
24
Chao, Hung-po, 1992. "Managing the Risk of Global Climate Catastrophe," Electric Power
25
Research Insitute.
26
27
Cline, W.R. 1992. The Economics of Global Warming. Institute for International
28
Economics, Washington, D.C.
29
30
Coase (1990)
31
32
Coase (1960)
33
34
Dasgupta, Gilbert, andStiglitz (**date**)
35
36
Dasgupta, and Stiglitz (1980a, 1980b)
37
38
Dasgupta, P. and K.G. Maler 1990. The Environment and Emerging Developing Issues,"
39
Proceedings of the World Bank Annual Conference on Development Economics, pp.
40
101-131.
41
42
Dean, A. 1994. 'The Effectiveness of Carbon Taxes at the International Level.' In:
43
Climate Change: Policy Instruments and their Implications: Proceedings of the Tsukuba
44
Workshop of IPCC Working Group III, 17-20 January, 1994. pp.46-59.
58
1
Dixon et al. (1994)
2
3
Edgeworth, Mathematical Physics,
4
5
Edmonds, J., Wise, M. and MacCracken, C. 1994. 'Advanced Energy Technologies and
6
Climate Change: An Analysis Using the Global Change Assessment Model (GCAM).'
7
Global Environmental Change Program, Pacific Northwest Laboratory. Wahington D.C.
8
Draft.
9
10
Energy Modeling Forum 1993. EMF 12 Global Climate Change: Impacts of Greenhouse
11
Gas Control Strategies. Palo Alto, Ca.
12
13
Environmental Protection Agency (EPA) 1989. 'Sea level rise.' In Report to Congress
14
on the Potential Effects of Global Climate Change on the United States.
15
16
Fair (1970)
17
18
Fankhauser (**date**)
19
20
Farrell (**date**)
21
22
Gleick, P.H. 1987, 'Regional Hydrologic Consequences of Increases in Atmospheric CO2
23
and Other Trace Gases'. Climatic Change 10, pp. 137-61.
24
25
Goulder (**date**)
26
27
Greenwald and Stiglitz (1986, 1988)
28
29
Greenwald, Stiglitz, and Weiss (1984)
30
31
Grossman (1981)
32
33
Grubb, M. et al. 1993. 'The Costs of Limiting Fossil-Fuel CO2 Emissions: A Survey and
34
Analysis.' Annual Review of Energy and the Environment 18. pp.397-478.
35
36
Grubb, M., Minh Ha Duong and Thierry Chapuis 1993. "Optimising Climate Change
37
Abatement Responses: on inertia and induced technology development," IIASA Workshop
38
on Integrative Assessment of Mitigation, Impacts, and Adaptation to Climate Change,
39
IIASA, October 1993.
40
41
Grubb, M., Thierry Chapuis and Minh Ha Duong 1995. "The Economics of Climate
42
Change: Implications of adaptability and inertia for optimal climate policy," forthcoming.
43
44
Hall et al. (1993)
59
2
DECISION MAKING FRAMEWORK TO
ADDRESS CLIMATE CHANGE
K.J. ARROW, J. PARIKH, G. PILLET
Contributors:
P.G.Babu, A.Beltratti, G.Chichilnisky, S.Fankhauser, S.Faucheux, G.Froger, F.Gassmann,
E.F.Haites, W.Hediger, J.-Ch.Hourcade, S.Kavikumar, M.G.Morgan, K.Parikh, D.Pearce, S.Peck,
R.Richels, R.Schubert, C.Suarez, U.Wellenmann
I
Decision Making Framework / Draft 05 / December 1994
2
CHAPTER 2
3
Draft 05 / December 1994
I
2
2.5
Economic Formulation of Individual and Collective Decision Making
27
3
2.5.1
Accommodating the supplementary uncertainty caused by climate change
28
4
DECISION MAKING FRAMEWORK TO ADDRESS
4
2.5.1.1
Strong uncertainty
29
5
2.5.1.2
Abrupt change adds unpredictability to uncertainty and risk
30
5
CLIMATE CHANGE
6
2.5.2
"Rationalities"
32
6
7
2.5.2.1
Utility-based decision criteria
33
8
2.5.2.2
7
Bounded rationality
35
9
2.5.2.3
Procedural rationality
35
8
IPCC Working Group III
10
2.5.3
Responses to uncertainty and risk
36
9
Writing Team II
11
2.5.3.1
Uncertainty and risk decision rules
37
12
2.5.3.2
Decision analysis techniques
38
10
13
2.5.3.3
Addressing strong uncertainty
39
14
2.5.4
Collective decision making
41
11
15
2.5.4.1
Steps in negotiating international environmental treaties and agreements
42
16
2.5.4.2
Collective decision criteria
43
12
TABLE OF CONTENTS
17
2.5.4.3
Decision procedure
44
18
13
19
20
2.6
A Sequential Decision Making Framework to Hedge Against Impacts of
14
2.1 Introduction
21
Uncertainty and Change
46
/
22
2.6.1
15
Decision making paths: Which strategy would be best?
46
2.1.1 Questions and issues
/
23
2.6.1.1
Scenario analysis
46
16
17
2.1.2 Climate change uncertainties and their relationships
24
2.6.1.2
Sequential decision analysis
47
2
18
25
2.6.1.3
Clearing up the uncertainties? The value of improved information
50
2.1.3 SCOPE OF THE CHAPTER:
6
26
2.6.2
Exploring mutual insurance decisions and markets for risks
52
19
20
2.2 NATURE AND GENESIS OF THE CLIMATE CHANGE PROBLEM LEADING TO THE FCCC 7
27
2.6.2.1
Use of financial markets
53
21
28
2.6.2.2
Insurability of risks associated to climate change
54
2.2.1 CONCENTRATIONS NOT EMISSIONS:
7
22
29
2.6.2.3
A market framework to respond to collective risks
55
2.2.2 ABSORPTION CAPACITY:
8
23
2.2.3 NATURE OF RISK:
30
2.6.3
Portfolio analysis analogy
58
8
24
31
2.6.3.1
2.2.4 COMMON BUT DIFFERENTIATED RESPONSIBILITIES:
Selecting a portfolio
59
9
25
32
2.6.3.2
No operational model is available
60
2.2.5 FCCC AS A KEY ELEMENT FOR DECISION MAKING
10
26
33
2.6.3.3
Three main forms of policy options
61
2.3 COLLECTIVE DECISION MAKING
27
II
34
28
2.3.1 How AND WHY COLLECTIVE DECISION MAKING:
11
35
2.3.2 FACTORS FOR COLLECTIVE DECISION MAKING:
36
2.7
REFERENCES
66
29
12
30
2.3.2.1 Vulnerability of food production systems:
12
31
2.3.2.2 Anthropocentric Vs Ecocentric approaches:
13
32
2.3.2.3 Poverty and Climate Change:
14
33
2.3.2.4 Intergenerational equity and discounting:
15
34
2.3.3 IMPLICATIONS OF COLLECTIVE DECISION MAKING:
16
35
2.3.3.1 Cost of delay and slippages (Free riding through delay):
16
36
2.3.3.2 Valuing Common Property (and Paying for It):
17
37
2.3.3.3 Accepting Adaptation
18
38
2.3.3.4 North-South transfers:
19
39
2.4 PROCESS OF COLLECTIVE DECISION MAKING
20
40
2.4.1 A SUMMARY OF THE STEPS IN ITERATIVE DECISION MAKING:
41
21
2.4.2 CRITERIA FOR INTERREGIONAL ALLOCATIONS:
42
22
2.4.3 NORTH-SOUTH COOPERATION
43
24
44
45
46
1
Chapter 2
2
DECISION MAKING FRAMEWORK TO ADDRESS
3
CLIMATE CHANGE
4
5
6
Summary for 2.5 & 2.6
7
8
9 Decision makers face strong uncertainties that are characterised by insufficient knowledge and lack of
10 information. There also is the possibility of large damages (low probability) whether climate change
11 is abrupt or not. In addition, models in the North do not usually build in the perspective of the South.
12 The presence of strong uncertainty, however, does not mean inaction. It means filtering paths of
13 action and choosing which strategy with which outcome would be best.
14 Decision theory leads to the assessment of alternative strategies and the valuation of likely outcomes.
15 As a consequence of insufficient knowledge and data, predictions of the rate and extent and timing of
16 climate change and their effects have been made in terms of scenarios rather than in usual probabilis-
17 tic terms. This situation reflects the gap between the information that is available on climatic pro-
18 cesses and economic effects and the one that would be relevant to Bayesian inference. Bayesian
19 scheme has two limitations with respect to climate change: first the existence of strong uncertainties
20 and, second, the existence of low probabilities attached to large potential damages. Alternatives,
21 however, are available: risk assessment (Chapter 7), fuzzy logic, or evidence theory.
22 Economic criteria for choice explicitly or implicitly are utility-based criteria. Ultimate choices are
23 made by individuals. Yet, because climate change engages considerations that are new in their glob-
24 ality and relate to unusual time-spans, choices also concern collective decision making. Under collec-
25 tive decision making, the valuation of outcomes is matter of negotiation.
26 The decision making framework to address climate change is made of stratgies at three different lev-
27 els. Sequential decision making. Sequential decision making paths are elected to hedge against uncer-
28 tainty and change. "Act-then learn" (then act) decision strategy would be best. Hypothesis: there is an
29 agreement that uncertainties cannot be resolved before a decision is taken. Exploring insurance
30 analysis. The word 'insurance' is sometimes abused in climate debate. Yet, the insurance industry
31 (individual, fixed, short-term contracts) cannot afford climate change related risk sharing. New
32 institutions should be envisaged, e.g., a market framework with new financial institutions, mutual
33 contracts and securities to deal with collective risks. Constructing a portfolio of policy measures.
34 Numerous policy measures are available to limit the impacts of climate changes. The decision
35 maker's problem is much more complex than the choice of a single policy measure to the exclusion of
36 all others. In practice, it is a question of constructing a portfolio of policy measures.
37 Jan. 18. 1995
Chapter 2/v05/Summary for 25 & 2.6 - page 1/1
responsibilities for current concentrations and development needs? In other words, who
should bear the burden of emissions abatement at different phases?
2.1. INTRODUCTION
4. What kind of policies will achieve the desired reductions of concentrations of GHG gases
The
and in turn, the emissions?
Framework convention of climate change (FCCC) signed by many heads of state at
Earth summit at Rio in 1992 has been ratified by more than 50 countries by March 1994 and
5. Considering that uncertainties will always remain, shall we wait for more precise
is now legally binding. To carry out the deliberations of this convention, a decision making
information that reduces uncertainties? If we want to collect more information, and since
framework is needed that encompasses global and national concerns. While short-term
collection of information is expensive, on what aspects of uncertainties should we
concentrate?
decisions are governed by FCCC, there are some long-term considerations, say for several
In this connection, we note that according to the earlier IPCC report, even
decades, which also need to be looked into in the decision making framework by interpreting
A
present information is adequate to go for the goal of reducing global emissions by 60% for
their consequences until 2050 and beyond. In the present chapter we construct such a
stabilisation.
framework.
To answer these questions, a framework is needed which will be characterized by
2.1.1 Questions and Issues:
uncertainties at all levels. We say that a decision situation is governed by uncertainty if the
consequences of decisions and the state of nature are not known apriori. The agent(s) who
Climate change due to accumulation of greenhouse gases (GHG) in the atmosphere
makes the decisions is known as the decision maker(s) (DM in short).
poses many economic, political and scientific challenges which have to be dealt with in a
collective decision making framework by national and international decision making bodies.
2.1.2 Climate change uncertainties and their relationships:
Policy decisions to address climate change have to be taken in a complex setting characterized
It is fair to say that most decisions are made under some uncertainty or other. Consider
by uncertainty. Apart from uncertainties, these decisions often involve use of conflicting
for example, the uncertainties in decision to invest in an energy supply system. During the
criteria such as equity, efficiency, sovereignity and ecological concerns.
lifetime of the project there could be changes in demand, interest rates, availability of
Decisions are usually based on one's beliefs about the likely occurrence or not of
alternative technologies and environmental and other regulations that may change from time
uncertain events which are relevant to the decisions at hand. The decision regarding climate
to time. However, the uncertainities of climate change decision making are different from other
change have to be made not only in the near future, but have to be reviewed on a continuous
decisions because:
basis, or at least periodically. The major decisions concenrning climate change that need to
Climate change decisions have to be taken collectively and not by a single entity;
be faced by all countries and their citizens are:
The consequences of these decisions affect everyone differently and for a very long term
1. What should be the level of concentration of GHG gases that can be tolerated at different
(Distributional/regional effects).
times?
Any decision made now with wrong outcome can only be identified after a long lag, by
2. By how much should GHG emissions be reduced to achieve these chosen levels of tolerable
which time it might be too late.
concentrations?
It is subject to non-linearity and can Icad to large damages, irreversibility and even
3. How should these abatement levels be distributed among countries taking into account the
catastrophic surprises.
I
That there is a rise in the accumulation of GHG is agreed upon by everyone. What is
2
uncertain is the extent of climatic impacts such as rise in temperature, rise in sea level or
technologies and their impacts on reducing emissions. As shown by Hafele et al. (1981), it
occurrence of extreme events and thereby changes in ecological balances in biosphere. Some
Refecter
takes nearly forty to fifty years for a new energy carier to capture 50% of the market from the
expect that in the long term, efforts to contain these impacts and offsetting them would have
significant economic costs. However, there are others who say that there are additional
JAP
time it captures 1% of the market. Judging by this thumb rule, there are not many technologies
to look forward to in the very near future. Though many technologies look promising, they
benefits in undertaking "hedging strategies" (see box) such as savings of fuel, local pollution,
have yet to enter in global energy balances. Even if there are many successful demonstrations
traffic congestion and so on. The uncertainties concerned with the climate change debate here
of carbon free technologies, the ultimate test is to enter the global energy balances, which at
can be classified into the following categories:
present still consist of coal, oil, gas, hydro and nuclear energy¹. In fact, it is possible that
1.
nuclear energy which is a part of the global and national energy balances today, might make
Scientific uncertainties regarding the effects of emissions of GHG on climate change
an exit without reaching its full potential. Thus, lead times required for transition in energy
characterized by several important variables among many others, viz., mean temperature rise,
system are very large. The lead times required for abatement technologies to be effective and
1
sea level rise, precipitation changes and changes in frequency of extreme events such as floods,
extent of abatement possible by each, thus, become important factors in decision making.
cyclones and draughts. One could also call these uncertainties relating to geographical and
2.1.2.1 Relating scientific uncertainties with impact uncertainties:
atmospheric changes. The interrelationships among these variables for a given level of
How are these uncertainties related? If one could structure the relationships among
concentration are also uncertain. These are dealt with by working group I of IPCC.
them, decision making can be guided with better understanding rather than getting
2.
overwhelmed by uncertainties, and some action paths can become obvious. Scientific
Impact uncertainties corresponding to each of the above scientific uncertainties where it is not
uncertainties concerning the changes in climate variables for a given concentration level, can
clear how much impacts different levels of climatological variables will have on, say, biomass
be linked with impact uncertainties as follows. Considering only the temperature variable,
productivity, eg., that of agriculture and forests, species loss - terrestrial and marine, and other
figure 2.1 illustrates that for low levels of climate change one could be in the zone of
discomfort and damage. These are dealt with by working group II of IPCC.
adaptation. Further emissions could lead to higher concentrations, that could increase
3.
(femperature and increase damages. Say, at Dt=2°C, there could be "discomfort" zone; but,
Economic uncertainties (corresponding to impact uncertainties) concern the effects of Climate
with some positive probability, this zone could mean disaster for some. We refer to disaster
Change on national and therefore individual incomes due to physical impacts such as coastal
as "an event occurring suddenly and causing great damanage or hardship". For example, a
erosion or cost of increased airconditioning. Cline (1992) has elaborated these and calls them
small percentage in sea level rise might cause serious damages only to Bangladesh and
as benefits of avoiding climate change. Both the costs and benefits of abatement are uncertain.
Maldives but, might not have as much effect on the rest of the globe. On the other hand, we
Costs of abatement could be different for each region and could have different impacts on GDP
define catastrophe as "a sudden and widespread disaster." This can occur with a very small
growth of different nations. However, the uncertainties regarding the costs of abatement
probability for say a temperature rise of 4°C but with a higher probability with a temperature
measures are less than those associated with the benefits these measures bring. Here again, the
change of 6°C. This has to be avoided because apparent and hidden costs of global and local
same level of geophysical impacts could lead to different economic impacts to countries at
irreversible damages and extensive adaptation are high. Figure 2.1 shows that as the GHG
different levels of economic development. Several chapters in the present report deal with
accummulation increase, the severity of damages could increase.
economic and technological uncertainties, viz., chapters 7 and 8.
4.
Technological uncertainties concern the timely availability (lead times) and cost of reliable
I Biomass consumed in developing countries in the form of fuel-wood. crop residues and animal dung are not recorded
?
satisfactorily. They provide 10% to 90% of household energy for cooking in many developing countries.
3
4
2.1.2.2 Relating impact uncertainties with economic uncertainties:
1
Now we shift from "damage" to "damage costs" and what they could mean in terms of
)
economic impacts to a country. Damages have different impacts on different countries for three
.
reasons.
Probability
I. Impacts themselves may be higher in geophysical terms. For example, for a given rise in
global mean temperature, sea level rise in Indian subcontinent could be higher than in, say,
7
North America.
ДТ = 2°C
II. Even the same change say, 1°C temperature rise, could have more impact at 25°C average
temperature than at location with 15°C average temperature.
AT . 4°C
AT = 0.2°C
AT = 6'C
10
III.
Even for identical climate change, adapation to climate change imposes greater burden
"
on poor countries than rich countries, i.e., the same amount of adaptation costs says $1
12
billion, could have differnt economic impacts due to differences in population,
Adaptation zone
Discount and
Disaster zone
Catastrophic
13
economic resources and technological capability.
impacts can
Damage zone
zone
have marginal
Frequency of
14
When damage costs for adaptation are considered, as figure 2.2 shows, the zone for
costs mostly
Substantial costs
extreme events
Irreversible
avoidable through
of discomfort and
increase with
damages to
"
acceptable costs for a pcor country compared to a rich country may be much smaller and the
adaptation.
damage
disastrous
global
consequences for
ecosystem.
16.
some.
disaster zone may be much wider beginning at a modest level of climate change.
No
Increasing Impact of Climate Change
Impact
Figure 2.1: Schematic linkage between scientific uncertainties and impacts
Area under each probability distribution curve is 1 At At = 2°C. the probability that there is no or little impact
is much larger than at at - 6°C. Reverse is the case for catastrophic zone. Different curves are meant to
represent qualitative differences between "small" climate change and "large" climate change. and the range of
temperatures corresponding to each curve remains a subject of discussion. Similar relationship may be there for
different levels of sea level rise.
5
Hence, in this chapter we undertake a detailed analysis of the consequences of various issues
and attitudes, such as, equity, poverty, ecocentric approach and discounting, information
3
uncertainty and distributional effects for climate change decision making problem marked by
uncertainty.
,
2.1.3 Scope of the chapter:
Welfare
Impacts
This chapter raises the basic questions involved in climate change decision problem;
it emphasizes the collective nature of the decisions to be taken at the international level; it
focusses on the various issues related to climate change decision such as vulnerability of
Catastrophe
food production systems, poverty, cost of delay, free riding, valuing common property,
10
inequalities in responsibilities, North-South transfers, and ecocentric views; and it provides
"
the linkages between various components of the sequential climate change decision
DIFFERENT WELFARE IMPACTS FOR THE SAME
CLIMATIC DAMAGE : AN ILLUSTRATION
Disaster
1 8 A rich country has more resources to deal with climate change,
it can avoid even discomfort and be within adaptation zone
2 $ The same level of climate damage can bring economic disaster
We have assumed that adaptation to climate change imposes greater hardship on poor countries than rich countries for identical climate impacts.
12
problem. It also discusses various decision tools such as the utility based criteria, the nature
"
of uncertainty, risk and non-linearity, the role of time, insurance and portfolio approaches to
14
climate change.
13
Discomfort
Figure 2.2
To address the climate change decision problem, one should know what are the
16
questions. Section 2.1 raises all the relevant questions related to climate change decision
2
17
making and the various uncertainties related to climate change and their interrelationships
II
are also discussed. Section 2.2 gives an overview of the nature and genesis of the climate
19
change decision problem; this section includes topics such as concentrations, absorption
Adaptation
to a poor country
10
capacity, nature of risk and inequalities in responsibilities. Section 2.3 presents factors that
"
affect collective decision making. In section 2.3.1, how and why of collective decision
22
making are analyzed. Various considerations that go into collective decision making such as
23
vulnerabilities, ecocentric views, poverty, and intergenerational equity are covered in section
24
2.3.2. The implications arising out of collective decision making such as cost of delay,
Probability
of welfare
25
valuing common property, accepting adaptations and North-South transfers are dealt in
impacts
26
section 2.3.3. In section 2.4, under the heading of process of collective decision making,
"
we look at the steps involved in iterative decision making, criteria for inter-regional
28
allocation, and North-South co-operation through global policy instruments. The economic
20
notions of uncertainty, risk, bounded rationality, as well as various decision tools are
30
discussed in section 2.5. Finally, in section 2.6, we provide a sequential decision making
"
framework to hedge against impacts of uncertainty and change. This framework is made of
32
three components, namely, decision making paths, mutual insurance decisions, and
6
2.2 NATURE AND GENESIS OF THE CLIMATE CHANGE
PROBLEM LEADING TO THE FCCC
2.2.1 Concentrations not emissions:
The climate change problem is one of increasing concentrations (i.e. accumulation of
,
gases over decades) and not just emissions in a given year (stock and not flows) of GHGs. Of
course, to contain the increase in concentrations, one has to worry about the rate of emissions
,
as well. At present targets are being discussed in terms of emission levels, say reaching 1990
2050
level by 2000. However, consequences of each path to 1990 levels have to be worked out in
terms of concentrations (Fig.2.3a). While there is a discussion on reporting emissions every
10
year at the Intergovenmental Negotiating Committee (INC), the concentration levels as shown
"
12
"
Cumm. Emission Targets Cum. Concen. for Fig.2.3b
A
B
1990
Fig.2.3c Risk is
more by A than B
in Fig.2.3b, are yet to be discussed. Both shape and area under emission curve are important,
in Fig. 2.3b
since risks are measured in CO2 -years. Hence, climate change decision problem need to be
addressed in terms of concentrations where earlier reductions should be rewarded to account
14
for their more beneficial consequences and to encourage early actions. If we decide to wait for
"
EMISSIONS, CONCENTRATIONS AND CLIMATE CHANGE RISK
2020
more information without acting, the stock of concentration of the order of 750 billion tons
16
since pre-industrial times gets larger, passing a larger burden to future generations.
A
B
1990 2000
Fig. 2.3b Area under
A & B are the same over
some decades, but not risks
2000
1
Fig. 2.3a Same emission level
for the target year in each
curve, but not concentrations
In Fig.2.3a, the emission levels in the target year are the same for all paths, but
each path has obviously different implications in terms of cummulated emissions.
Fig.2.3b shows that the cummulative emission targets are the same for paths A and B
i.e., area under each path is the same. However, as shown in Fig.2.3c, the climate
change risk is more under path A than path B due to higher "C02-years" in the
atmosphere. Therefore, earlier reduction should be rewarded.
Figure 2.3
Emission Targets
2
3
1990
7
2.2.2 Absorption capacity:
adjustions
Therefore, small but regular reductions may be desirable than hoping to reduce suddenly and
3
The difference between cummulated emissions and atmospheric concentrations is
a lot later, which can be traumatic, if not impossible. It is better to initiate efforts to be on the
due to absorption capacity of the atmoshpere. Due to lack of understanding this is often
learning curve which is gradual. If new information suggests that scaling up is not required,
.
represented as a factor by which one reduces cummulated emissions.
one could relax. However, in the meantime due to learning experience, understanding of how
to deal with it develops. One could compare this with developing and testing fire fighting
,
Greenhouse gas emissions are absorbed by forests, oceans, agriculture and so on.
systems before the fire breaks, not after. This capability is to be developed as early as possible.
Thus, not all emissions end up as concentrations in the atmosphere. While some hold a
As pointed out by Chichilnisky (1994), climate risk is endogenous, collective and correlated.
view that absorption capacity, also known as sinks, may increase slightly with increased
Delay ia response is associated with two risks: (a) need to adapt to the changes and (b)
concentrations, others consider that this is eroding fast due to denuded forests and polluted
irreversible damages. Can irreversible damages, say, by extreme events, be compensated by
oceans. Assuming that about 4 billion tons of carbon dioxide is annually absorbed by the
10
insurance? The bigger issues such as increased costs and damages due to temperature rise, sea
10
sinks, one implication could be that if humanity lived a lifestyle corresponding to about
level rise, loss in biodiversity and other unforeseen damages are unlikely to be covered by the
II
0.66 tonne of carbon equivalent emissions per person, per year, there would be no rise in
12
insurance industry. Therefore, COP may need to exogenously decide on a risk minimization
12
atmospheric concentration of greenhouse gases from current levels. This figure is applicable
13
strategy in terms of concentrations of GHG, which the North could lead. For more deatils, refer
13
for the present population levels. According to this principle, those who emit less than this
14
chapter on equity.
14
amount may not be considered responsible for climate change effects. In other words,
15
responsibility of marginal damage begins after going above this limit. However, is this the
"
2.2.4 Common but differentiated responsibilities:
16
right way to allocate absorption capacity? This is an important issue. While absorption
16
Climate change problem is beset with most difficult problem of inequalities in
17
capacity of forests and agriculture could belong to the state in which they are, capacity of
17
responsibilities. Annex I continues to emit 70% of CO₂ emissions and has been responsible for
18
oceans is a common global property. There is a need to understand how absorption takes
18
85% of the cumulated emissions from the pre-industrial period and 77% from 1950 onwards.
19
place and how the sink capacity can be distributed.
19
Decisions to reduce would have been easier, if there was equality among nations. Many non-
20
2.2.3 Nature of risk:
20
annex I countries have inadequate living standards with insufficient nutrition, provision for safe
21
Climate change risk is a trend risk and not a sudden risk, or a random event say as in
21
drinking water and access to sanitation. Nearly, 20% of them do not have even light bulbs.
22
nuclear accidents, earth quakes or floods. That is, if there is a risk at time t, then, there will
22
It is not correct to say that everyone emits. In the next 5 decades despite expected rise in
23
be similar risk at time t+1.² However, correspondingly there would be a stream of benefits of
23
South's emissions, concentrations contributed by North would be larger than the concentrations
14
abatement, i.e., if there is benefit at time t, there will be benefit at t+1 also, and they are
24
contributed by South, even if North goes back to the 1990 levels and reduces further.
25
additive. This has two implications: once climate change impacts get in motion, it would be
25
Therefore, there would have to be simultaneous trends of the rich (viz., North) reducing and
difficult to halt them. Corollary of that is, that the problem has to be addressed in a systematic
26
the poor (viz., South) increasing their emissions, to give south the time needed to develop and
27
fashion by reducing the trends of growth, i.e., more like gradually decelerating a car, rather
27
fulfill basic needs. Here, the North has a stake in South's development. That is, to see that
than swerving to avoid an accident. Second, higher abatement costs are justified in case of
28
South pursues on pathSthat requires least greenhouse gas emissions for given development
trend risk, as it has stream of benefits for all years starting after a given year, in distant future.
goals. If, North's contribution to past and future concentrations are put to zero, South can go
That is, abatement avoids adaptation costs not in say, 2025 but for many years subsequently.
on with its development for another century without worrying about the climate problem.)This
1
That is, presently, there is 750 giga (billion) tons of carbon equivalent greenhouse gases in the atmosphere accummulated
I
1 The non-Annex I countries are not homogenously poor; hut. on average their living standards are much below that of
2
due to human activities. If there is risk at time t, then the risk at time [+] will be similar.
2
Annex I countries. We also note that there are poor in North and rich in South.
8
9
I
point is analyzed in detail in the section on North-South transfers.
2.3 COLLECTIVE DECISION MAKING
2
2
2.2.5 FCCC As A Key Element For Decision Making
2.3.1 How and Why collective decision making:
3
Some of the features described above are directly or indirectly incorporated in the
The Conference of Parties (COP) will be the apex forum which makes the collective
FCCC. Some may yet guide future negotiations. The following features of Framework
global environmental decisions for global welfare within the constraints set by equity and
,
Convention on Climate Change (FCCC) are relevant for the decision making framework.
efficiency. A conservative choice gives South the space it needs to grow and develop and
future generations more options to chose from. However, it can not be too conservative so as
1. Objectives: The ultimate objective is to achieve stabilization of GHG concentrations in the
to deprive the present generation. There is a hierarchy of decisions involved here.
,
atmosphere at a level that would prevent dangerous anthropogenic interference with the
climate system.
Climate change means different things to different people. To Annex I countries, it is
the cost of abatement, to non-Annex I countries it is the burden of adaptation, and concern for
2. Principles: The parties should protect the climate system in accordance with their common
vulnerability, irreversible damages and rise of frequency of extreme events and most of all
10
but differentiated responsibilities. The developed countries should take the lead in
11
costlier development. To island states, it is their existence and to oil exporters, it is their
11
controlling climate change. The specific needs and special circumstances of developing
12
revenue loss, if fossil-fuel demand drops. Again, it concerns differently economists and
12
countries should be given full consideration.
"
environmentalists. All these concerns need to be voiced in a comprehensive manner.
13
3. Commitment: Developed country parties should take lead to adopt national policies and
14
First stage as pointed out in section 2.1, involves decisions at global level, such as
14
measures by limiting their anthropogenic emissions of GHGs so as to modify longer term
14
choosing concentration level stabilization and emission growth paths and their interregional
14
trends. In order to promote progress to this end, each of these parties shall communicate
16
allocations. These decisions have to be taken collectively (by the COP). Utility based criteria
16
detailed information on measures taken with the aim of returning individually or jointly to
17
(see box 2.3, section 2.5.1) would be quite inadequate for this purpose. Global climate change
"
their 1990 levels of these anthropogenic emissions.
18
will affect each country differently. Their risks and welfare costs are different. At the same
"
4. Technology transfer: It also supports technology transfer and training to developing
19
time, those who have to take major abatement actions and therefore bear the costs of abatement
19
countries by developed countries.
20
are in general different countries from those who are at largest risk from climate change. We
21
20
do not have a global welfare function that integrates the utilities of different stakeholders. Since
5. Incremental costs: Global Environmental Facility (GEF) would be set up to provide "full
22
21
agreed incremental costs" to developing countries, when needed.
global climate change affects every country, each country is a stake holder. In fact, since
21
developing countries are more vulnerable and have contributed little to the pollution so far,
22
Keeping these FCCC guidelines in view, we discuss various issues related to climate
24
they could have unequal rights, and could have more say in this matter. In fact, even in
21
change decision making in this section.
29
levying income tax, certain limits on taxable income is recognized widely and also progressive
26
income tax. It is seldom that polluters decide what level of pollution could be tolerated
"
particularly when they are less vulnerable to the possible impacts of the pollution. Yet, their
2K
difficulties in responding quickly has to be kept in view. In such a situation one can use risk
24
based criteria, discussed later in this chapter. We discuss below the factors to be kept in view
while choosing the level of risk.
10
11
2.3.2 Factors for collective decision making:
2.3.2.2 Anthropocentric Vs Ecocentric approaches:
2
2.3.2.1 Vulnerability of food production systems:
So far, we have viewed climate change decision making purely from the point of homo
sapiens. At this juncture we want to emphasize the existence of an alternative viewpoint, viz.,
1
FCCC suggests that " stabilization of greenhouse gas concentrations in the
that of ecocentricism, which gives adequate, if not equal weightage, to all the living species
4
atmosphere at a level that would prevent dangerous athropogenic interference with the climate
in the world.
,
system. Such a level should be achieved within a time-frame sufficient to allow ecosystem to
adapt naturally to climate change, to ensure that food production is not threatened and to enable
Climate change may affect flora and fauna and their habitats, such as corral reefs and
,
economic development to proceed in a sustainable manner --". Parry & Rosenzweig's study
physical environments that are a part of life support system. The earlier IPCC impact
(1994) provides an idea of what is in store for future due to climate change in terms of food
assessment indicates substantial pressures on biosphere and biodiversity due to climate change.
vulnerability.
For example, bleaching of coral reefs is expected. Due to the movement of temperature zones,
without the rest of the life support system, species loss on land and sea is indicated. As we
10
They provide global assessment of the potential effects of climate change on crop
"
mentioned earlier, such a loss of species or damage to biodiversity has been traditionally
"
yields, world food supply, and regions vulnerable to food deficits. The national yield changes
12
ignored in the decision making framework. Only recently researchers have started developing
12
for the climate change scenarios were used as inputs into a world food trade model, the Basic
"
tools such as Environmental Impact Assessment to study the value of biodiversity for the
13
Linked System (BLS), developed at IIASA by Parikh K. et al (1988) and Fisher et.al. (1990).
humans. Such studies are mostly anthropocentric valuations, carried out to find impact of the
14
Simulation outputs from the BLS provided estimates. of changes in cereal production, food
15
biodiversity loss on human recreation, medicines, education, research, chemicals, bio-
15
prices and the number of people at risk of hunger.
16
engineering etc.
16
The study concludes that the world cereal production is estimated to decrease between
"
An ecocentric view, on the other hand, views homo sapiens as one of the species on
17
1 and 7%, depending on the climate change scenarios used, by 2050. The study also indicates
the planet earth expected not just to live in harmony and balance with other species but also
ID
that the impacts would be more severe on developing countries, in comparison with developed
14
to share the biosphere (which includes oceans, rivers, atmosphere, etc.) which provide
"
countries. With farm level adaptations, reduction in global cereal production is expected to be
sustenance to other species too. To put it sharply, humankind's abrogation of decision making
between 0 and 5%. However, the largest negative changes would occur in developing countries.
21
powers regarding nature is questioned. The ecocentric view proposes to curtail such decision
21
While, in the North (including China) there would be an increase in cereal production to the
22
making power. This view extends the ethical stand to include other animals and plants; Equity
22
tune of 2 to 14% after taking into account the direct physiological effects of CO₂ and the farm
"
is not only among human beings but also between humans and other living species.
23
level adaptations, there would be a decrease in developing countries to the tune of 6 to 12%
24
24
even after taking the fertilization effects of carbon dioxide and the farm level adaptations. This
Though ecocentric viewpoint may seem extreme, it raises one crucial point. The link
25
"
loss of production in developing countries, together with rising agriculture prices, according
between the existence of other species and that of the human beings should be kept in mind
to the study, is likely to increase the number of people at risk of hunger, in the order of 5-50%,
while making any decisions regarding the activities of humans which would adversely affect
"
depending on the GCM scenario.
the environment and inter alia other species.
Thus, CO₂ concentration levels of double the value of preindustrial values in 2060 may
29
not be acceptable to the South. If 550 ppm is reached in 2060 AD, it is likely that given the
long lead times required, stabilization reached by say, 2100 AD, would be much higher and
31
the outcome may be severe.
12
13
2.3.2.3 Poverty and Climate Change:
concern. Such a concern automatically raises the issue of equitable distribution which we
2
2
Another important issue while making climate change decisions is that of poverty and
discuss later.
the linkages between poverty and climate change impacts. Suppose there is increased
2.3.2.4 Intergenerational equity and discounting:
occurrence of extreme events such as floods, cyclones, typhones, hurricanes and so on occur
,
due to climate change. Then, in such a situation, it is the poor who are always vulnerable
Climate change problem will affect many future generations because of long life time
compared to the well off. Take the case of the recent earthquake in India; in this natural
of GHG. But, whenever cost benefit analysis is used to evaluate long term benefits and costs,
disaster, more than 10,000 people died compared to only a few in the earthquake of similar
as in the case of global climate change, the practice of discounting the future is severely
7
intensity in California.⁴ In essence, the threshold level to withstand these impacts would be
questioned. One set of critics base their criticism on the belief that discount rates place a low
less for developing countries when compared to the developed ones. In such a case, given that
weight on future generations. Another group argues that because discount rates dominate
the present accumulation of GHG in atmosphere is due to advanced countries, will they be
decisions relating to the future, one should use low discount rates whenever one evaluates
10
II
forthcoming to insure the poor against such calamities?
projects with long term implications. This is the position taken in Cline (1992) when he uses
II
a low discount rate ranging from 0 to 2%.
12
Large scale out-migration from coastal zones is expected due to sea level rise. Intrusion
12
13
of sea water in the ground water and changes in temperature can reduce agricultural incomes
Ethically, generational welfares are to be treated equally. But, future generations could
13
14
everywhere. This will create a large number of environmenal refugees. Will they get migration
be better or worse off than the present generation depending on the GDP growth rates. In the
14
15
rights (to the North)? Share of GDP from agriculture in total GDP ranges from 16% to 64%
former case, if the future benefits outweigh the present abatement costs, one would expect that
15
16
in low income developing countries and 12% to 37% in middle income developing countries
the future generations should pay for the abatement cost. Such an argument is based on the
16
"
compared to only 3% of USA (World development report, 1994). Also, the share of population
assumption that in the intergenerational welfare function, atmospheric capital (environment
17
"
depending on agriculture are of similar proportions in the south. In short, the impact of climate
space) is substitutable by man-nade capital. This substitutability assumption is highly
18
19
change will differ from region to region and sector to sector.
questionable. If there is no such substitutability between physical and atmospheric capital, one
19
cannot delay abatement actions in favour of setting aside some man-made capital for future
While the rich can argue about whether to enjoy more now or pay later in the future
20
generations. Chapter 3 on "Intergenerational equity and discounting" suggests that discounting
21
and if so what discount rate to use, this is unlikely to be a major issue for the poor who are
21
considerations change if the present generation is richer or poorer compared to the future
22
on the margin of subsistence. If a poor country is forced to reduce their GHG emissions or
22
generation. Most global modelling scenarios (chapter 11 on Integrated Assessment) show that
23
pushed to adapt to climate change without aid or efficient technology, they would have no
23
on the average, future generation of South will still be poorer compared to present generation
24
option but to reduce their current consumption below subsistence level. Rising concentrations
24
of North even by 2050 or beyond. Thus, global average of all regions can be the reference
25
of GHG in atmosphere will bring more constraints on their future development. To conclude,
24
point for such considerations. Present generation of North could choose discount rate as though
26
since the poor, whereever they are, have a very high marginal utility of income, if required to
26
global average would be less in future compared to their generation at present, while South can
27
curtail their GHG emissions in a comparable way to the rich, the poor will bear a
27
choose discount rate as though the next generation would be richer.
20
disproportionate burden of global warming caused mainly by the lifestyles and consumption
7
2a
patterns of the rich. Any reasonable decision making framework needs to address this primary
On the other hand, suppose that the future generations are worse off than the present
7
generation. In such a case, what the future would pay for the present abatement will be much
less than the cost of abatement. Some argue that intergenerational equity should be dealt with
4 Of course, two carthquakes are never quite comparable. The point, however, is in the Indian carthquake if only the houses
"
separately as a political decision with ethical dimensions. For a forceful argument of this point
2
were built with modern construction material, even low cost ones, the loss of life would not have been as much. The people
,
could not afford even such low cost modern housing.
14
15
I
we refer the readers to Schelling (1994), and chapter 3 of thus document.
after 1990 which would be retroactively adjusted against their future share, after negotiations
3
Not doing anything for the benefit of future generation could be acceptable. But, if the
are completed. This would prevent free riding and precipitate some action even while matters
,
present generation is responsible for the losses of the future, as is the case in the climate
are being negotiated. While issues such as past concentrations, population can be discussed
change problem, then not doing anything is morally unacceptable. That is, not giving them
later, accepting this principle of accountability after 1990 could lead to prudent behaviour and
large inheritance is acceptable, but we should not pass on liabilities. Abatement is not to be
act as some deterrent to all against wasteful consumption.
looked at as an investment option giving returns in the future but as a liability reduction
2.3.3.2 Valuing Common Property (and Paying for It):
exercise. Hence, the level of abatement of concentrations could be chosen to minimise climatic
,
The emissions emitted beyond earth's absorption capacity then test the biosphere's
impacts on poor and then look for the cost effective ways of fulfilling those commitments. Low
adaptation capacity. This capacity to adapt may be also limited i.e., it is an exhaustible
levels of discounting could play a role in the later step, but, not in the former.
resource. How is the atmospheric resource different from many other exhaustible resources we
10
We have described some of the considerations that go into collective decision making.
10
use every day that include metals, and fossil fuels? First of all, these resources have property
"
Having collectively decided the concentration levels and interregional responsibilities,
II
rights and they are paid for, each time they are used. "Paying for what you use" is a sound
12
individual countries could formulate their strategies to follow respective emission paths. While
12
economic principle that should be enforced whenever atmosphere is used to prevent misuse.
13
deciding who needs to do what, one has to take into account various factors such as
D
If one does not have to pay for common property, it encourages appropriation of atmospheric
14
responsibility interregional equity, North-South transfers and various policy tools to achieve
14
capacity. The user charge should be either paid to others when the property also belongs to
15
the required emission reductions.
"
them or put in a common pool for future use. Second way it is different from other exhaustible
resources is that there are substitutes available or likely to be available for them. This is
16
2.3.3 Implications of collective decision making:
"
unlikely to be the case for atmosphere of planet earth.
17
2.3.3.1 Cost of delay and slippages (Free riding through delay):
II
How does one pay for using atmosphere and to whom? Collective decision is required
"
International negotiations are difficult and take long time. Even if the decisions are
to decide this. How does one value absorption capacity and adaptive capacity? How does one
19
taken at international level, it cannot be guaranteed that they will be followed up by people at
20
allocate property rights for both? One can think of several ways.
20
various social strata. Therefore, slippages may occur. However, in the mean time concentrations
21
(a)
Consider that those who wish to emit more than their share should pay rent.
21
mount, if no directions are given. A ortion of every tonne of GHG emitted occupies some
22
22
environmental space for about 200 years, which is not available to either future generations
(b)
Consider that this is a permanent loss and pay damage costs.
23
everywhere, or constrains the development needs of non-Annex I countries. For example,
"
(c)
If and when property rights are allocated, trade or lease them.
24
during the nineties, after signing the FCCC, the Annex I countries would emit about 40 to 50
25
billion tonnes of CO2. This could have been sufficient for South to grow and develop for more
26
than 25 years. In addition, high emissions by one party push the other parties either to adapt
"
or to forego their development.
20
Therefore, if no decisions can be reached soon, at least it should be agreed that Annex
I countries would be liable for whatever concentrations emitted after 1990. That is, whatever
the outcome of negotiations, it would hold parties accountable for the cumulated emissions
16
17
I
2.3.3.3 Accepting Adaptation:
I
come.
1
There is ever increasing reference to need for adaptation. Adaptation ranges from
2
2.3.3.4 North-South transfers:
1
switching to different crops, incurring increased cooling costs to migrating or giving up land
,
4
rights if they are in the coastal areas. Parry & Rosenzweig (1994) have indicated increased
Several justifications for the North-South transfers are given below. These are not
4
,
hunger in developing countries which is also one way they adapt. That is, due to delay in
necessarily mutually exclusive. That is, transfers for one may overlap with other. However,
,
these are different ways of looking at North-South transfers¹.
abatement, global risks are externalised for which there will be no compensation. In any case,
,
adaptation will be too all-pervasive and costs too difficult to isolate for compensation. Will the
1. Currently there are many countries whose use of environmental space far exceeds their
North pay for dykes in Bangladesh or give migration rights and compensate for shifting and
7
justifiable share of that space. J.Parikh (1993) has indicated that these privilages are worth
rehabilitation to people from Maldives?
US $70 billion per year. Also, North's use of environmental space has a temporal
10
characteristic; that is, these emissions will occupy the environmental space for hundreds of
U.S. estimates for costs to build walls along the vulnerable zones to sea level rise is
10
"
$130 billion (refer chapter 4). That may be a small share of U.S. GDP, but such measures,
years in the future (Smith, 1991). According to the "polluter pays" principle, those who
II
12
even scaling for their coast lines, for say, India, could require very large share of their GDP.
have occupied this environmental space should pay rent for it for their excess use of the
12
13
Who shall pay India for such a wall? Moreover, one still does not know what would be the
global environment to those whose use is below their entitlements. In addition, when the
13
14
latter need that space, that should be made available to them. (Rent for environmental
environmental impact assessment of such walls. They could also have adverse ecological and
14
19
economic impacts on coastal zones.
space).
15
16
It is said that there may be rise of mean temperature of 0.25 degree per decade. This
2. The consumption patterns of the rich countries have put constraints on the development
16
17
is sometimes multiplied to get 2.5°C for a century but there could be non-linearity and
alternatives of the poor countries. Due to the climate change possibility, the poor will be forced
"
18
temperature rise could be more with more GHG accumulations. Moreover, 2.5 degree
to spend more resources for adapting to the climate change problems; this will constrain the
"
19
resources available for their immediate development priorities. (Compensation for adaptation
centigrade mean temperature rise could mean 5.5 degree centigrade rise in the Northern
19
20
hemishpere due to extensive land areas compared to oceans. Indeed, if this happens, it would
burden imposed).
21
be difficult to live in many parts of Indian subcontinent and other populated countries in the
20
3. Those who have emitted more than their share of the absorption capacity of the atmosphere
22
mid-east and South Asia and large scale migration may take place. Thus, adaptation costs
21
(in per capita basis) and benefitted, are liable to others for the possible damages arising from
23
passed on to others should be internalised in one'sactions. It is important to note that marginal
22
global climate change. (Liability payments for damages and excess past concentrations).
24
damage of one extra degree of temperature is far more serious in the South than in the North,
"
4. The cost of delay in emission reduction (by the North) in terms of South's foregone
25
where they may even benefit from that extra degree. Thus, risk tolerance of say, South should
24
opportunities to development is substantial. This will impose many constraints on the way the
26
matter more than that of North.
25
South decides on policy options regarding issues such as how to generate power, how to use
27
Most models developed in the North do not effectively build in the perspectives of the
20
land, and what crops to grow and so on. Hence, North-South transfers of large amounts are
28
South in adapting to climate change. The resources to adapt often do not exist and the outcome
"
needed to compensate South for the development opportunities foregone. (Lost opportunities
20
is further human suffering, manifested in, for example, frequent occurrences of the homeless
28
due to concentrations after 1990).
dying from heat stress. South does not have the capacity to adapt except through increased
29
As against all these, Global Environment Facility (GEF) gets replenishment of US $ 2
31
human suffering and is poor enough to have problems of their own. Homeless die of heat stress
12
even today. Finally, adaptation does not solve the problem; it only postpones what more is to
I
'K. Smith (1991) has used a concept of natural debt to address these issues.
18
19
billion for three years to address four different global environmental concerns, viz., climate
2
change, biodiversity, international water, CFC reducing to serve Montreal protocal. One could
1
also think of North-South transfers through global policy instruments such as tradeable
4
emission permit regime, joint implementation and so on. We briefly discuss these issues in the
,
next section. The reader can also see the separate Chapter on policy instruments for more
details.
,
Having discussed considerations for and implications of collective decision making, we
discuss some theoretical considerations involved in making the national and international
decision making in the following section.
"
2.4 PROCESS OF COLLECTIVE DECISION MAKING
12
Climate Research and
Policy Research
Integrated Assessment
NATIONAL DM
(Utility based Criteria)
Response Options
Technological Options
Life Style etc.
Let us now turn to the process of climate change decision making. In this section, we will first
"
provide a summary of steps involved in iterative climate change decision making problem,
14
criteria for interregional allocation, global policy instruments and consider sequential decision
15
making procedures.
16
2.4.1 A summary of the steps in iterative decision making:
17
Necessary elements of collective decision making are linked and brought together with
18
national decision making either directly or preferably through global policy instruments as
DECISION MAKING FRAMEWORK
Figure 2.4
21
19
shown in figure 2.4.
20
The first step in long-term decision making process is to select concentration level
21
through collective decision making by considering equity of stakeholders, of those stakeholders
22
whose emissions so far have not contributed significantly to the current atmospheric
COLLECTIVE DM
CONCENTRATION LEVELS
(Risk based Criteria)
EMISSION PATHS
INTERREGIONAL ALLOCATIONS
GLOBAL POLICY INSTRUMENTS
Equity & Efficiency
(Utility based Criteria)
for interregional transfers
South) (North-South)
23
concentration levels. For stabilizing atmospheric concentration of the main greenhouse gas,
Tradeable Permits
-Joint Implementation
24
namely CO2, at any level between 350 to 750 ppmv calls for significant reductions in the
Carbon Tax
"
emission trends in the near future. Some alternatives are indicated in figure 2.5. If a very low
20
level is chosen, considerable sacrifice is expected from the present generation. If high level of
27
concentration is chosen, there would be danger of irreversible damages and high adaptative
28
costs. Remember that, if 550 ppm is chosen, it is difficult to go back to 440 ppm, but vice
29
versa is possible.
20
I
than north. Emissions per GDP are sometimes taken as indication of inefficiency. However,
2
high emissions per GDP can be attributed only partially to inefficiencies in South. High
3
emissions are mainly due to low level of development which requires filling the backlog for
4
infrastructures such as, schools, hospitals, power plants, due to unsatisfied development needs
,
such as health, nutrition, education, lighting, and so on where relatively high share of GDP
6
goes (World Devp. Rep., 1994). This is referred as purchasing power parity. This is
,
comparable to the fact that, growing children need more food per kg. of body weight than
2250
adults past fifty. Higher need for new infrastructure building explains high emission per GDP
and also calls for higher growth in future. Apart from investment in infrastructure, pricing and
$750
$650
5550
$450
S350
10
valuation of non-tradeables also matter in these calculations. This can be captured by
11
adustment factor for purchasing power parity.
2200
12
One could think of a time table in terms of limiting emission by Annex I and non-
"
Annex I countries through reduction of emission growth rates.
14
15
Illustration of Strategic Decisions to be taken in Sequence
16
Time Period
Growth rate of emissions (dC/dt)
Rate of change of growth rate
(d²C/dt²)
17
Annex I
Non-Annex I
Annex I
Non-Annex I
countries
countries
countries
countries
18
≥ 0
> 0
<0
> 0
tigure 2. 5
ANEXURE Figure a
Profil:s of Atmospheric CO, Concentration
Reading to Stabilisation at 350, 450, 550, 650 and 750 ppny
2150
Upto 1990
2100
19
1990-2000
= 0
> 0
<0
> 0
20
2001-2025
<0
> 0
<0
= 0
20
2026-2050
<0
≥ 0
<0
<0
12
Beyond 2050
< 0
<0
<0
<0
" C Carbon dioxide emissions
24
2050
25
This is only an illustration of the process where Annex I Countries take lead and are
16
followed by non-Annex I countries after some lag. It is a sobering thought that over such a
27 long period, distinction between Annex I and non-Annex 1 countries may get blurred and
28
countries have to be grouped in terms of per capita emissions below and above the world
Prescribed CO, (pprmv)
Double
Pre-Industrial
Pre-Industrial
29 average.
1990 2000
30
The discussions regarding future entitlements for GHG emission so far have excluded
800
700
600
500
400
300
200
31 the historical record of past emissions. Such an omission might be wrong for the following
32 reason. Increases in the concentration of carbon in the atmosphere have taken place over a
33 prolonged period of time. The contribution of North America and Western Europe to the
I $ 2 &
34 concentration is much greater than their share of current emissions. On the other hand, the
22
historical emissions from developing countries have been low and hence their share of the
2
overall concentration is also low. The current generation of developed countries are the
,
beneficiaries of resource transfers from past generations. These resource transfers have been
possible only due to the exploitation of global environmental resources by their past
,
anthropogen CO,
erissions in 1990
generations. As a result, the current generation living in the developing countries have the right
to claim a part of these resource transfers, or, transfer of emission free technologies, or, right
7
to pollute (for development needs), as environmental resources are common property.
S750
S650
S550
S450
2050
2.4.3 North-South Cooperation through Global Policy Instruments:
S350
Once interregional obligations are decided, one could discuss global policy instruments,
10
which make it possible to reduce the cost burden by either trading emissions or by meeting the
2040
"
obligations jointly or by contributing to carbon tax, which can be used to reduce emissions in-
13
Comparison of IPCC Emissions Scentrins and
Emissions Profites Consistent with Stabilization of
Atmospheric Concentrations of 350 to 750 ppmv
IS92c
non-Annex I countries. Global Policy is also needed to discuss ways of North-South transfer,
13
either through the above mentioned or other means. These are referred as global policy
14
IS92a
2030
instruments. They can be selected based on utility (global or national) based criteria.
15
Figure: 2. 2.6
Institutional failures, such as lack of well defined property rights, and distortionary subsidies,
16
always lead to market failures such as gaps between private and social costs of production and
17
ANEXURE :Figate
2020
consumption activities. As a result of this, producers and consumers do not appreciate the true
"
scarcity value of the resources they use up, such as environmental resources. The emerging
IS92e
10
pattern of ecnomic growth from such production and consumption plans is usually
20
unsustainable. Economic instruments which are aimed to correct such unsustainable resource
Anthropogenic emissions (G:C/yr)
2010
21
use, bridge the gap between social costs and private costs by internalizing all the external costs
22
and risks. The use of economic instruments at the global level to deal with climate change
23
requires special considerations. We emphasize here these, as they relate to some suggested
2000
24
policy instruments. For a more detailed discussion on the policy instruments, refer chapter 10
25
of this document.
24
To elaborate, economic instruments such as property rights, pollution taxes, and
13
12
11
10
6
8
7
6
5
4
3
2
1
0
1990
"
tradeable emission permits bring about a realignment of resource allocation with the social
28
objectives. Such economic instruments achieve the objective of sustainable growth through (1)
%
providing a corrective to the functioning of the market, (2) encouraging cost minimization, (3)
30
permitting flexibility of use, and (4) adjustability to changing environment. The key to the
,
"
promise of economic instruments is their ability to harness the power of the market. Economic
32
instruments can be broadly classified into the following categories: (1) allocation of property
24
63c alean 2260-692-91r :# MAION THE 111 TREATMENT 8 Spiff NCW F6.-92-835
I
rights (ownership rights, use rights and development rights), (2) creating markets (tradeable
emissions. A global level of tolerable emissions is set for greenhouse gases and allocated
2
emission permits, tradeable offsets), (3) economic incentives (emission taxes, subsidies), (4)
among the nations by some criteria. Tradeable permits scheme on the other hand, give some
)
financial instruments (such as GEF, green fund, location incentives), and (5) liability
developing countries the possibility to earn additional foreign exchange for their development
instruments (liability standards, damage liability, liability insurance).
till such time that they have surplus permits. Sooner this is done, earlier the benefits. It will
also indirectly encourage them to reduce GHG emissions early, unlike JI, which induces them
,
Joint implementation (JI), Tradeable permits and carbon tax are some of the global
to reduce directly but sporadically. However, on what basis emission permits should be
policy instruments that lead to economic efficiency but require international negotations.
distributed is a thorny issue about which much is written. Though the desirability of
,
Among these options, JI is already referred in the FCCC and is in the advanced stage of
distributing permits (quotas) on per capita basis is recognised at least in the long run (Manne
negotiations. Pilot phase of JI is approved by the INC. Л gives opportunity for co-operation
and Richels, 1993), population growth in developing countries is an issue of concern. A
between North and South where North can invest in South to reduce GHG emissions, if it is
10
compromise could be to have the quotas on the basis of the population fixed in the year in
10
cheaper to do so. South gains in terms of energy efficiency if better technologies are provided.
"
which the agreement is reached. This will encourage North to reach agreement sooner and at
II
Key question is whether international credit or "offsets" should be given for such efforts. That
12
the same time South does not profit from population growth.
12
is, should GHG emissions reduced through JI accrue to the investing country, giving them
"
opportunity to increase their emissions to that extent? If yes, then further negotiations are
11
In addition to these options, other forms of co-operation also exist. This could be
14
needed on how much emission reductions through JI can be credited, mode of monitoring and
14
regional approaches, technology transfer, technology fund or joint technology development.
15
verification.
15
Apart from the North-South co-operation, regional approaches can also be developed keeping
16
in view the common features shared by the regions which may include socio-economic,
16
There are concerns that perhaps JI will delay new technology development or
17
ecological or meteorological features, viz., temperature, precipitation, solar radiation and so on.
17
consumption patterns of North will not change. Accountability of global GHG emissions when
IS
South does not have limits on the emissions is another concern. Therefore, there is no
IS
Once the goals and policy instruments are agreed upon, national level DM begins. This
10
agreement on the credit issue so far. Even if this is undecided, credits or incentives for the JI
19
strategy will depend on utility based criteria or regulatory mechanisms. Their responses will
in the form of carbon tax rebate can be given internally by North. If carbon tax is not imposed
vary from fuel substitutes, life style changes, technological alternatives, or increasing sinks.
21
by that country government, then it could permit rebate under some other category such as that
21
22
of treating JI as a category of R&D expense or charity or other tax deductible expenses. Thus,
23
internal incentives can be given to private entrepreneurs by the Governments in the North.
Why should the governments in North do so without international credit? JI can have an
25
immense value in terms of capacity building in South, so that when they will have to reduce
emissions, they will be ready for it. They may choose to go for energy efficient mode of
27
development. J. Parikh (1994) shows that if South makes some effort to reduce emissions
28
now, it will reduce pressure on North in future as well. Thus, it could have long term benefits
in terms of preparedness of South and reduced global emissions. JI will require monitoring of
the projects by certified auditors to account for carbon reduction, especially if international
"
credits are given.
"
Tradeable emission quotas (or permits) is one way of creating a market for GHG
25
26
1
2.5
Economic Formulation of Individual and Collective Decision Making
I
fecting decision making: among them one can cite attitude, poverty and intertemporal equity.
2
"Prior judgements that decision makers bring to the problem,
3
and the decision rules they employ, could be far more important
2
2.5.1
Accommodating the supplementary uncertainty caused by climate change
4
in controlling the policy conclusions that they reach
5
than are the results of scientific discoveries."
3 Cognitive and computational limitations make sense with respect to climate change concerns. Deci-
6
Dowlatabadi and Morgan, 1992
7 This section concerns methodological insights and guidance for making choice between a limited
4 sion making is concerned with these limitations.
3
Indeed, ignoring climate change would simplify decision making. However, it would be irrational
8 range of actions that have attached consequences, and especially establishing a point of convergence
9 between individual and collective decision making. How we think of choice involves issues of ratio-
6
to return to our previous lower level of uncertainty awareness. We thus have to accommodate the
10 nality and related decision criteria and rules, and is influenced by various factors among which uncer-
7 supplementary uncertainty caused by possible changes in climate (Caselton and Luo, 1994) and their
11 tainty comes first.¹ Whereas a key element of the decision making framework to address climate
8 effects on individual and social welfare.
12 change is the ratified FCCC (which indicates the objective to achieve, namely, the stabilisation of
9
As a consequence of insufficient knowledge and data, predictions of the rate and extent and timing
13 GHGs concentration), providing a steady economic basis to decision making paths and actions to
10
of climate change and their effects have been made in terms of scenarios rather than in usual proba-
14 bring about probably constitutes the most important contribution to the negotiating process.
11 bilistic terms. In particular, this situation reflects a gap between the information that is available on
15
The long term perspective and the globality of climate changes designate uncertainty as a key fac-
12 climatic processes and the one that would be relevant to Bayesian inference (the common way to use
16 tor in decision making. Uncertainty is related both to the science and to the economics of climate
13
new information to update beliefs "rationally"), and that is in short supply at present time.
17 change. On the one hand, there is a genuine uncertainty of the climate system-which has bounds but
14
What is the issue? In decision theory, decisions are based on the assessment of alternative strate-
18 which is not quantifiable (Bolin, 1994). For economics, on the other hand, estimates of the costs and
15
gies and the valuation of likely outcomes. The decision process describes uncertainties by means of
19 benefits of global warming are much uncertain. In particular, the uncertainty of impacts on individual
16 the concept of "state of nature" (a description of possible "states of the world"-that may be uncer-
20 and social welfare seems predominant.
17 tain, as well2). Common decision making is served by the Bayesian approach which represents any
21
For example, models in science, in economics and in social sciences provide quantitative estimates
18 information or knowledge concerning states of nature by assigning them probabilities. In this respect,
22 of effects based on critically important assumptions that openly leave large band of uncertainty in ref-
19 the representation of information in the form of conventional, necessarily precise, probability distri-
23 erence to climate change. Lack of knowledge about fundamental mechanisms that equations cannot
20 butions has been systematically developed in the statistical and economic literature since the 1930s.³
24 currently master represents limits to computer models in each domain. It is also true that most models
2 To formalise the conditions of uncertainty, it is convenient to refer to Arrow's concept of state of the world (SOW), a de-
acription of the world so complete that, if true and known, the consequences of every action would be known. The mean-
25 in the North do not effectively build in the perspective of the South in slowing down or adapting to
ing of certainty is that the agent knows the state of the world while that of uncertainty is that the agent does not know the
state of the world. Uncertainty characterises all activities; "it affects all fundamental variables that determine behaviour,
26 climate change-both in terms of resources and outcomes.
explain choices and bring about decisions" (Kessler, in Edckhoudt and Gollier, 1995).
3 In decision making under uncertainty and risk, the analyst is provided with precise, either objective or subjective proba-
27
A priori, however, the existence of uncertainty does not mean inaction. Making decision under un-
bilities. When it is not possible to arrive at such a unique probability distribution for an event. one tries to construct it
from more fundamental events. Such a procedure is known as the Bayes' rule. Suppose the hypothesis B is true and that an
28 certainty rather implies to filter paths of action and to choose which strategy with which outcome
event A has occurred. Then, the probability of hypothesis B under the condition that event A has taken place is given by
P(BIA) = p(AB)/p(A). (2(BIA) is known as the posterior probability and depends upon p(A) which is known as the a
29 would be best. This can be done at various analytical levels such as economic rationality, utility-
priori probability. How do we obtain the prior probabilities? If there are past empirical evidences one can obtain the priors
from them. But if there is no evidence available to us about, say, whether global warming or climate change of unaccept-
30 based choice criteria and rules, collective decision procedure (the one FCCC brings forth), short run -
ablc level is going to occur or not, then, according to the "principle of insufficient reason", one has to assign equal prior
probabilities 10 all the mutually exclusive events (or to forget about the Bayes' rule if climate change simply does not al-
31 long run trade-offs, risk analysis, value of improved information, as well as other particular factors af-
low to make prior probability assessments-Heap et al., 1992). In the first case, if new scientific information is forthcom-
ing, one can use that information to update the prior probabilities or beliefs using the Bayes' rule (or shifts and alternatives
1 Decision analysis also includes a range of techniques such as cost-benefit analysis or multi-attribute utility analysis.
to the Bayesian scheme-see below) to get the posteriors. In summary, the Bayesian theory yields instructions on how to
use new information to update beliefs "rationally" in certain circumstances. In essence, the Bayesian decision theory is
Each tool has a methodological foundation in decision theory but is suitable for different purposes. See 2.4.3.2, below.
subjective expected utility theory.
Chapter 2/page 27
Chapter 2/page 28
I
In view of the climate change concerns, and lack of knowledge and data, we undoubtedly have to
1
probability can always be formulated "as a bet on a certain event". Under these circumstances, the
2
address decision situations that go beyond conventional levels of uncertainty. How to treat informa-
2
condition for making strong uncertainty computable is that agents at least believe that they know,
3
tion sources (these are diverse with respect to climate change, each of which providing only a small
3
even approximately, the relative frequencies of the states of the world.
4
component of the total information-IPCC reports provide the most comprehensive assessment in
5
this field), and how to allow probability assignments to which subsets becomes significant.
4
2.5.1.2
Abrupt climate change adds unpredictability to uncertainty and risk
5
Nevertheless, some controversy exists with respect to probabilities and possible climate changes.
6
2.5.1.1
Strong uncertainty
6
On the one hand, the climate change context makes associated risks endogenous (they are affected by
7
This means that decision making to address climate change appears as a category of decision making
7
our actions) and collective (they are not statistically independent risks) (Chichilnisky and Heal,
8
which must envisage shifts to the conventional Bayesian scheme (Caselton and Luo, 1994) when
8
1993). For example, circumstances have changed for the insurance industry "in ways that mean
9
confronting the "strong" uncertainties created by climate change.
9
traditional practice and the old statistical methodologies are no longer adequate"-"past experience is
10
"Strong" uncertainty refers to situations where chance is included (which by turn designates
10
no longer a safe guide to the far future" (Ayres and Weaver, 1994). On the other hand, strong
11
common uncertainty and risk). but knowledge is insufficient-in other words, situations for which
11
uncertainty persists with regard to the characteristics of the evolution (rate, abruptness, etc.) of
12 probabilities cannot conventionally be defined.
12 possible climate changes. In sum, climate changes seem to imply a certain degree of irreducible
13
In short, climate change gives rise to a type of uncertainty that might not be straightforwardly re-
13
unpredictability beyond the common categories of risk and uncertainty.
14
duced to traditional risk. Phrased in economic terms, technological and economic "common" uncer-
14
On the theoretical side, for example, the important processes influencing the global climate are
15
tainties (or, "variability"), while including chance, include the prospect of being pleasantly surprised,
15
identified and can be understood at least in principle. Particularly, gases influencing the greenhouse
16
whereas climatic and biological uncertainties, while including chance, lie beyond our current cogni-
16
effect are analysed in detail (absorption bands, infrared emissions, diffusion) and so, its anthro-
17
tive and computational capabilities-and (in as far as they are seen as catastrophes) normally imply a
17
pogenic modification can be rather exactly calculated. However, the basic assumptions made in such
18
downward trend of expected revenue (Price, 1993).4
18
calculations are that the structure of the climate system remains unchanged. Especially, atmospheric
19
A more sensible distinction would be between situations in which there is some experimental or
19
and oceanic circulation patterns and cloud formation, in many models, are assumed to remain quali-
20
observational knowledge on the whole system or parts of the system (i.e., conceivable outcomes
20
tatively the same, at most being quantitatively slightly affected and so contributing to a positive or
21
would be known, but their probability of occurrence would not), and when there is no such knowl-
21
negative feedback, amplifying or damping anthropogenic climate effects respectively. Easy to under-
22
edge. In the first case, strong uncertainty can generally be dealt with by assessing "subjective" proba-
22 stand, that the prognosticated anthropogenic warming is gradual without discontinuities (+0.3 K per
23 bilities to the possible states of the world. In the second case, the situation leads to ignorance.⁵ From a
23
decade).
24
probabilistic standpoint, one can however argue that the distinction between common uncertainty and
24
In contrast to this, the picture drawn by the newest ice core results (Anklin et al., 1993) empha-
25
strong uncertainty (and so between risk and uncertainty as well) is a gradual one, as "objective"
25
sises variability and instability in a surprisingly unexpected way. Not only does the last ice age show
26
probability assessment or statistical inference does not exist.6 From such a standpoint, a subjective
26
a large number of sharp transitions between a glacial and an interstadial climatic regime (what has
4 However, this is not the rule for climate change in as far as it might also represent an opportunity, e.g., for innovation,
27 been known before) but also reveals the last interglacial age ("Eem", 125'000 years ago) as being
institutional and technological progress; i.e., it can be used positively whether true or not.
5 Ignorance basically designates a situation where there is absolutely no information about the unknown state of nature.
In the Bayesian theory, ignorance is meant as an informationless prior. Ignorance, and similarly near ignorance, can also
known/knowable (risk) or not (uncertainty). In this respect, a risk-like uncertainty (knowable probabilities) is sometimes
be viewed from the perspectives of belief and plausibility.
referred to as "weak" uncertainty. There exist more schemes of distinction between risk and uncertainty though
6
differences can rest on science-philosophical grounds (see Vercelli, 1991, Faucheux and Froger, 1994). An alternative is
The distinction IS frequently made between risk and uncertainty on the basis of whether the probabilities are
fuzzy logic. The one application of such logic to climate issues is contained in a paper by Leimbach (1993).
Chapter 2/page 29
Chapter 2/page 30
1
composed of a series of sharp climatic transitions. The largest uninterrupted interval of Eem climate
1
next for what change. We would thus face a "globally weak but regionally strong uncertainty".
2
optimum (roughly 2K warmer than our Holocene level) encompasses only ca. 2500 years and around
2
These perspectives add to the complexity of making decision under uncertainty with respect to
3
20 important abrupt transitions between different climate levels can be observed. The most catas-
3
climate change. Climate change may be a sudden event. Yet, this does not preclude quantification.
4
trophical fluctuation yet analysed happened towards the end of the Eem, where the climate changed
4
In sum, one can argue that although there is some consensus on the ability to deal with uncertainty
5
into a glacial regime within about 15 years, to make the same abrupt change back to the original
5
and risk through subjective probabilities, some uncertainties and risks associated to climate changes
6
warm climate only 70 years later (Anklin et al., 1993)!
6
may stay outside the reach of traditional practice. Although this is not the right place to expand
7
The discrepancies between the paleoclimatic observations of abrupt changes and the smooth grad-
7
"classical" and "abrupt change" scenarios, and although probability distributions should not be at-
8
ual warming prognosticated by today's models could not be more pronounced. They are the result of
8
tached to such alternative scenarios, subjective probabilities reflecting low probability but great
9
the fact that the models simulate climate changes as marginal variations around a stable basic climate
9
severity are needed (see 2.5.3.3, below). Alternatively, one must be also aware of low probability
10
mode. That the dynamics of this basic mode could dramatically change, being replaced by a different
10
outcomes (e.g., in the case GHG emissions would in future be captured). This means that shifts to the
11
one, is only beginning to be modelled (eg., Manabe and Stouffer, 1993).
11
Bayesian decision theory must be taken into account. To conclude, Baysian decision theory has two
12
In the framework of the basic properties of non-linear systems with circular causality, small
12
limitations with respect to climate change: first the existence of strong uncertainties and, second, the
13
stochastic fluctuations can become important factors influencing the time evolution of a system. As
13
existence of low probabilities attached to large potential damages. Indeed, the search for variations as
14
soon as the system approaches a critical value of a control parameter, an unpredictable fluctuation
14
well as alternatives to the Bayesian scheme for reasoning under strong uncertainty has intensified
15
might be the cause triggering an abrupt phase transition into one of several possible nearby states.
15
over the past decades to make reliable inference available from unreliable information, vague, partial
16
Even if all the critical values would be exactly known, neither the transition time nor the follow-up
16
and contradictory data (see 2.5.3.2 and 2.5.3.3, below). This stresses that, essentially, uncertainties
17
state could be predicted due to the "strong" uncertainty coming from stochastic fluctuations and weak
17
lead to judgement, and that it is the task of political processes to make judgements.
18
causality. Indeed, the main characteristics of the global climate system is its structure consisting of
19
many interlinked feedback-loops, some of which being highly non-linear. Under these circumstances,
18
2.5.2
"Rationalities"
20
the most natural behaviour to be expected is not a gradual warming but rather an abrupt phase change,
19
What is choice? On which criteria do individuals make choices in an economic context?
21
a transition to another climate regime.
20
Alternatively, do choices with respect to environmental problems convincingly differ from choices in
22
It turns out that in addition to the "classical" scenario characterised by gradual warming. two
21
economic situations-which constitute the reference framework in which economic choices apply?
23
"abrupt change" scenarios can be considered for illustration purposes:
22
The general picture of the theory of choice (equivalently, decision making theory) is as follows.
24
The "abrupt global change" scenario, with a sudden important temperature increase within a
23
The individual can take a limited range of actions that have attached consequences. He or she "has in
25
short period (within a decade), which seems to be a natural candidate based on theoretical reasons as
24
mind an ordering of all possible consequences of actions". Among the actions actually at hand, "that
26
well as new scientific results. The time point where the abrupt change will take place as well as the
25
action is chosen whose consequences are preferred to those of any other available action" (Arrow,
27
26
amplitude of the step and the regional responses will be unknown, so implying "strong uncertainty".
1951). Such situations constitute the reference framework in which economic choices apply.
27
28
The "abrupt regional change" scenario with sudden important climate changes in certain re-
Central conditions are, for example, a theory of market interactions and clearing as well as any co-
29
28
gions is a possible combination of the above and "classical" scenarios. While the global average tem-
herent theory of reactions to the economic stimuli (prices, in essence). Discounting, scarcity of the re-
30
29
perature increase would be rather continuous and reliable, it would be unknown which region will be
sources, and opportunity costs matter. Assumptions also compose the common economic framework.
Chapter 2/page 31
Chapter 2/page 32
1
For instance, a decision-maker's beliefs can always be represented by a unique probability measure;
1
in terms of relative desirability" (Arrow, 1951, Blaug, 1992). It designates a "maximization of the ex-
2
economic agents are said to be perfectly competent to deal with the difficulty of the processes in-
2
pected value of a given utility function (Savage, 1954; Arrow, 1987).
3
volved in the decisional situation. Basically, economic situations (and theory) are based on the
3
Subjective arguments are compatible components of the assumption of a consistent utility-max-
4
consistency of human choice.
4
imising behaviour. Expected utility (together with the notion of expected value, the return) is the
5
Economic criteria of choice primarily rest upon representations and orderings of consequences of
5
prime economic criterion for choice. It can take account for risk through various utility-based criteria.
6
actions by means of the assumption of a utility function. Such a utility function renders intelligible
6
Utility-based criteria mean addressing the maximisation of expected utility (or value-cf. Box
7
with regard to observable situations the ability of the consumers to rank their preferences in a
7
2.5-1). They entail decisions that are based on the valuation of outcomes-wealth levels, time streams
8
consistent order (in like manner, entrepreneurs will prefer more profit to less if each is equally risky).
8
of consumption, etc.-(Morgan and Henrion, 1990).
9
9 The assumption of a utility function, therefore, postulates a consistency of human choice-in short,
Box 2.5-1: Examples of some utility-based decision criteria
10
rationality. This postulate tells the decision-maker that the focus is on the decision to reach, not on
Source: After Morgan and Henrion, 1990: 26
11
the process per se. Verily, inasmuch as individuals are assumed to be perfectly knowledgeable and
Benefit-cost:
Estimate the benefits and costs of the alternatives in economic
terms and choose the one with the highest net benefit.
12
rational, the most valuable option-in terms of utility, or human wishes-is always chosen (Price,
Maximise multi-attribute utility:
This is the most general form of utility-based criterion. Rather
than use monetary value as the evaluation measure, MAU in-
13
1993).7
volves specifying a utility function that evaluates outcomes in
terms of all their Important attributes (Including uncertainties
14
Rationality and attached utility-based criteria of choice can still play the same pivotal role in
and risks). The alternative with maximum utility is selected.
15
climate change-oriented decision making processes as in economic situations.
Cost effectiveness:
Select a desired performance level, perhaps on non economic
grounds. Then choose the option that achieves the desired
16
However, decision-makers must initially be aware of the following differences. First, with respect
level at the lowest cost.
17
to the climate change perspective, choices are of collective and of longer-term view nature-they are
Minimise chance of worst possible outcome
Political and behavioural considerations frequently dictate the
Maximise chance of best possible outcome:
use of such criteria.
18
no longer individual and fixed choices (for instance, a country individually cannot take abatement
Set the maximum budget to devote to risk management
Maximum budget (bounded cost criterion):
19
measures in so far as emissions in the rest of the world will affect the country in question and the
activity.
20
gains from abatement by this country will accrue to the rest of the world). It turns out that a collective
Minimise maximum regrets:
Choose measures of "regret" and minimise the maximum
regret.
21 decision procedure, based on a "procedural" kind of rationality, must be set up (see 2.5.4, below).
Minimax Loss (Minimax critenon for catastrophic
For each decision alternative find the loss associated with the
risk):
extreme event. From among these losses find one with the
22 Second, choices are not made with respect to a fixed set of alternatives, either. Third, choices are not
maximum loss and try to minimise it. This is a conservative cri-
terion.
23
confined to market related decision situations, but are also concerned with non-market related
Maximin Gain:
Find the gains associated with each action and try to maximise
24 decision situations (including issues of poverty or fragility of natural ecosystems and species).
the minimum of these gains.
25
2.5.2.1
Utility-based decision criteria
10
For example, if the climate damage is sufficiently great, or the uncertainty of such effect suffi-
26
Economists do not assert that human agents are "rational" by definition. Rationality is a law-like
11
ciently strong, it will even pay a poor country to adapt or contribute to slowing down climate change,
27
refutable statement which "simply means behaviour in accordance with some ordering of alternatives
12
even at a low level of income. Poor countries have invested heavily in industrial and irrigation pro-
7 Simply. it is assumed that in evaluating uncertainty and risk situations, the decision-maker replaces the monetary value
13
jects, thereby sacrificing present consumption to seek higher incomes in the future. In some countries,
of final outcome by the utility (a function) of that one. Therefore, utility can be interpreted as a method "of passing from a
more objective level to a more subjective level: that of the amount of satisfaction achieved" (Eeckhoudt and Gollier,
14
such as Singapore, people have voluntarily engaged in high savings (45% of income approximately)
1995).
Chapter 2/page 33
Chapter 2/page 34
1
I
even at low present consumption levels to achieve a better life in the future.
climate change. In addition, lack of knowledge about fundamental mechanisms that equations cannot
2
2
The implicit argument is a redistributionist argument based on the higher marginal utility of
currently master represents limits to computer models in both domains. These difficulties typically
3
3
income at lower levels.
occur with respect to decisions that are taken on a national or international scale when governments
4
negotiate some form of agreement and countries differ in their economic wealth, reliance on fossil
4
2.5.2.2
Bounded rationality
5
fuels, endowmens of alternative resources and technical capabilities.
5
Rational behaviour is a "first approximation" to actual behaviour. Indeed, as soon as one faces either
6
Specifically, "procedural" rationality designates the rationality of a decision in terms of the man-
6
imperfect forms of competition or non-market resources allocation decisions, knowledge require-
7
ner in which it is made while economic rationality names the rationality of a decision considered in-
7
ments radically shift and exceed those usually required for a pure market framework. In addition,
8
dependently of the manner in which it is made (in the case for economic-equivalently, substantive-
8
there might be as many rationalities as there are cultures.
9
rationality, rationality evaluation refers exclusively to the results of the choice, not to the decision-
9
It is anyhow widely accepted that (economic) rationality hypotheses be supplemented by assump-
10
making process in itself-Faucheux and Froger, 1994).
10
tions of a different character (e.g., information-gathering ability and limitation of both knowledge and
11
Which kind of theory can better support and explain the decisions at issue under FCCC? Shall we
11
computational capacity). It is accepted that rationality, obviously, is bounded (Simon, 1957, Arrow,
12
place interest in the decisions to be reached only, or shall we focus attention on the (international)
12
1987). In principle, these cognitive difficulties are connected to decisions that lie beyond individual-
13
decision procedure as such? Answering to these questions and the like would command the choice
13
related, common economic situations; they relate to decisions that must be reached through interna-
14
between economic and procedural rationality frameworks (Simon, 1987). From an economic stand-
14
tional conventions, or to decisions that, independently of the valuation of the outcomes, are destined
15
point, unfortunately, the fact that the manner in which the decision is made-and not the result of the
15
to eliminate risks.
16
choice-is at issue in the decision evaluation makes the procedural kind of rationality mostly unde-
16
As a concept, "bounded rationality" has been coined to name rational choice that takes into ac-
17
termined. This tells decision-makers that procedural rationality should be bounded to the processes of
17
count the "cognitive limitations of the decision-maker" (Simon, 1987). Basically, this tells the deci-
18
collective decision making, i.e., the steps in negotiating treaties and agreements. We will return to this
18
sion-makers that rationality can be also concerned with the ways in which the actual decision-making
19
in subsection 2.5.4., below.
19
process influences the decisions that are reached and not only with reaching decisions assuming the
20
2.5.3
20
maximisation of expected utility, as we are used to consider from a common economic viewpoint.
Responses to uncertainty and risk
21
21
This concept thus designates a different type of rationality, namely procedural rationality.
Getting back to the distinction between uncertainty and risk made on the basis of whether the proba-
22
22
Incidentally, a decision which optimises co-operation will possibly be in opposition with a decision
bilities are not (uncertainty) or are known/knowable (risk), let's consider how (intuitively) to proceed.
23
23
which optimises the expected value of utility because the latter places the focus on the decision to
The classical ingredients of a decision problem under uncertainty are a set of feasible acts (A), a
24
24
reach while the former places the emphasis on the decision process in itself.
set of consequences (C), and a set of states of nature (S). The decision-maker does not know for sure
25
which consequence will result from a chosen act (probabilities for the events are not known in ad-
25
2523
Procedural rationality
26
vance). The role of the states of nature is to indicate the matching of feasible acts to consequences,
26
The former cognitive limitations may break usual limits on specific information. In the words of
27
thus resolving the uncertainty which relates acts to consequences. This makes states of nature exoge-
27
Simon, they are almost always limits on the adequacy of the scientific theories that can be used to
28
nous (uncertainty becomes in like manner exogenous, too) so that probabilities can be computed. As a
28
predict the relevant phenomena (Simon, 1987). Meteorology and atmospheric chemistry, on the one
29
result, the decision-makers basically have to assign utilities to the consequences (utility U assigned
29
hand, as well as economics, on the other hand, openly leave large bands of uncertainty in reference to
Chapter 2/page 36
Chapter 2/page 35
I
to each consequence c in C) and probabilities to the states (probability P assigned to each state 9
I
Minimax (Savage, 1951)
We should select measures of "regret" (i.e., the difference between an
2
actual pay-off and what the pay-off would have been if the correct strat-
2
in S) (Schmeidler and Wakker, 1987).8 In choosing between acts, the decision-makers will elect that
3
egy had been chosen). This is a cautious principle and ensures that, if the
3
one with the highest "expected utility" (or, expected value)-that is, the one with the highest sum of
4
worst happens, we make "the best of a bad job".
5
4
the products of probability and utility.
Emphasises the cost of making the wrong decision (Price, 1993).
6
We should seek to minimise the maximum regrets.
5
A decision making under risk introduces a special formulation of the decision problem to the ex-
7
We should seek to minimise maximum loss (Dowlatabadi and Mor-
6
tent that probabilities for the events are known in advance. In other words, the decision problem in-
8
gan, 1992).
7
cludes the probability distribution of an act over the consequences. Accordingly, decision-makers
9
Second, basic individual/collective preferences towards risk can be outlined as follows:
10
a
only have to determine utilities of consequences before choosing between acts.
11
Risk aversion:
Higher expected return is needed as compensation for an increase in risk. A
9
Yet, this presentation reflects individual election and addresses common economic or technologi-
12
risk averter can be a diversifier and spread his portfolio over different assets-
10
13
cal variability; it is not straightforwardly designed to address unpredictability and judge risks like
or a "plunger" and invest wholly in bonds and money.
11
those climatic and biological possible catastrophes might imply (see 2.5.3.3, below). As a first rough
14
Diversifier:
Risk is likely to be the cause of portfolio diversification: if risks increase then
12
15
approximation with respect to climate change, the expected utility criterion says that it is possible to
the diversifiers will move back to holding wealth in the form of money.
13
retain the notion of expected value (the return) while, with reference to uncertainty decision rules (see
16
Risk lover:
A risk lover accepts a higher level of risk for a given expected return.
14
below) and efficient risk sharing (see 2.5.2), also taking risk into account.
17
Risk neutral:
A risk neutral individual would be indifferent whether or not he or she ac-
18
cepted "fair" insurance. Under risk neutrality, the risk premium is zero.
15
2.5.3.1
Uncertainty and risk decision rules
16
19
Cautiousness:
One can move from the pure decision criteria that appear in Box 2.5-1 to more simple decision rules;
Defined as decreasing risk aversion (Pratt, 1995). The cautious person has no
20
dislike for uncertainty as such, but feels that he or she "can expect to do better
17
i.e., rules which do not need to elucidate all possible outcomes and their respective probabilities.
21
in the long run by playing it safe for the present and thus increasing probabil-
18
First, approved individual/collective decision rules with respect to uncertainty can be outlined as
22
ity that he/she will have the funds available to take advantage of more favor-
19
follows (MIT, 1986):
23
able deals which may arise in the future (Schlaifer, 1961; after Pratt, 1995).
24
Cautiousness is thus derived from risk aversion.
20
Maximax
The decision maker should opt for that course of action which yields the
21
highest pay-off (net benefits, or whatever) regardless of any view taken
25
2.5.3.2
Decision analysis techniques
22
about which state of nature will occur. It's the optimistic rule.
26
Application of utility-based criteria and rules involve a full range of techniques, each of them being
23
Maximin (Wald, 1950)
We should first observe all the minimum pay-offs and then select the
24
highest of these.
27
suitable for different purposes. Some are designed to serve as an aid to collective decision making.
25
Amounts to expecting the worst (Price, 1993)-presumably in our case,
28
One is a balancing of benefits and costs in a Benefit-cost analysis framework. Game theory is one
26
the failure of technology or substitution to relieve climate change con-
29
another. Techniques developed by management scientists over the last 10-20 years (which are some-
27
straints.
28
Decision-makers who employ this criterion are extreme pessimists.
30 times collectively referred to as "Soft System Analysis" and include Strategic choice analysis, Rob-
31
ustness analysis, Hypergame theory and a number or others) are other ones (cf. Ríos, 1994). One
8 A specific way allowing to cover the case for uncertainty-originally about the availability of resources and about
consumption and production possibilities-is the idea of contingent commodity introduced by Arrow (1953) and
32
another grouping of methods for reasoning under uncertainty is Evidence Theory, or defeasible
developed by Debreu (1953). This device is added to the current specifications of the commodity such as its physical
characteristics and location to make the availability of the commodity contingent on the occurrence of some (uncertain)
33
reasoning and uncertainty management techniques (Kohlas et al., 1994). This group concerns the
environmental event such as weather conditions.
Chapter 2/page 37
Chapter 2/page 38
I
study of symbolic and numerical methods for the representation and management of incomplete
I tainty, which in this matter might not be exactly an exceptional situation but maybe the rule. In this
2 information. Basic research activities cover domains like nonmonotonic logics, fuzzy logic and
2 context, one may choose two different ways.
3 possibility theory, "surprise" (Shackle, 1967), "imprecise" probability (Walley, 1991), Bayesian net-
3
First, as regards strong uncertainty, the approach based on the exploration of the future on the ba-
4
works, Dempster-Shafer theory of evidence (Dempster, 1967, Shafer, 1976; Caselton and Luo, 1994)
4
sis of the use of scenarios appears as one of the few relatively-useful instruments. This can address
5
and "Non-expected utility models" (Yaari, 1987). It attacks, notably, problems related to uncertain,
5 mitigation and adaptation to climate changes, the phasing out existing policies, as well as climate and
6 unreliable, vague, partial and contradictory data.9
6 technology research.
7
A different conception is the Multi-attribute utility analysis which is suitable for situations where
7
Second, in so far as climate change risks can be delineated, the risk assessment of potential cli-
8 options involve "collections of attributes that are incommensurate" (Keeney and Raiffa, 1976,
8
mate change can deliver probability distributions of the different estimates to policy makers (see
9 Morgan and Henrion, 1990). The resulting decision model can avoid the dilemma of converting each
9 Chapter 7, below). On the other hand, in so far as we accept to bet on future events according to a set
10 attribute to an equivalent monetary value.
10 of possible descriptions of collective risk, it could be natural to allow agents to trade securities
11
Variations to utility-based decision criteria are available. A Cost effectiveness criterion can be
11 insurance contracts (see 2.5.2, below).
12
used where the value of benefits cannot be estimated-but a choice must be made between alterna-
12
In sum, an open question is: are policy makers better served by best guess scenarios than by prob-
13 tives (Morgan and Henrion, 1990). A Maximum budget can be set to devote to risk management ac-
13 ability distributions of these estimates (Shlyakhter et al., 1995)?
14 tivity (bounded cost criterion). Last, measures of "regret" can be chosen and the maximum regrets
14
These options can be reconciled by drawing distinctions between scenarios that are probable (risk
15 minimised.
15 situations) and those that are possible (strong uncertainty situations). According to Shlyakhter et al.
16
Other methods or tools are thought to help diagnose the negotiation process itself, for example,
16 (1995), those risks that fall below a particular threshold of probability-and thereby are ignored by a
17 Cognitive mapping, Simulation modelling, Rule-based systems, an Information management (HASA,
17 particular group or society-are called "de minimis" risks. "How societies and governments decide
18 1993). Techniques belonging to decision analysis can be reoriented to serve as negotiations supports
18 what constitutes de minimis risk in particular situations or contexts is largely a matter of political
19 as well. One can cite Cross-impact analysis that facilitates the modelling of complex situations or
19 judgement" (Shlyakhter et al., 1995). Although there is no clear definition of a de minimis risk, it can
20 stakeholder analysis which uses information on the position, interest, priorities, and preferences of
20 generally be seen to be closely akin to a related concept, namely, the probability of "surprise".
21 various stakeholders on a particular issue to facilitate analysis of the range of differences among
21
The next question is: "At what probability of a serious effect should society take action?" How to
22 stakeholders and the potential for coalition formation (IIASA, 1993). Multi-attribute utility analysis
22 compute a de minimis risk-equivalently, a probability of surprise? Although many conceptions and
23 techniques have been applied by researchers at IIASA to evaluate coalition building and preference
23 definitions are possible, the use by Shiyakhter et al. (1995) of the word surprise is meant to denote
24 adjustment in the UN Conference on Environment and Development.
24 those situations where the true values of a particular uncertain parameter, e.g., climate sensitivity to
25 CO₂ doubling, appear at least 2.6 standard deviations away from its current "best guess" value. For a
25 2.5.3.3 Addressing strong uncertainty
26 random variable that is assumed to be normally distributed, the probability that the "true" value is
26 Under climate changes situations, decision-makers need to pay attention to situations of strong uncer-
27 more than 2.6 standard deviations from the current "best guess" is just 1 percent. As Shlyakhter et al.
9 An essential aspect of a decision analysis is the ability to combine information from various independent sources. For
28 (1995) continue, it is interesting to note that public opinion polls suggest that many people are uncon-
instance, in the Dempster-Shafer context, if two basic probability assignments (BPA) on the same unknown state of nature
are obtained from two different independent sources of information, they can be combined to yield a new resultant BPA.
29 cerned about a 5 percent chance of a climate-related catastrophe within their lifetime, although they
This kind of scheme also allows at specifying probabilities on intervals of the state of nature, leading to a possibilistic
specification of probability.
30 are concerned about a 1 percent chance of a nuclear accident. They also note that an airliner with a
Chapter 2/page 39
Chapter 2/page 40
1
calculated chance of failure around 5 percent in its 70 year life would not be allowed to fly in com-
1
decision making should read intergovernmental "negotiated" decision making. 10
2
mercial service.
2
2.5.4.1
Steps in negotiating international environmental treaties and agreements
3
Last, qualitative and probabilistic changes relate to the expected value of a single experience. As
3
Environmental accords tend to be partial and incremental. On the one hand, they tend to be inte-
4
such, they have "close affinities with the meaning of discounting" (Price, 1993; also cf. Chapter 3).
4
grated, complex packages of multiple linked issues that are difficult to negotiate internationally. On
5
2.5.4
Collective decision making
5
the other hand, they are made of several steps such as pre-negotiation, negotiation and post-nego-
6
To discuss the choices of a group rather than of an individual, one can use an index of social welfare
6
tiation processes. This is due largely to scientific uncertainty concerning the problems and possible
7
as a "utility" function. This fits into the framework of maximising the expected value of a function.
7
solutions, and the (a priori) need for continued learning about effects and consequences. Partial
8
The group then makes collective risky investment decisions and insurance decisions by using this
8
agreements are also customary because of the political uncertainty in devising fair and equitable ap-
9
function. Though collective decision making is central in economic analysis, many applications are
9
proaches at a global level (IIASA, 1993).
10
analysed in this way; that is, using microeconomic analogies and inferences.
10
Sustained negotiation or post-agreement negotiation are also an attribute of current international
11
Specifically, however, a group has possibilities of risk retention that are much greater than that of
11
environmental negotiations. This is the case, namely, for the UN FCCC, its former Intergovernmental
12
an individual. "This is partly due to the opportunities of diversification within the group and partly
12
Negotiating Committee (INC) and its new Conference of the Parties (CoP). Post-agreement negotiation
13 due to the transfer of risk towards the least risk-adverse members (or the richest members if absolute
13
can be defined as the dynamic and co-operative systems, procedures, and structures which are institu-
14
risk aversion is decreasing)" (Eeckhoudt and Gollier, 1995). This is why organisations have always
14
tionalised to sustain dialogue on issues that cannot, by their nature, be resolved by a single agreement.
15
been best suited to undertake risky activities-adapting to or slowing down climate changes is a risky
15
A framework convention or a protocol mechanism often have no clear end-point. A broad framework
16
activity by itself due to the existence of strong uncertainties. "Without this diminution of risk
16
of principles is agreed to first, followed by additional talks to specify details (IIASA, 1993).
17
aversion in an economy, thanks to the creation of risk pools, many risky projects would not have been
17
Procedural issues need to be addressed further. It results from a study by IIASA (Sjöstedt, 1993)
18
undertaken and we would undoubtedly not have known the economic expansion that we have
18
that the following ones need to be dealt with by practitioners and researchers alike:
19
observed over the last two centuries" (lbid.). In practice and with respect to climate change,
19
Many new groups and individuals beyond traditional governmental officials are now important
20 applications of collective decision making regard two out of three key rationales this chapter will
20
participants in the negotiation process.
21
21
select: insurance decisions (2.5.2) and a policy measures portfolio (2.5.3) (the first rationale is time-
The complexity of international environmental negotiations is increasing. Policy makers have
22
linked environmental issues with other important policy concerns, such as development and
22
oriented, i.e., sequential decision making paths-see 2.5.1).
23
trade. The implications of these linkages are often not well understood. Ways must be found to
23
Specifically again, 'collective' decision making means the choice of weights for the members of the
24
develop a common understanding of these linkages.
25
24
group. For this reason, the valuation of outcomes will become matter of negociation under a collec-
Bridging the differences between industrialised and developing countries will be a continuing
26
problem, given their different interpretations of what constitutes fair and equitable solutions.
25
tive decision making framework. Yet, negociation proceeds along a particular procedural scheme, to
27
New approaches to innovative problem solving need to be found and implemented, perhaps
26
which we turn now.
28
using the assistance of third parties or creative reasoning heuristics.
29
27
As a procedure, collective decision making concerns the processes of co-ordinated decision-
New strategies to implement negotiated environmental agreements are needed to reduce
30
national ratification delays and increase the likelihood of compliance (DASA, 1993).
28
making, i.e., the steps in negotiating international treaties and agreements, the appropriate weights,
10 The choice of the country as decision unit has formal justification in international law. However, this does not solve
29
decision criteria and/or rules if any, and the particular decision procedure. In this context, 'collective'
real world decision making in as far as major energy producers and users are directly involved in it. Therefore the
paramount importance of utility-based decision making criteria and rules.
Chapter 2/page 41
Chapter 2/page 42
I
Unlike the processes of pre-negotiation and negotiation, post-agreement negotiation has received
1
of assuming technological progress which might fail to materialise. =
2
little attention from researchers in terms of their procedural rationality. Options can be:
2
Convention-related principles are still one another possibility. Indeed, the need for an equitable
3
Involving domestic stakeholders in negotiations from the beginning.
3
distribution of tasks both at the national and international levels is the cornerstone of the Framework
4
Modifying the structure of the accords themselves: Clusters of simple, single-issue agreements
4
5
may be more practical.
Convention on Climate Change (FCCC). For example, while all the Parties are required to contribute
6
Initiating a new learning process in the post-agreement negotiation process in as far as
5
to the collective effort by elaborating a national programme, the application of the principle of
7
international problems can be more abstract than local environmental issues.
6
"common but differentiated responsibilities" implies that the industrialised nations should take
8
Post-agreement negotiations present policy makers with the challenge of progressively reframing
7
strong measures-first by adopting specific national measures to stabilise greenhouse gas emissions,
9
problems, adjusting strategies and perceptions, and refining solutions (HASA, 1993).
8
second by promoting the transfer of financial resources, know-how and/or technologies to developing
10
2.5.4.2
9
Collective decision criteria
countries, in particular the least developed countries (FCCC, 1992, BUWAL, 1994).
10
11
At the world level, decisions cannot be easily based on individual-like utility functions (index of
The precautionary principle also belong to the FCCC-related principles. Basically, it is a rights-
11
12
social welfare). In addition, once it is accepted that we have no world sovereignty and no global
based criterion. According to FCCC, however, it has been thought to encourage nations and individu-
12
13
social welfare function, the outcome is a matter of negotiation. This is the way to address the choice
als to take early actions to prevent environmental damage even when these actions are not
14
of weights for the members of the group.
13
"economically attractive". In this sense, it might look like some legal or ethical principle, and
15
14
How much each country has to say is, however, more a matter of what is than a matter of ethics.
consequently be in contradiction with cost-effectiveness demands (that also are requested by the
15
16
The literature on negotiation analysis (Fisher and Ury, 1981; Raiffa, 1982; and others) stresses that
framework convention on climate change!).
17
each party to an agreement must find himself better off with the agreement, than without. Otherwise
16
2.5.4.3
Decision procedure
18
he has no reason to agree.
17
The Convention establishes flexible procedures, based on the submission by governments of national
19
The Conference of the Parties (CoP) will be the apex forum which makes the collective global en-
18
communications, in order to ensure that States implement their commitments. These procedures are
20
vironmental decisions for global welfare within the constraints set by environmental risk and eco-
21
nomic efficiency.
11 Here are examples of some collective, rights-based criteria (Morgan and Henrion, 1990):
Zero risk:
22
One possibility is to use collective decision criteria (or, rights-based criteria).
Independent of the benefits and costs, and of how big the risks are, eliminate the
risks, or do not allow their introduction.
23
One another possibility is-in as far as every country has its predispositions in the debate, and
Bounded or constrained risk:
Independent of the costs and benefits, constrain the level of risk so that it does not
exceed a specific level or, more generally, so that it meets a set of specified criteria.
24
there can be no certainty about which position will be vindicated in the passing of time (Price,
Approval/compensation:
Allow risks to be imposed only on people who have voluntarily given consent, perhaps
after compensation.
25
1993)-to adopt traditional criteria for decision making under uncertainty.
Approved process:
Not strictly a decision criterion for analysis, but widely applied in risk management
26
The maximin criterion (Wald, 1950; see also 2.5.3.1, above) amounts to expecting the worst-
decision making. If all relevant parties observe a specified set of procedures (e.g.,
under some international convention). then any decision reached will by definition be
27
presumably the failure of energy and technology substitution to relieve climate change constraints-
acceptable.
Liability insurance:
Parties which impose risk might be eligible for liability insurance (added example).
28
and adopting a policy based on that expectation (Price, 1993).
Rights-based criteria have been thought to deal with processes and constrained actions or activities (Morgan and Henrion,
29
The minimax regret criterion (Savage, 1951; also see 2.5.3.1) emphasises the cost of making the
1990). For instance, the approved process criterion insists "that all the relevant parties observe a specified set of
procedures" such as some internationally signed convention defining a due process (this leads us back to procedural
30
wrong decision-presumably, in the climate change context, the very high cost of damages as a result
rationality). Some of these criteria are inconsistent to deal with exogenous. naturally occurring risks as for instance the
zero risk criterion which simply does not allow the introduction of the risk. Hybrids of utility- and rights-based criteria are
sometimes used.
Chapter 2/page 43
Chapter 2/page 44
1
entitled to an overproportional share of tax recycling. In this respect, however, liability would be
1
also designed to allow if necessary to amend the Convention in the light of new knowledge, for ex-
2
impossible for emissions prior to the discovery of the possibility of anthropogenic warming (ca.
2
ample through protocols dealing with specific areas (BUWAL, 1994).
3
1988). Instead, it should be recognised that more GHG efficient technologies are available to
3
In line with the provisions of the FCCC (Art. 4.2e), industrialised countries must also co-ordinate
4
developing countries than it was the case when developed countries were in development phase, and
4
economic and administrative instruments. International co-ordination also helps at avoiding undesir-
5
that the only path to development is not the one that has been travelled by developed countries
5
able side-effects such as trade distortions and unwanted impacts on the environment of third coun-
6
(otherwise, it must be acknowledged that developing countries will be the source of most pollution in
6
tries.
7
All countries should participate collectively in efforts aimed at solving global environmental
future years).
7
8
In this respect, joint implementation (JI) can be a valuable instrument though controversies-that
8
problems, based on an equitable sharing of responsibilities. In this respect, increased solidarity be-
tween the planet's rich and poor regions is essential. 12 Attention should also be paid to providing fi-
9
mainly are political and not economic-still surround the concept. Used in very flexible manner, it
9
nancial support to the Central and eastern European nations, in order to encourage their full participa-
10
would induce additional transfers to the South. Developing countries would not invest in "climate
10
11 change abatement" but might change their development strategy in ways that are compativle with
11
tion in the work undertaken within the FCCC.
12 lower GHG emissions. Left without efficient devices then the South is unlikely to invest in abatement
12
Under FCCC, indeed, the developing countries have no obligation to control emissions and there-
13
13
of climate change. This will have to come from the North. Second is cost-effectiveness. Any Northern
fore do not need to pay "global carbon taxes". However, if a CO₂ tax is imposed, efficiency requires
14
investment in abatement would be most cost-effective through joint implementation (Chapter 4; Pillet
14 that it be imposed on everyone. Energy efficiency is just as important in poor countries as in rich
15 countries. The issue is exceptionally sagacious: (a) Equity may demand that the proceeds of the tax be
15
et al., 1993). Last, it now clearly appears to many developing countries that solving the climate
16 remitted to the developing countries; (b) however, there seems to be no objections raised against
16
change issue is closely linked to future development paths (Villanuova, 1994). So it is in the global
17 tradable permits; (c) as a rule, transferable permits amount to a tax on emissions by all countries.
17
interest that North South trade reliable, cost-effective mitigation.
18
For example, if India or any other developing country seeks to increase its own combustion, it re-
18
2.6
A Sequential Decision Making Framework to Hedge Against Impacts
19 duces its sale of permits to the developed countries and therefore loses that revenue; this is equivalent
19
of Uncertainty and Change
20
to a tax. 13
20
This section concerns the decision making framework for hedging against possible climate changes
21
It turns out that all countries have to reduce their emissions, at least compared with what they oth-
21
and related impacts, and especially establishing decision making paths relative to uncertainty and
22
erwise would do. The growth of non-Annex I countries depends both on emission reductions in
22 risk, exploring the use of financial markets. and pooling different kinds of policy options.
23 Annex I countries and on increased efficiency in their own energy usage. Developing countries
23
2.6.1
Decision making paths: Which strategy would be best?
24 should not be exempted from the tax-as this would lead to inefficiencies. 14 Instead, they could be
24
2.6.1.1
Scenario analysis
12 This argument does not imply that poor countries cannot afford to take any thought for the future. They do, and have to
do, though day-to-day adversity can become pervasive, as it is the case in large parts of Africa. This does not means that
25
The decision making framework to address climate change is served by scenario analysis. Scenario
there is nothing for Northern parties to gain from an agreement, either: a growing prosperous South contributes to the
North through international trade and reduces the possibility of environmental refugees and northward migration pressure
26 analysis is one of the few relatively-useful instruments at hand to look at the here-and-now
in general; Northern countries do have moral and humanitarian interests in reducing poverty of some countries; though
smaller than likely Southern adverse impacts, there are adverse impacts that the North will suffer too; last, we all have
27 significance of alternative futures. Uncertainty is indirectly treated; that is, the very existence of
interest in avoiding an irreversible catastrophe.
13 This example also concerns the problem of decision making at the firm or at the household level (Chapter 8 deals with
28 uncertainty makes alternative scenarios different. but uncertainty is not incorporated within scenarios.
the process of innovation and technology adaptation by firms and households-see 8.3.1.3).
14 In addition, one of the major thrusts of environmental economics is that society must internalise the costs of "free"
natural resources (the atmosphere). This obligation applies to developing and developed countries alike. The carbon tax is
one means of doing this.
Chapter 2/page 45
Chapter 2/page 46
1
1
Within scenarios, therefore, the analysis proceeds as if we definitely had the opportunity to learn the
2
2 state of the world before taking action-in other words, it proceeds as if all uncertainties are resolved
Energy Decisions Sector
3
3 prior to the date at which decisions are taken. Each scenario develops within a certainty-like context.
Damage Potential
4
LOW
4
A decision-making framework is made of a combination of paths that lead to a decision making
5
5
strategy. Scenarios are no decision-making paths. Acting is a path. Learning is one another path. The
Moderate
6
6 best combination of such paths would give the best strategy.
7
7
The size of the strategy will depend on the probabilities assigned to alternative outcomes attached
8
a) Learn. Then and Beyond Adt
Decisions Sector
8
to possible states of the world (which can be served by alternative scenarios). The value of improved
Damage Potential
9
Decisions Sector
Low
9
information (expected value of perfect information) will play a central role.
10
Moderate
10
2.6.1.2 Sequential decision making
11
High
11
Alternative combinations of decision-making paths give alternative characterisations of the decision
and Beyond
12
12 problem.
13
2000. 2010
13
One strategy is to proceed as if we definitely had the opportunity to learn the state of the world
b) Act Then Leam
14
14 before taking action; in other words, as if all uncertainties were resolved prior to the date at which
15
Figure 2.6-1: Alternative decision making strategies
15 decisions will be taken. This decision making strategy can be called a "learn-then act" strategy.
16
A circle denotes a chance node-a poins at which an uncertainty is resolved.
17
A square denotes a decision node-a point at which actions are required.
18
16 Learning comes first. Then, once all uncertainties are more or less resolved, action can be taken.
Manne and Richels (1992).
19
17
One another strategy is to hedge against the probability of surprise such as those risks that fall
20
18 below a particular threshold of probability. Such a strategy can be sequential, proceeding under an
Box 2.6-I: Emissions level under "learn-then act" procedure (hypothetical)
Damage potential
Carbon emissions (2010 target)
19 "act-then learn" characterisation of the decision problem: decisions are taken (acting) for the years to
Low
Unlimited (hypothetical)
20 come without knowing which of the possible states of the world will occur in a sequence of decisions
Moderate
20% below 1990 level (hypothetical)
21 allowing new information (learning) to be incorporated into a complex decision process. 15
High
50% below 1990 level (hypothetical)
21
Source Manne and Richels, 1993
22
See Figure 2.6-1(a) on which a circle denotes a chance, and a square a decision node.
23
To illustrate, Box 2.6-1 shows hypothetical emissions constraints ascribed to each of the three
22
If the damage potential proves low, then carbon emissions would remain unconstrained. With a
24 states of the world under the "learn-then act" procedure (Fig. 2.6-1a).
23 moderate damage potential, the 2010 target would be 20 percent below 1990 levels, and with a high
25
Figure 2.6-1 shows the decision trees:
24 damage potential, the 2010 target would become a 50 percent reduction (incidentally, this might not
26
(a) for a "leam-then act" decision strategy (learning about damage potential, then acting) and
25 be possible to be implemented at that time).
27
(b) for an "act-then learn" decision strategy (hedging against damage potential by acting first, then
26
Under a "act-then learn" hedging policy, decisions would be taken at discrete points of time
28 learning about the damage potential, and then acting again).
27
separated by intervals of one decade. It would be assumed that prior to 2020, the energy sector's
28
supply and demand decisions should be made under uncertainty about reaching a consensus on
15 Hedging is not equal to "act-then learn" as learning does not necessarily take place.
Chapter 2/page 47
Chapter 2/page 48
1
carbon limits. From 2020 onward, it would be supposed that decisions would be made after the
2
resolution of uncertainty. Under the "act-then learn" strategy, we would make decisions for 2000 and
Strenghtened
3
2010 without knowing which of the states of the world (low, moderate, and high damage potential)
Continued
4
would occur and should select an optimal hedging policy for the emissions levels¹⁶.
Aggressive
5
A sequential hedging strategy consists of adopting an emissions level that lies between the extreme
Relaxed
6
cases shown by different emissions scenarios, and the size of the hedging strategy will depend on the
Strenghtened
7
probabilities assigned to alternative outcomes. To illustrate a case in which the expected value of
Moderate
8
perfect information (EVPI) would be high, and in which the probability of low damage would be 60
Continued
9
percent and that the probability of a 20 percent reduction 1.5 times that of a 50 percent reduction,
Relaxed
10
then, Manne and Richels (1992) conclude, if controls are required, they are more likely to be
I
11
moderate than high.
2
Figure 2.6-2: Decision tree for sequential policy choice.
3
At the beginning of the first period, (1992) either the moderate reduction or aggressive reduction near-
12
One another example of a sequential decision making is reported by Hammitt el al. (1992).
4
term abatement policy is selected. Ar the beginning of the second period (2002). uncertainties about
5
climate sensitivity and the damages of climate change are substantially resolved and the chosen policy is
13
According to well publicised works (see also Singer et al., 1991; Manne and Richels, 1991, 1992;
6
amended to limit climate change to the "optimal" level. Hammitt et al., (1992).
7
14
IUCC, 1993; Richels, 1994; Peck, 1994), first decisions might be to take those actions where the cost
8
Hammitt et al. consider two near-term policies, aggressive and moderate reduction (Fig. 2.6-2).
15
is not high while considering more aggressive-reduction policies (Fig. 2.6-2).
9
Under the aggressive-reduction policy, conservation and fuel-switching are adopted together
16
First, energy conservation represents a set of low-cost, quickly implemented policies. To illustrate,
10
starting in 1992.
17
if adopted at Lo, energy intensity falls 30% over 20 years (Hammitt et al., 1992). This corresponds to
11
Under the moderate-reduction policy, only the conservation policy is adopted and implemented in
18
some minimal agreement made of two complementary actions: conserve energy by discouraging
12
1992. At the beginning of the second period (2002), uncertainties about climate sensitivity and the
19
wasteful use globally, and improve efficiency in energy use (Singer et al., 1991).
13
damages of climate change are thought to be substantially resolved and the chosen policy is amended
20
Second, fuel switching represents a more aggressive-reduction policy. It is a set of high-cost,
14
to limit climate change to some "optimal" level (though this could appear as an irreversible process).
21
slowly implemented emission-reduction technologies. They are produced by long-lived capital
22
15
2.6.1.3
Clearing up the uncertainties? The value of improved information
equipment using either emitting or non-emitting technologies (fossil fuel vs. nuclear, solar, biomass).
23
16
Under a "learn-then act" sequential decision making, the value of new information depends on
Both have construction periods of about 10 years, and an operating period of roughly 30 years
24
(Hammitt et al., 1992). The motto is: use non-fossil fuel energy sources wherever this makes eco-
17
changes in the probabilities assigned to the (alternative) scenarios before and after the study. "If the
25
18
nomic sense (Singer et al., 1991). "The policy decisions are whether to begin substituting non-emit-
probabilities of (alternative) scenarios remain equal, then the value of the study is zero; if, on the
26
19
other hand, only one scenario can be selected, a study might be worth as much as 100 billion dollars"
ting or emitting equipment, and if so, the rate at which to substitute, specified as the transition half-
27
20
life, or the time until half the capital stock then operating is non-emitting" (Hammitt et al., 1992).
(Shlyakhter et al., 1995).¹⁷
28
17 Values of situations before and after imperfect or perfect information is received can be computed. When there is risk
neutrality, these values also represent expected utility. Suppose that expected results are as follows: (a) without
information: +150; (b) imperfect information: +325; (c) perfect information: + 500. This hypothetical example "illustrates
a perfectly general result: the value of perfect information (500 150) clearly exceeds the value of imperfect information
consumption (Global 2100 by Manne and Richels, 1992).
16 The period-to-period allocation of carbon emissions is determined so as to maximise the expected discounted utility of
(325 150)"; it also indicates that some information-albeit imperfect-is better than no information at all (Eeckhoudt
and Gollier, 1995). Hourcarde and Chapuis (1994) have shown that, accounting for climate change related surprises under
Chapter 2/page 49
Chapter 2/page 50
1
Under an "act-then learn" sequential decision making, there is an agreement that uncertainty
1
moderate, or high (conditional probabilities; cf. Manne and Richels, 1992).
2
cannot be resolved before a decision is taken. This does not mean that uncertainty is an obstacle for
2
This representation allows the use of standard probabilistic techniques to incorporate the outcomes
3
decision making. This means that first, its resolution is at a too high a cost and, second, that maybe it
3
of the research program in the decision model (Global 2100, in this case).
4
will not facilitate a decision-clearing up the uncertainties could make the greenhouse warming issue
4
Maximum accuracy, or perfect information, would eliminate the need to hedge against surprise.
5
more difficult to resolve (Waterstone, 1993). Accordingly, it is for instance assumed that the damage
5
Accuracy of the forecast, however, will not be maximum. We therefore have to accommodate with
6
potential will remain unknown until 2010-2020. As a consequence, the appropriate emissions level
6
hedging under imperfect information. In this case, the research outcome will lead to a revision of our
7
would be known only well into the second decade of the twenty-first century. The actual level of
7
beliefs. It will help at modifying near-term decisions.
8
investment is also influenced by the conflict between the long term nature of the problem (and the
8
A sequential decision making strategy made of acting and learning decision making paths would
9
time that would be required for resolution of the uncertainties) - and the short term public and private
9
be the best strategy.
10
decision makers horizons (sometimes called NIMTOO phenomenon). Today decisions must thus
11
consider the partial disclosing of the uncertainties along with the long-term effects of short term
10
2.6.2
Exploring mutual insurance decisions and markets for risk
12
decisions. Sequential decision making can address this point.
11
Insurance is one another way to respond to risks and uncertainty which is worth to explore."
13
12
Transferring risk from one agent (country) to another or sharing risk collectively does not make
14
13
the risk disappear though restrictions can be attached to that kind of transfer. Yet, under what condi-
15
Research
Research Outcome
Energy Sector
True Damage
Energy Sector
14
tions, and why will an agent (country) demand insurance to cover possible climate change damages?
Investment
(damage forecast)
Decisions
Potential
Decisions
16
Low
Low
15
Can market evaluate the price of climate change risk? How would the allocation of risk be decided?
17
16
Exploring insurance decisions means addressing the problem of determining the economic value
Moderate
Moderate
18
17
of things. In particular, it goes much further than simple knowledge of costs and benefits of climate
High
High
19
18
change; that is, risks. Exploring insurance decisions and, consequently, markets for risks is addressing
Today
2000. 2010
2020
20
and Beyond
19
the possibility of determining the final allocation of these risks. Who bears the economic risk of the
21
20
possibility of climate changes which is one of the large risks of our modern economies?
22
Figure 2.6-3: Decision tree for value of information analysis.
21
In particular, do transfers of risk improve welfare in an economy (or is it instead a 'zero sum
23
Source: Manne and Richels (1992)
24
22
game')? It appears from the literature that "the transfer of risk is a potential source of large improve-
25
A decision tree reflecting this type of sequential decision making can be adapted from the one in
23
ments in economic efficiency and social welfare" (Eeckhoudt and Gollier, 1995).
26
Figure 2.6-1b. Figure 2.6-3 shows this for the energy sector's near-term supply and conservation
24
Insurance markets leave uncovered collective or correlated risks such as the risk induced by
27
decisions that must be made with respect to the uncertainty regarding the ultimate damage potential.
18 The word 'insurance' is sometimes abused in climate debate. For instance, the insurance premium of, e.g., Manne and
Richels (1992) is in fact a risk premium in the Arrow-Pratt sense (the Arrow-Pratt insurance premium describes situations
28
In this case, the degree of uncertainty is represented by a likelihood table indicating the probability
where, because of risk aversion, people are willing to pay more than expected yearly damage), and the insurance scheme
of AOSIS is in fact a liability scheme (see below). In addition, one can think that there is no analogue between the impact
29
that the damage forecast will be low, moderate, or high-given that the true damage potential is low,
of climate change and the insurance market because climate change is not an isolated event, and that to control portfolios
of insurances requires the risks to be fairly well known. Yet, the insurance industry is currently giving up because of
increasing losses that are not indispensably linked to climate change-related natural catastrophes such as a new insurance
orchestration must anyhow be envisaged. Simple evidence is given here to decision makers (the approach is purely
a sequential decision model, a resolution of uncertainty had high value relative to no-regret potentials and technical
normative due to the absence, to date, of any applied analysis of climate change insurance scheme). Note that the impacts
innovation. Peck and Teisberg (1992) and Manne and Richels (1992) show similar results.
of extreme events on property insurance market and adaptation to climate change by property insurers are the concern of
IPCC WGB Chap. B9 on "Financial Services" and consequently are out of the scope of this section.
Chapter 2/page 51
Chapter 2/page 52
1
ignorance of the true frequency distribution of harmful events. According to Chichilnisky and Heal
I
either laissez-faire or to let suboptimal policies being implemented-which for sure are biased
2
(1993), our ignorance of the frequency of the impacts of climate change constitutes a collective risk.
2
against the ecological-economic long run-, trading climate change risk positions that are contingent
3
This collective risk can be allocated through markets for securities that pay off contingent on that
3
on which model of the change is correct would interlace systems that basically are highly differenti-
4
frequency. For the individual risks that remain, it is more practical to use mutual insurance contracts:
4
ated ones. In addition, countries would have to express their estimates on global climatic risks
5
this is done by having a different individual insurance contract for each possible frequency of
5
through a market commitment and therefore will have to express their true views. If a country sells
6
impacts.
6
securities that pay off in the event that climate changes are not serious and buys those paying off if
7
7
2.6.2.1
they are serious on terms that do not correspond to its true beliefs, it will suffer a financial loss
Use of financial markets
8
(Chichilnisky, 1994, Pillet, 1994).
8
According to Chichilnisky and Heal (1993) an interesting aspect of financial markets is that they can
9
provide a natural mechanism for reconciling differences in assessments of the likelihood of important
9
2.6.2.2
Insurability of risks associated to climate change
10
climate changes between countries, and for testing the conviction behind publicly-stated positions.
10
The insurance issue can also be accorded our attention in contrast with mitigation strategies.
11
International markets for risks of climate change would also provide an objective test of the
11
Basically, in view of uncertainties about consequences of carbon emissions, the question is: Is it
12
seriousness with which countries adhere to their publicly-professed positions on the risk of climate
12
worth taking costly and disruptive measures to reduce levels of emission far enough to stop the build-
13
change.
13
up of CO₂ (Chichilnisky, 1994)? Next, should you consistently decide to wait until more is known
14
The Arrow-Chichilnisky-Heal idea is then to settle a framework involving securities whose pay-
14
and better estimates of the consequences are available, would it not be cautious to insure against
15
offs depend on which description of the (collective) risk is correct. "Betting on which of several
15
harmful events even if the chance of their occurring is small and not immediate? Is a trading of risk
16
alternative descriptions of the way the world works is correct, is in effect what one does when
16
positions between regions of the world conceivable?
17
choosing one research strategy over another. (...) For example, a market for the securities of high-
17
Climate change associated risks are first difficult to quantify (in a classical statistical sense, the
18
technology firms pursuing different research strategies towards the same goal is a financial market in
18
probabilities describing them are unknowable). They are then endogenous (once steps have been
19
which these bets are made" (Chichilnisky, 1994).
19
taken toward reducing climate associated risks), correlated (which demands complex analysis to
20
The trading of climate risk positions from a global ecological-economic viewpoint is amply
20
assess probabilities), and collective (by which the uncertainty about the relative frequency of harmed
21
thinkable. In short, ecological models are energy oriented models whereas economic ones are based
21
agents is meant). In principle, however, once frequencies are at least approximately known, actuarial
22
on the behaviour of individual agents and groups. Under the case for climate change prospects, the
22
calculations can be set up, and the problem can be conceived as insurable.
23
behaviour of the agents contributes to make particular energy profiles to appear and develop, with
23
For instance, to the extent that carbon emissions lead to uncertainties about climate change and the
24
unknown feedbacks on future economic outcomes (Pillet, 1994).
24
corresponding damage that such changes cause, it can be said that the uncertainty is induced by the
25
Under this perspective, one better has to place the focus on the various matching and diverging
25
industrial countries (to date), and borne by the world as a whole. The available data also indicates that
26
points that dynamically occur; that is, on what will possibly emerge from such interconnections. This
26
developing countries are more vulnerable to climate change. Any insurance scheme must take this
27
amounts to allowing agents to identify the set of possible descriptions of the risks that humanity faces
27
distribution of causes and risks into account.
28
and to bet on which model of the climate change related risk is correct.
28
The floods of 1993 in the US, in Bangladesch, and in Europe (France, Switzerland) have reminded
29
Ecological systems (assets) cannot easily be allocated on the marketplace. Therefore, instead of
29
us of the profound vulnerability of human settlement to climate. There is considerable natural varia-
Chapter 2/page 53
Chapter 2/page 54
1 tion in these phenomena, although climate scientists expect that variation induced by human activity
I greenhouse gas emissions depend on their discount rates and degrees of risk aversion (even if there
2 will soon come to except the natural variation.
3
2 were complete agreement about all of the scientific aspects of the global change problem, there could
Box 2.6-2: Natural events and attached Insured losses (after Weilenmann, 1994)
3 still be disagreement about policy responses). Differences in policy positions could then be attributed
Event
Date
Insured Losses (in billion US$)
4 to differences in discount rates and degrees of risk aversion rather than, or in addition to, different
Hurricane Gilbert
September 1988
0.05
Hurricane Hugo
September 1989
5.8
Winter storms Europe
January/February 1990
10.0
5 interpretations of the current scientific evidence.
Colorado storms
July 1990
1.0
Hurricane Bob
August 1991
0.62
6
Yet, different perceptions of the risk involved do not however preclude efficient solutions.
Hurricane Andrew
August 1992
15.5
Hurricane Inki
September 1992
1.6
East coast storms US
7 Differences in attitudes toward risk could merely lead to the introduction of new markets in which
March 1993
1.6
Midwest floods US
Summer 1993
0.765
4
8 different risk positions are traded, with efficiency gains. Economics is about differences in prefer-
5
Correspondingly, one must bear in mind that before 1989, natural disasters caused by storm events
9 ences leading to trade (Chichilnisky, 1994). Betting on climate states could be envisaged, not only
6 with insured losses greater than one billion dollars were unknown (Weilenmann, 1994). For insurers,
10 because insurance can be provided within a market framework, but basically because there exists an
7 the most costly storm event to date has been 1992 hurricane "Andrew" with insured losses of
11 ecological-economic ground for the use of financial instruments in addressing climate change collec-
8 US$15.5 billion (see Box 5.2-II). This kind of losses has become more frequent and a potential rise is
12 tive risks induced by uncertainty about the overall/regional distribution of adverse effects (Pillet,
9 predicted on the standpoint of the insurance industry (Weilenmann, 1994). However, the causes for
13 1994).
14
10 that trend are not clear and are not conclusively linked to current "greenhouse activity". They seem to
The institutional structure to be developed would use two types of financial instruments in order to
11 primarily depend upon the increasing concentration of people and insured economic values in areas of
15 deal with the two above-mentioned aspects of the problem. The aim is to obtain an efficient allocation
12 high risk although losses in climate related disasters have been increasing since the 60s at a rate 3:
16 in the face of those risks. One instrument is a mutual insurance contract which is tailored to deal with
13 times that in earthquakes (Yokohama document, 1994).
17 the risks faced on a regional basis by communities contingent on the distribution of harmful effects
18 world-wide. The other one is related to Arrow securities because we need such securities to deal with
14 2.6.2.3 A market framework to respond to collective risks
19 collective risks induced by strong uncertainty about the overall distribution of adverse effects.
15 Next, what a society believes it to be worth paying to reduce the risk of climate change depends inter
20
A mutual insurance contract is an agreement between parties subject to similar risks that those
16 alia on two key characteristics of that society: its degree of risk aversion and its discount rate.
21 who are harmed will be compensated by those who are not.20 In the climate change context, one can
17
The degree of risk aversion is a measure of how much an individual or group is willing to pay to
22 think of groups of communities subject to the possible impact of that change, with those unharmed
18 avoid a risk. 19 The discount rate is an equivalent measure of the value of the future relative to the
23 compensating the others. To illustrate, the Alliance Of Small Island States (AOSIS) has proposed that
19 present (see Chapter 4).
24 an "International Insurance Pool" be established to cover climate change impacts. The Pool would
20
It consequently turns out that different country's positions with respect to measures to restrict
25 take the financial burden suffered by affected developing countries and distribute it amongst the
26 industrialised countries. The members of the Pool would be divided into two groups. The first group
19 An indication of the degree of risk aversion can be obtained by the following exercise: offer a person the choice
between either $0 if a tossed coin comes down heads and $100 if it comes down tails or $50 with certainty. Most people
27 would consist of low-lying coastal countries and small island states that would receive insurance
will choose $50 with certainty, even though on average they will get the same by choosing to bet 00 the tossed coin and
they stand some chance of doing better (and some change of doing worse). In fact most people will choose $45 with
28 coverage from the fund. The second group would include industrialised countries who would
certainty to a bet on a tossed coin giving $0 for heads and $100 for tails. Suppose that $45 is the smallest amount for
which this is true. Then the difference between $45, the reward in the risk less situation, and $50, the average reward in
the risky situation, is a measure of the person's risk aversion. The more risk averse you are, the more you will insure
20 Examples are agricultural co-operatives of the type recorded in Europe at least since the fifteenth century, and the nine-
(Chichilnisky, 1994).
teenth century UK workers' associations and friendly societies. These involved agreements between a group of workers
that if one were sick and unable to work, he or she would be compensated by the others.
Chapter 2/page 55
Chapter 2/page 56
1
contribute funds (after IUCC, 1993).
I
safe guide to the far future" (Ayres and Weaver, 1994). The current trend in the insurance industry
2
Mutual insurance contracts would mean making transfers, the size of the latter depending on the
2
with respect to weather events is a clear attempt at modifying the underwriting guidelines by setting
3
overall level of prosperity, and the transfer per se being contingent on the overall incidence of
3
restrictions on coverage (exclusions), at setting total exposure limits for a specific high risk zone to
4
negative effects (in the sense of high or low overall incidence of negatives climate change-related
4
avoid the cumulation of risks (for instance the Eastern coast of US, Japan, Australia, low lying coastal
5
events in the population and in ecosystems).
5
areas), or at improving overall risk exposure by selective underwriting.²
6
The AOSIS proposal calls for Pool contributions to be collected in 2004, ten years after the Climate
6
To conclude, the markets just described can provide a natural mechanism for reconciling differ-
7
Change Convention enters into force-provided the rate of global mean sea-level rise has by then
7
ences in assessments of the likelihood of important climate changes between countries, and for testing
8
reached an agreed figure. This ten-year delay has been thought to give the funding countries sufficient
8
the conviction behind publicly-stated positions (Chichilnisky, 1994). Accordingly, this could be help-
9
time to make their initial investments. If sea-levels have not risen too much by the time the ten-year
9
ful in advancing negotiations between industrial countries with different perceptions of the risks, such
10
period expires, a review of conditions would be undertaken.
10
as the US and the EU (also see Pillet, 1994)23 as well as in correcting the North-South balance
11
The securities we would need to deal with collective risks induced by uncertainties about the
11
between the causes of climate risks (in the North), and the ecological-economic impacts of the risks
12
overall distribution of adverse effects would pay out if and only if there were a particular frequency
12
(which are believed to be more serious in developing countries).
13
of damage in the world as a whole. This would involve identifying the set of possible descriptions of
13
2.6.3
Portfolio Analysis Analogy
14
the risks that humanity as a whole face, which are called here collective risks. Climate-related events
14
Numerous policy measures are available to limit the impacts of climate change. Countries will typi-
15
such as floods, tropical storms or certain temperature patterns are examples.
15
cally implement several of these policy measures since no one measure is clearly superior. Indeed, the
16
Introducing, therefore, securities whose payoffs depend on which description of the risk is correct
16
decision-maker's problem is much more complex than the choice of a single action/policy measure to
17
would allow agents to bet on which model of the risk is correct. A similar financial instrument is
17
the exclusion of all others. In practice, it is a question of constructing a portfolio of actions/policy
18
already in use to hedge against environmental uncertainties and risks: CAT Futures (for Catastrophe
18
measures with the goal of best using the opportunities to diversify that are present in the economy.24
19
Futures). This instrument is a security which pays contingent on the frequency of the occurrence of a
19
Although governments will be making climate change decisions for most of the next century, this
20
major weather risk, such as tomadoes in the East Coast of the US, and floods in the Midwest. CAT
20
does not mean that they have unlimited opportunities to adjust the size and mix of their portfolio of
21
Futures are traded by insurance companies to hedge against their exposure to major climate risks as
21
climate change policies. The total resources devoted to measures to limit climate change (the size of
22
some sort of reinsurance scheme.
22
the portfolio) and the mix of measures implemented (the mix of the portfolio) cannot be changed as
23
Yet, would insurance companies be able to address global climate change risks? Currently not. A
23
often as governments deem appropriate because of lock-in effects as well as other time-, attitude-, and
24
new financial structure has to be set up.
24
scientific results-related factors. In many cases, instead of making decisions, options will be taken.
25
For the insurance industry, which is used to drive fixed, short-term contracts, climate change intro-
26
duces new circumstances "in ways that mean traditional practice and the old statistical methodologies
22 Swiss Insurance Industry, anonymous executive. Pers. com., 1994.
27
are no longer adequate". Climate change implies a certain degree of irreducible unpredictability be-
23 Suppose for example the US believes it most likely that there will be little climate change, and the European Union
believes otherwise. Then through the market for securities whose payoffs depend on which description of climate change
28
is correct, the US will naturally sell insurance to the EU. The US would wish to be a seller of securities which pay if
yond the category of risk-like uncertainties For insurance companies, "past experience is no longer a
climate change is serious, because of its belief that this event will not occur, and a buyer of securities that pay if it is not,
because of its belief that this will be the outcome. The EU would be on the opposite sides of these markets. (Chichilnisky,
21 Others would say: the island countries would have to contribute some funds to the Pool. Note that the AOSIS proposal
1994).
was presented to the negotiators of the Climate Convention but was not included in the final treaty-and that it is closer to
24 This portfolio is not the same as a portfolio of an investor; however, decision makers should envisage the evaluation of
a liability scheme than to an insurance contract.
the expected return and the risk of that portfolio.
Chapter 2/page 57
Chapter 2/page 58
I
2.6.3.1
Selecting a portfolio
1
political system must take for itself (countries do not use information; government departments may
2
In this respect, the key for selecting an optimal portfolio is to understand how the options interact
2
use information).
3
-e.g., better greenhouse information will influence technology development and emission reduction
3
2.6.3.2
No operational model is available
4
decisions (Manne and Richels, 1993).
4
5
An operational model to analyse the optimal portfolio of climate change policies for a country is not
In principle, the policy measures/options available to countries to limit climate change and its im-
5
feasible given the state of our knowledge. But even the sparse information on the expected costs and
6
pacts include:
7
technology research aimed at improving the cost-effectiveness of greenhouse gas mitigation
6
benefits of some amongst the broad policy options can yield some useful insights.
8
technologies such as energy efficiency and non-fossil energy sources;
7
9
Peck and Teisberg (1993), and Manne and Richels (1993) have estimated the value of spending
joint implementation, technology transfer and other forms of international co-operation to limit
8
on climate research. It is shown that the expected return is several times the current level of
10
climate change;
9
climate research spending. The same analysis could be performed for technology research. To
11
phasing out existing policies, such as subsidies to fossil fuels, that reduce welfare and directly
10
better encounter uncertainty, the main idea is to consider spending on climate research and cli-
12
of indirectly increase greenhouse gas emissions;
11
13
mate-related technology improvements as an risk premium, and not as subsidies to be randomly
measures to reduce emissions, or increase sequestration, of greenhouse gases,
12
14
allocated. Indeed, the key for selecting an optimal portfolio is to understand how the options in-
measures to adapt to the consequences of climate change;
13
15
teract, e.g., better greenhouse information will influence technology development as well as
insurance orchestration to hedge against risks revealed by strong/unexpected uncertainties;
14
emission reduction decisions.
16
insurance for additional costs incurred due to adaptation.
15
Several researchers compare the costs of unilateral action and international co-operation to ad-
17
The specific policy measures available vary from country to country; research capabilities vary,
16
dress climate change. The analyses consistently show large returns to international co-
18
the inefficient existing policies vary, the costs and availability of measures to reduce greenhouse gas
17
operation. International co-operation should however contain an instrument by which welfare
18
can be redistributed among nations. If it is not the case, as shown by Pillet et al. (1993), small
19
emissions vary, the costs and availability of measures to increase sequestration of greenhouse gases
19
countries with a comparative advantage in the production of clean commodities and a high
20
vary, the damages due to climatic change and the measures for adapting to these damages vary.
20
emissions reduction marginal cost, might have arguments for doing nothing and thus adopting a
21
A portfolio manager attempts to get the best return given a level of risk (as derived from a given
21
free-rider behaviour until a complete international climate agreement is concluded.
22
22
An OECD analysis of the effects of phasing out existing inefficient policies suggested that
level of uncertainty of a given type). The appropriate level of risk can be derived from the investor's
23
greenhouse gas emissions would be reduced by over 25% with a net economic benefit. In many
23
assets using assumptions about maximising expected value behaviour. The investor can also impose
24
cases eliminating existing subsidies is politically very difficult, nevertheless a large net benefit
24
constraints on the portfolio to reflect his or her preferences. Investing some or all of the funds in
25
is possible from this policy.
26
25
An emissions reduction supply curve or other analysis of the costs and potential emissions re-
"environmentally friendly" assets or prohibiting investments in specified industries, regions, or assets
27
ductions that could be achieved through a wide range of specific measures has been conducted
26
for ethical reasons are examples of such restrictions.
28
for many countries. Virtually every study shows that some emissions reductions are possible at
27
Countries likewise will select a portfolio of climate change policies that reflect, at least implicitly,
29
very low, possibly negative, cost. Many of the measures also yield other environmental benefits
30
28
such as reductions in acid rain, ground-level ozone precursors, and toxic discharges.
a variety of objectives and constraints. A country may look for the economically optimal portfolio of
31
Adaptation measures that reduce some of the damage due to climate change are available in
29
climate change policies. However, considerations of the impacts of the policy mix on different socio-
32
some countries; building dikes to protect low lying coastal areas from sea level rise for exam-
30
economic groups, international equity, and intergenerational equity are likely to come into play.
33
ple. Such measures can be implemented where they cost less than the damage that would oth-
34
erwise occur.
31
These objectives will have different weights and will vary from country to country. The "right" blend
35
Climate change insurance contracts and transfers, and trading of risk positions between regions
32
of options will include political costs and institutional capacities and is a decision which each
36
and countries can be referred to face unexpected uncertainties and collective risks.
Chapter 2/page 59
Chapter 2/page 60
I
2.6.3.3
Three main forms of policy options
I
strictly economic considerations mentioned above. Expected damage to ecosystems, expected sea
2
It turns out that three mains forms of policy options dominate current discussions: continued intensive
2
level rise, potential socio-economic impacts, international equity concerns, intergenerational equity,
3
science research to reduce climate and impact uncertainties, development of new supply and
3
and economic development strategies based on development of mitigation and adaptation technolo-
4
conservation technologies to reduce abatement costs, and immediate emissions reductions in order to
4
gies are examples of other considerations that may influence a country's climate change limitation
5
slow down climate changes. The issue for the "option-taker" is not one of either-or but one of finding
5
spending.
6
the right blend of options. Policy makers must decide how to divide greenhouse spending among
6
Incorporating option values into the benefit-cost analyses of the policy measures indicates that
7
competing needs. What portion goes to resolving climate uncertainties? What portion goes to tech-
7
some action on climate change is economically justified by every country. No country can be confi-
$
nology development? And what portion goes to immediate abatement of emissions? (Richels, 1994).
8
dent that it will not incur damage as a result of climate change. Thus, delaying action to limit climate
9
Using the portfolio analysis analogy, a country must use the available information to decide upon the
9
change is an option that has a cost, at least an insurance premium. Every country has available policy
10
size of its investment in climate change limitation and on the allocation of that investment across
10
measures that have negative or very low costs after accounting for any associated benefits. Thus ev-
11
policy measures.
11
ery country has some climate change mitigation measures that are economically justified.
12
The size of the investment will reflect the country's objectives and its obligations under interna-
12
Normally, the mix of assets in the portfolio is adjusted so that the risk-return ratio is equal at the
13
tional agreements (FCCC, 1992). Assume that the country has ratified an international agreement that
13
margin. As discussed above, only sketchy information is available on the expected costs and benefits
14
specifies national greenhouse gas emissions limits. Then the size of its climate change investment
14
of the main policy measures. The variability of the costs and benefits for these broad policy options is
15
reflects the most cost-effective means of achieving the country's agreed emissions limit plus
15
not known. Governments can make subjective assessments of the variability of costs and benefits of
16
expenditures on research and adaptation measures. In the absence of an international agreement a
16
different policy measures. The risks associated with phasing out existing inefficient programs or im-
17
country may have little alternative but to implement necessary adaptation measures. Since almost
17
plementing energy-efficiency measures are lower that those associated with reforestation to sequester
18
every country accounts for only a small fraction of global greenhouse gas emissions, unilateral
18
carbon or construction of sea walls. The latter offer only climate change benefits while the former
19
actions will have only a negligible effect on the pace and extent of climate change. No country, acting
19
also yield other benefits.
20
alone will be able to appreciably reduce the damage it suffers due to climate change except by
20
21
implementing adaptation measures. This suggests that all countries have an economic incentive to
22
reduce emissions of greenhouse gas emissions.
23
At a global level, benefit-cost analyses can help establish an appropriate level of climate change
24
investment. Climate change policy does not involve a one-time choice-or some go/not go type of
25
decision-between distinct, long-term options (do nothing or implement policy "X" for the next
26
century), the types of decisions to which benefit-cost analysis has historically been applied. Since
27
climate change policy can be modified frequently and incrementally, the "option values" of decisions
28
must also be considered (the option value of deferring action on climate change can be interpreted as
29
an insurance premium to cover the cost of damages that may result).
30
The level of their investment in climate change policy will be influenced by factors beyond the
Chapter 2/page 61
Chapter 2/page 62
2.6 LINKAGES OF CLIMATE SYSTEM WITH DECISION MAKING
2
FRAMEWORK
,
4
In this section, we briefly outline the linakges of decision making framework with
,
climate system illustrated in figure 2.8. Each square is fully discussed by various chapters in
this report, so we restrain from elaboration. There are two other working groups on scientific
,
and impact assessments, summary of which have to be reflected also in the decision making.
Thus, the framework encompasses outcomes of all the chapters in this report and highlights of
the other working groups.
2
3
10
The figure shows that the chain of events start from cumulative emissions of
"
greenhouse gases (square 1 in the figure 2.8). These cumulated emissions may trigger off
12
various possible climate related changes such as change in the temperature, sea level rise,
SCIENTIFIC EFFECTS
Change in Temp.
-Sea Level Rise
-Moisture Loss
-Precipitation Loss
Extreme Events
IMPACTS
-Damage to Agricul.
and Fishing
-Heat Stress
-Migration
Species Loss
-Irriy. & Hyd. Power
IRREVERSIBILITIES
Need for Adaptation
" moisture loss, precipitation change and extreme events (square 2). These climate changes would
LINKAGES WITHIN CLIMATE CHANGE DECISION
Long term simulation
through Climate Research
Policy Research
Integrative Assessment
16
in turn have impacts such as damage to agriculture and fishing, migration of population from
19
the coastal areas, species loss, and scarcity of water and so on, which in turn, will constrain
16
irrigation and hydro power generation (square 3). However, the decision making today has to
MAKING FRAMEWORK UNDER UNCERTAINTY
5
9
17
consider climate changes and impacts that may occur 25, 50, 100 years and beyond and
"
respond to it_now by formulating, say, 10 or 20 year strategies that have to be periodically
M
revised. Since we do not know what may be in store, we have to rely on climate change
CUMMULATIVE EMISSIONS
ABSORPTION
-Oceans
Forests
- .
Other
COLLECTIVE DM
(Risk based Criteria)
RISK Tolerance
-Equity of Stakeholder
Future Generation
-Ecocentric Approach
DECISION TOOLS
(Utility based)
Optimization
-Cost-Benefit
-Discounting
Integrative
Assessment
Figure 2.8
64
20
research to say how the situation would look at different time periods. Economic costs of
-Poverty
21
adaptation and irreversible damages for these impacts have to be compared with the abatement
6
22
8
costs using various decision tools. Thus research results are fed into collective decision making
"
process, which takes into account issues such as inter-regional and intergenerational equity,
24
ecological and poverty considerations apart from the various country specific factors such as
EMISSIONS
25
risk tolerance and vulnerabilities (square 4). To assess the impacts and to facilitate the decision
SOCIO-ECONOMIC SYSTEMS
C C/B E/Y Y/N N
C/E Carbon Intensity of
energy
E/Y : Energy Intensity of
Y/N : Income levels per
person
Population ..
RESPONSE OPTIONS
L
GDP
26
-Technological
makers in designing the specific abatement policies, one can use various decision tools such
-Life Style
-North -North-South
Transfers
Increase Sinks
POLICY INSTRUMENTS
Market:
-Tradeable Permits
-Joint Implimentation
Carbon Taxes
Non-Market:
-Quotas
-Climate Research
N
27 as decision theory, optimization, simulation, forecasting and more importantly integrative
20 assessment models (square 5).
7
63
I
Thus, collective decision making is required to arrive at revised strategies to take into
I
2.7
References
2
account the new information from climate research, and policy research. This helps to choose
,
the desired concentration levels and emission paths. Allocation of these emission levels among
2
Allais, M., 1953: "Le comportement de l'homme rationnel devant le risque, critique des postulats et
3
axiomes de l'école américaine". Econometrica 21: 503-546.
4
parties requires consideration of current and past emissions, need for development for meeting
4
Ancklin, M. et al., 1993: "Climate Instability During the Last Interglacial Period Recorded in the
,
basic needs and resource endowments and global cost effective criteria. For global cost
5
GRIP Ice Core". Nature 364: 203-207.
minimisation, collective decisions are needed to choose among various global policy
6
Arrow, K.J., 1951: "Alternative Approaches to the Theory of Choice in Risk-Taking Situations".
,
instruments, such as market based policy instruments, viz, tradeable quotas, carbon taxes, joint
7
Econometrica 19: 404-437. Reprinted in Arrow, KJ., 1984: Collected Papers of Kenneth J. Arrow:
8
3, Individual Choice under Certainty and Uncertainty. Cambridge, Mass.: The Belknap Press of Har-
implementation, as well as other non-market policies such as setting standards, reduction
9
vard University Press, pp. 5-41.
targets, etc. One can choose a set of instruments or a combination of these instruments to
10
Arrow, K.J., 1971: "Exposition of the Theory of Choice under Uncertainty". Decision and Or-
11
10
ganization, ed. C.B. McGuire & R. Radner. Amsterdam: North-Holland, PP. 19-55. Reprinted in
achieve the targets of emission reductions (square 6). This also forms a part of policy research.
12
Arrow, K.J., 1984: Collected Papers of Kenneth J. Arrow: 3, Individual Choice under Certainty and
"
The choice of a specific policy instrument at national level could be based on a variety of
13
Uncertainty. Cambridge, Mass.: The Belknap Press of Harvard University Press, pp. 172-208.
12
utility based criteria such as cost effectiveness, or cost benefit analysis, which may depend on
14
Arrow, K.J., 1984: Collected Papers of Kenneth J. Arrow. Cambridge, Mass.: The Belknap Press of
15
Harvard University Press.
13
the discount rate used. The national policy instruments would generate in certain response
16
14
Arrow, K.J., 1987, 1990: "Economic Theory and the Hypothesis of Rationality". The New Palgrave.
strategies; these may use ustility based criteria (square 7). The response strategies could be
17
New York-London: W.W. Norton & Company, pp. 25-37.
19
chosen from among portfolio of options such as alternative technologies, changes in
18
Arrow, K.J., 1994: "Discounting and Climate Change Prospects". Communication at the IPCC
16
consumption patterns and life style, increase in the sinks, etc. These response strategies would
19
Working Group III Writing Team II Montreux Meeting, March 3-6. Montreux, Switzerland.
17
again have an effect on the socio-economic emissions (square 8). The emissions of GHG
20
Arrow, K.J., and Fisher, 1974: "Environmental Preservation, Uncertainty and Irreversibility.
21
"
Quarterly Journal of Economics, pp. 312-319.
would depend on socio-economic factors such as carbon intensity of energy, intensities of the
19
GDP, per capita GDP and population. Land and forest use patterns also matter, but there may
22
Atomic Energy Commission, 1975: An Assessment of Accident Risks in US Commercial Nuclear
23
Power Plants. Washington, D.C.: US Atomic Energy Commission, WASH-1400.
be a decreasing trend of the emissions associated with these. Some of these emissions are
24
Ayres, R.U., and P.M. Weaver, 1994: "The Case for Proactive Risk Management". Environmental
21
absorbed by oceans, forests, agriculture and other related activities (square 9). The fraction of
25
Hazard: Liability or Opportunity? Report of the SANDOZ/INSEAD Colloquium held 29 June 1993.
26
22
Paris: INSEAD, Centre for the Management of Environmental Resources, pp. 3-9.
the total emissions which could not be absorbed by the sinks accumulate in the environment
23
over a period of time and leads to cumulation (square 1) completing the cycle. Sequential
27
Beltratti, A., "Environmental Problems and Attitudes Towards Risk and Uncertainty". In: Pillet and
28
Gassmann (Eds.), Steps Towards a Decision Making Framework to Address Climate Change.
24
decision making is required periodically to verify strategies.
29
Report of the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen
30
& Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer Institute, PP. 67-74.
25
31
Bernthal, F.M. (Ed.), 1990: Climate Change. The IPCC Response Strategies. Geneva: World Meteo-
32
rological Organization, and Nairobi: United Nations Environment Program. Intergovernmental Panel
33
on Climate Change, 270 pp.
34
Björkman, M., 1984: "Decision Making, Risk Taking and Psychological Time: Review of Empirical
35
findings and Psychological Theory". Scandinavian Journal of Psychology 25: 31-49.
36
Blaug, M., 1992: The Methodology of Economics-Or How Economists Explain. Cambridge
37
Surveys of Economic Literature. Cambridge: Cambridge University Press, second edition, 286 pp.
38
Böhm-Bawerk, E., 1884: The Positive Theory of Capital. New York.
39
Bolin, B., 1994: "Opening Address". IPCC WG III Lead Authors Meeting, Geneva, 8-10 Sept.
40
Bray, M., 1983, "Convergence to Rational Expectations Equilibrium". In: Frydman and Phelps
41
(Eds.), Individual Forecasting and Aggregate Outcomes. Cambridge: Cambridge University Press.
65
I Bray, M., 1987: "Perfect Foresight". The New Palgrave. New York-London: W.W. Norton &
2
Company. pp. 144-148.
2 I Edmonds, J.A., and J.M. Reilly, 1986: Uncertainty in Future Global Energy Use and Fossil Fuel
CO₂ Emissions 1975 to 2075. Prepared for the U.S. Department of Energy.
3 Broome, J., 1992: Counting the Cost of Global Warming. Cambridge: Cambridge University Press.
3 Edmonds, J.A., and J.M. Reilly, 1994: Energy Journal 4: 21.
4 BUWAL/OFEFP, 1994: Le réchauffement planétaire et la Suisse: bases d'une stratégie nationale 1.
5
Global Warming and Switzerland: Foundations for a National Strategy. Beme (Switzerland): Federal
4 5 Eeckhoudt, L., and Ch. Gollier, 1995: Risk-Evaluation, Management and Sharing. Translated by
V. Lambson. New York/London/Toronto/Sysney/Tokyo/Singapore: Harvester Wheatsheaf, 347 pp.
6
Office of Environment, Forests and Landscape (BUWAL/OFEFP), 112 pp.
7 Caselton, B. and W. Luo, 1994: "Dempster-Shafer Theory and Decision Making Under Near-igno-
6 7 Energy Modeling Forum, 1993: Reducing Global Carbon Emissions-Costs and Policy Options.
EMF-12. Stanford, CA: Stanford University.
8
rance". University of British Columbia and British Columbia Hydro. Working Paper, 17 pp.
8 EPA, 1990: Reducing Risk, Relative Risk Reduction Strategies Committee of the US EPA Science
9 Cass, D., G. Chichilnisky, and H.M. Wu, 1992: "Individual Risks and Mutual Insurance".
9
Advisory Board. Washington, D.C.: Environmental Protection Agency.
10
University of Pennsylvania: Department of Economics. Working Paper CARESS. Econometrica,
11
forthcoming.
10 Fankhauser, S., 1994: "Global Warming Damage and the Risk of a Climate Catastrophe". In: Pillet
11
and Gassmann (Eds.), Steps Towards a Decision Making Framework to Address Climate Change.
12 Chichilnisky, G., 1993a: "What is Sustainable Development". Columbia University, New York,
12
Report of the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen
13
N.Y. Research Paper, 19 pp.
13
& Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer Institute, PP. 19-29.
14 Chichilnisky, G., 1993b: "Markets with Endogenous Uncertainty: Theory and Policy". Stanford
14 Faucheux, S. and G. Froger, 1994: "Decision-making Under Environmental Uncertainty". In: Pillet
15
University: Department of Economics. Working Paper.
15
and Gassmann (Eds.), Steps Towards a Decision Making Framework to Address Climate Change.
16
Report of the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen
16 Chichilnisky, G., 1994: "Global Environmental Risks and Financial Instruments". In: Pillet and
17
& Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer Institute, pp. 51-65.
17 Gassmann (Eds.), Steps Towards a Decision Making Framework to Address Climate Change.
18
Report of the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen
18 FCCC, 1992 see UN Framework Convention on Climate Change (UN FCCC).
19
& Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer Institute, pp. 99-109.
19 Fishburn, P.C., 1987, 1990: "Utility Theory and Decision Theory". The New Palgrave. New York-
20 Chichilnisky, G., and G. Heal, 1992: "Financial Markets with Unknown Risks". Columbia
20
London: W.W. Norton & Company. pp. 303-312.
21
Business School. First Boston Working Paper Series.
21 Fisher, 1., 1930: The Theory of Interest. New York: The Macmillan Co. 550 pp.
22 Chichilnisky, G., and G. Heal, 1993: "Global Environmental Risks". Journal of Economic Per-
23
spectives 7 (4): 65-86.
22 Folland, Parker, and Palmer, 1986: "Sahel Rainfall and Worldwide Sea Temperature". Nature
23
320: 602-607.
24 Cline, W.R., 1992: The Economics of Global Warming. Washington, D.C.: Institute for International
25 Economics.
24 Froger, G., and E. Zyla, 1994: "Decision Making for Sustainable Development: Orthodox or System
25
Dynamics Models? Proceedings of the International Symposium on Models of Sustainable Develop-
26 Cline, W.R., 1993: "Greenhouse Policy After Rio: Economics, Science, and Politics". In: Kaya,
26
mens. Paris, March 16-18, 1994. Volume II, pp. 1061-1074.
27 Nakicenovic, Nordhaus, and Thoth (Eds.). Costs, Impacts, and Benefits of CO₂ Mitigation. Procee-
28
dings of a Workshop held on 28-30 September 1992 at IIASA, Laxenburg, Austria. Laxenburg,
27 Gassmann, F., 1993: "Non-linear Dynamical Systems, An Introduction. In: Greppin et al. (Eds.),
29 Austria: IIASA, pp. 41-56.
28
Some Physico-chemical and Mathematical Tools for Understanding of Living Systems. Geneva:
29
University of Geneva.
30 Dansgaard, W. et al., 1993: "Evidence for General Instability of Past Climate from A 250-kyr Ice-
31
core Record." Nature 364: 218-220.
30 Gassmann, F., 1994: "Abrupt Change Scenarios". In: Pillet and Gassmann (Eds.), Steps Towards a
31
Decision Making Framework to Address Climate Change. Report of the IPCC Working Group III
32 Dempster, A.P., 1967: "Upper Lower Probabilities Induced by a Multivalued Mapping". Ann. Math.
32
Writing Team II Montreux Meeting, March 3-6. Würenlingen & Villigen (Switzerland): PSI-Bericht
33
Statist. 37: 355-374.
33
94-10, Paul Scherrer Institute, pp. 3-9.
34 Dowlatabadi, H., and M.G. Morgan, 1992: A Model Framework for Integrated Studies of the
34 Graham, N.E., and W.B. White, 1988: "The El Ni-o Cycle: A Natural Oscillator of the Pacific
35
Climate Problem. Department of Engineering and Public Policy, Carnegie Mellon University, Pitts-
35
Ocean - Atmosphere System". Science 240: 1293-1302.
36 burgh, PA. Draft manuscript, 25 pp.
36 Grinols, E.L., 1985: "Public Investment and Social Risk-sharing". European Economic Review 29:
37 Dowlatabadi, H., and M.G. Morgan, 1993: "A Model Framework for Integrated Studies of the
37
303-21.
38
Climate Problem". Energy Policy 21: 209-221.
38 Haites, E.F., 1994: "Portfolio Analysis Model of Climate Change Decision Making". In: Pillet and
39 Dowlatabadi, H., and M.G. Morgan, 1993: "Integrated Assessment of Climate Change". Science
39 Gassmann (Eds.), Steps Towards a Decision Making Framework to Address Climate Change.
40
259: 1813 & 1932.
40
Report of the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen
41
& Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer Institute, pp. 111-114.
42 Haken, H., 1977: Synergetics. Berlin: Springer Verlag.
1
Hammitt, J.K., R.J. Lempert and M.E. Schlesinger, 1992: "A Sequential-Decision Strategy for
I
Kohlas, J., L. Cardona, R. Haenni, U. Hänni, B. Anrig, and N. Lehmann, 1994, "Defeasible
2
Abating Climate Change". Nature 357: 315-18.
2
Reasoning and Uncertainty Management Systems". University of Fribourg (Switzerland), Institute of
3
Informatics: Research Projects 1993-1994, pp. 11-12.
3
Heap, S.H., M. Hollis, B. Lyons, R. Sugolen, and A. Weale, 1992: The Theory of Choice-A
4
Critical Guide. Oxford, UK and Cambridge, Mass.: Blackwell, 398 pp.
4
Krutilla, J.V., 1967: Conservation Reconsidered. American Economic Review 57: 777-786. Re-
5
printed in Smith (Ed.), Environmental Resources and Applied Welfare Economics-Essays in Honor
5
Hediger, W., 1994: "On the Opportunity Cost of Climate Change and Policy". In: Pillet and Gass-
6
of John V. Krutilla. Washington, D.C.: Resources for the Future, 1988, pp. 263-73.
6
mann (Eds.), Steps Towards a Decision Making Framework to Address Climate Change. Report of
7
the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen &
7
Lind, R.C., K.J. Arrow, G.R. Corey, P. Dasgupta, A.K. Sen, T. Stauffer, J.E. Stiglitz, J.A.
8
Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer Institute, pp. 41-49.
8
Stockfisch, and R. Wilson, 1982: Discounting for Time and Risk in Energy Policy. Washington,
9
D.C.: Resources for the Future, 468 pp.
9
Hope, C.W., J. Anderson, and P. Weinman, 1993: "Policy Analysis of the Greenhouse Effect-
10
An Application of the PAGE Model". Energy Policy 21: 327-338.
10 Lorenz, E.N., 1963: "Deterministic Nonperiodic Flow". J. Atm. Sci. 20: 130-141.
11
Houghton, J.T., G.J. Jenkins, and J.J. Ephraums (Eds.), 1990: Climate Change. the IPCC
11
Luce, R.D. and H. Raiffa, 1957: Games and Decisions. New York: Wiley.
12
Scientific Assessment. Cambridge - New York - Port Chester - Melbourne Sydney: Cambridge
13
University Press, 365 PP.
12
Manne, A.S. and R.G. Richels, 1993: CO2 Hedging Strategies-the Impact of Uncertainty Upon
13
Emissions. Prepared for the OECD/IEA Conference on the Economics of Climate Change, Paris,
14
Hourcade, J.-Ch., 1993: "Economic Issues and Negotiation on Global Environment. Some Lessons
14
June 14-16, 17 PP.
15
from the Recent Experience on Greenhouse Effect". In: Carraro (Ed.), Trade, Innovation,
16
Environment. Dordrecht: Kluwer Academic Publ.
15
Manne, A.S., and R.G. Richels, 1992: Buying Greenhouse Insurance. Cambridge, Mass.: The MIT
16
Press, 182 pp.
17
Hourcade, J.-Ch., and T. Chapuis, 1994: "No-Regret potentials and technical Innovation: A Viabil-
18
ity Approach to Integrative Assessment of Climate Policies". Contribution to the IPCC Working
17
Markandya, A., and D.W. Pierce, 1991: "Development, the Environment, and the Social Rate of
19
Group III Writing Team " Montreux Meeting, March 3-6. Forthcoming in Energy Policy.
18
Discount". The World Bank Research Observer 6 (2): 137-152.
20 IIASA, 1993: Processes of International Negotiation. IIASA's options, June 1993, pp. 4-12.
19
Mégie, A., 1992: "Modélisation prospective, critères de décision et négotiation internationale d'un
20
développement durable: le cas des risques climatiques". Paris: Séminaire du PRISTE-CNRS et du
21
INSEAD/SANDOZ, 1994: Environmental Hazard: Liability of Opportunity? Report of the SANDOZ/
21
Programme ECLAT du Ministère de l'Environnement.
22
INSEAD Colloquium held 29 June 1993. Paris: INSEAD, 35 pp.
22
Meyer, A., and A. Sharan, [undated]: Equity and Survival.-Who Provides Global Benefit; Who
23
IUCC. 1993: "Insurance Against Climate Change? The AOSIS Proposal". Climate Change-Fact
23
Causes Global Disbenefit? A Basis for the Reform of global Institutional Arrangements. London:
24
Shees 238. Geneva: UNEP/WMO, 2 pp.
24
Global Commons Institute, 4 pp.
25
Jevons, W.S., 1871: The Theory of Political Economy. London.
25
MIT, 1986: The MIT Dictionary of Modern Economics. D.W. Pearce (Ed.). Third Edition. Cam-
26
bridge, Mass.: The MIT Press, 462 pp.
26
Johansson, P.-O., 1993: An Introduction to Modern Welfare Economics. Cambridge: Cambridge
27
University Press, 176 pp.
27
Morgan, M.G., and M. Henrion, 1990: Uncertainty. A Guide to Dealing with Uncertainty in Quanti-
28
tative Risk and Policy Analysis. Cambridge New York Melbourne: Cambridge University Press,
28
Joint Climate Project, 1992: Joint Climate Project to Address Decision Makers' Uncertainties.
29
332 pp.
29
Washington, DC: Science and Policy Associates, Inc., 86 pp. + Appendices.
30
National Research Council, 1983: Risk Assessment in the Federal Government: Managing the
30
Jungermann, H., and F. Fleischer, 1988: "As Time Goes By: Psychological Determinants of Time
31
Process. Washington, D.C.: National Academy Press.
31
Preferences". In: Kirsch, Nijkamp, and Zimmermann (Eds.): The Formulation of Time Preferences
32
in a Multidisciplinary Perspective. Aldershot.
32
Nordhaus, W.D., 1982: "How Fast Should We Graze the Global Commons?". American Economic
33
Review, AEA Papers and Proceedings 72 (2): 242-6.
33
Kates, R.W., Hohenemser, C., and J.X. Kasperson (Eds.), 1985: Perilous Progress: Managing
34
the Hazards of Technology. Boulder, Colorado: Westview Press, Inc.
34
Nordhaus, W.D., 1991: "To Slow or not to Slow: The Economics of the Greenhouse Effect". The
35
Economic Journal 101: 920-937.
35
Kaya, Y, N. Nakicenovic, W.D. Nordhaus, F.L. Toth (Eds.), 1993: Costs, Impacts, and
36
Benefits of CO₂ Mitigation.. Proceedings of a Workshop held on 28-30 September 1992 at ILASA,
36
Nordhaus, W.D., 1991: American Economic Review 81: 146.
37
Laxenburg, Austria. Laxenburg, Austria: IIASA, 597 pp.
37
Nordhaus, W.D., 1992: The DICE Model: Background and Structure of a Dynamic Integrated
38
Keeney, R.L. and H. Raiffa, 1976: Decisions with Multiple Objectives: Preferences and Value
38
Climate-Economy Model f the Economics of Global Warming. New Haven: Yale University.
39
Tradeoffs. New York: Wiley.
39
Foundation Discussion papers 1009 & 1019.
40
Kneese, A.V., 1968: Economics and the Quality of the Environment-Some Empirical Experiences.
40 Nordhaus, W.D., and G. Yohe, 1983: "Future Carbon Dioxide Emissions from Fossil Fuels". In:
41
Reprint Number 71. Washington, D.C.: Resources for the Future.
41
Changing Climate. National Research Council. Washington, D.C.: National Academy Press.
1
Parikh, J., P.G. Babu, and K. Parikh, 1994: "Decision Making Framework to Address Climate
I
Pillet, G., W. Hediger, S. Kypreos and C. Corbaz, 1993: The Economics of Global Warming-
2
Change". In: Pillet and Gassmann (Eds.), Steps Towards a Decision Making Framework to Address
2
National and International Climate Policy: The Requisites for Switzerland. Würenlingen & Villigen
3
Climate Change. Report of the IPCC Working Group III Writing Team II Montreux Meeting, March
3
(Switzerland): Paul Scherrer Institut, PSI Bericht Nr. 93-02, 156 pp. Second edition: 1994.
4
3-6. Würenlingen & Villigen (Switzerland): Paul Scherrer Institute, PP. 77-97.
4
Price, C., 1993: Time, Discounting & Value. Cambridge, Mass. and Oxford, UK: Blackwell, 393 pp.
5
Parikh, J., et al., 1992: Consumption patterns: The driving force of environmental stress. IGIDR
6
mimeograph.
5
Prigogine, I., 1980: From Being to Becoming. San Francisco: Freeman.
7
Parikh, J., et al., 1993: Natural resource accounting. IGIDR mimeograph.
6
Querner, I., 1993: An Economic Analysis of Severe Industrial Hazards. Berlin: Physica-Verlag, 271
7
pp.
8
Parikh, J., 1994: "Joint implementation and sharing commitments: A southern perspective", in
9
Nakicenovic et al. (eds.), Integrative assessment of mitigation, impacts and adaptation, IIASA con-
8
Ramsey, F.P., 1928: "A Mathematical Theory of Savings". The Economic Journal 38: 543-565.
10
ference volume.
9
Richels, R., 1994: Decision Making Under Uncertainty and the Global Climate Debate Contribution
11
Parikh, J., and J. Painuly, 1993: "Population, consumption pattern and climate change: A socio-
10
to the IPCC Working Group III Writing Team II Montreux Meeting, March 3-6, 3 pp.
12
economic perspective from South", forthcoming in Ambio.
11
R'os, S. (Ed.), 1994: Decision Theory and Decision Analysis: Trends and Challenges. Boston/
13
Parikh, K., et al., 1988: Towards free trade in agriculture. Dordrecht: Kluwer Academic Press.
12
Dordrecht/London Kluwer Academic Publichers, 294 pp.
14
Parry & Rosenzweig, 1994: "Climate change and global food security". Nature.
13
Rothmans, J., 1990: IMAGE: An Integrated Model to Assess the Greenhouse Effect. Dordrecht:
14
Kluwer Acad. Publ.
15
Paté-Cornell, M.E., 1992: "Risk Analysis and Relevance of Uncertainties in Nuclear Safety
16
Decisions". In: Bailey (Ed.), Public Regulation: New Perspectives on Institutions and Policies.
15
Safra, Z., 1987: "Contingent Commodities". The New Palgrave. New York-London: W.W. Norton
17
Cambridge, Mass.: The MIT Press, pp. 227-253.
16
& Company. pp. 22-24.
18
Pearce, D., E. Barbier, and A. Markandya, 1990: Sustainable Development. Economics and Envi-
17
Sandmo, A, 1972: "Discount Rates for Public Investment under Uncertainty". International Economic
19
ronment in the Third World. Aldershot.
18
Review 13: 287:302.
20
Peck, S.C., 1993: "The Implications of Non-linearities in Global Warming Damage Costs". In: Kaya,
19
Savage, H.A., 1951: "The Theory of Statistical Decisions". American Statistical Association Journal
21
Nakicenovic, Nordhaus, and Thoth (Eds.). Costs, Impacts, and Benefits of CO₂ Mitigation. Procee-
20
46: 55-67.
22
dings of a Workshop held on 28-30 September 1992 at IIASA, Laxenburg (Austria). Laxenburg:
23
IIASA, PP. 209-210.
21
Savage, H.A., 1954: The Foundations of Statistics. New York: Wiley.
24
Peck, S.C., 1994: "Candidate Framework for Analysis of Climate Change Decisions and Uncer-
22
Schelling, .,: 1992: American Economic Review
25
tainties". Communication at the IPCC Working Group III Writing Team " Montreux Meeting, March
26
3-6. Montreux (Switzerland).
23
Schlesinger, M.E. and X. Jiang, 1991: Nature 350: 219-221.
27
Peck, S.C., and T.J. Teisberg, 1993: "Global Warming Uncertainties and the Value of Information:
24
Schmalensee, R., 1993: "Symposium on Global Climate Change". The Economic Journal 38: 543-
28
An Analysis Using CETA". Resource and Energy Economics 15: 71-97.
25
565.
29
Pethig, R. (ed.), 1994: Valuing the Environment: Methodological and Measurement Issues. Amster-
26
Schmalensee, R., 1993: "Symposium on Global Climate Change". Journal of Economic Perspectives
30
dam: Kluwer Academic Publishers.
27
7(4): 3-10
31
Pigou, A.C., 1932: The Economics of Welfare. London.
28
Schmeidler, D. and P. Wakker, 1987: "Expected Utility and Mathematical Expectation". The New
29
Palgrave. New York-London: W.W. Norton & Company. pp. 70-78.
32
Pillet, G., 1993: Economie écologique - Introduction a l'économie de l'environnement et des
33
ressources naturelles. Geneva: Georg Publ. Co., 223 PP.
30
Schneider, S.H., 1983: "CO2, Climate and Society: A Brief Overview". In: Chen, boulding, and
31
Schneider (Eds.), Social Science Research and Climate Change, PP. 9-15.
34
Pillet, G., 1994: "Betting on Climate States-The Impact of Climate Change Prospects on Decision
35
Making under "Strong" Uncertainty and Collective Risks". In: Pillet and Gassmann (Eds.), Steps
32
Schubert, R., 1994: "Climate Change and Discount Rates". In: Pillet and Gassmann (Eds.), Steps
36
Towards a Decision Making Framework to Address Climate Change. Report of the IPCC Working
33
Towards a Decision Making Framework to Address Climate Change. Report of the IPCC Working
37
Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen & Villigen (Switzerland):
34
Group III Writing Team " Montreux Meeting, March 3-6. Würenlingen & Villigen (Switzerland):
38
PSI-Bericht 94-10, Paul Scherrer Institute, pp.
35
PSI-Bericht 94-10, Paul Scherrer Institute, pp. 33-40.
39
Pillet, G., and F. Gassmann (Eds.), 1994: Steps Towards a Decision Making Framework to
36
Shackle, G.L.S., 1967: Décision, déterminisme et temps. Centre d'Econométrie de la Faculté de Droit
40
Address Climate Change. Report of the IPCC Working Group III Writing Team " Montreux
37
et des Sciences Economiques de Paris, Paris: Dunod.
41
Meeting, March 3-6. Würenlingen & Villigen (Switzerland): PSI-Bericht 94-10, Paul Scherrer
42
Institute, pp. 115-137.
38
Shafer, G., 1976: A Mathematical Theory of Evidence. Princeton, N.J.: Princeton University Press.
I
Shlyakhter, A, L.J. Valverde A., Jr., and R. Wilson, 1995: "Integrated Risk Analysis of Global
2
Climate Change". Chemosphere, Forthcoming.
1
Weilenmann, U., 1994: "Insurable Risk Associated to Climate Change". In: Pillet and Gassmann
2
(Eds.), Steps Towards a Decision Making Framework to Address Climate Change. Report of the
3
Simon, H.A., 1957: Models of Man. New York: Wiley.
3
IPCC Working Group III Writing Team II Montreux Meeting, March 3-6. Würenlingen & Villigen
4
(Switzerland): Paul Scherrer Institute, pp. 11-17.
4 Simon, H.A., 1987: "Bounded Rationality". The New Palgrave. New York-London: W.W. Norton &
5
Company, pp. 15-18.
5 Yaari, M., 1987: "The Dual Theory of Choice Under Risk". Economitrica 55: 95-116.
6 Singer, F., R. Revelle, and C. Starr, 1991: "What to do about Greenhouse Warming". Cosmos,
7
PP. 28-33.
8
Sjöstedt, G. (Ed.), 1993: International Environmental Negotiation. Sage Publications.
9 Smith, K., 1991: "Allocating Responsibility for Global Warming: The Natural Debt Index". Ambio 20
10
(2): 95-96.
11
Solow, R.M., 1986: "On the Intergenerational Allocation of Natural Resources". Scandinavian Jour-
12
nal of Economics 88: 141-49.
13
Solow, R.M., 1992: An almost Practical Step toward Sustainability. Washington, D.C.: Resources
14
for the Future.
15
Sugden, R., and A. Williams, 1990: The Principles of Practical Cost-Benefit Analysis. Oxford: Ox-
16
ford University Press (First publication, 1978), 275 pp.
17
Svenson. O., 1991: "The Time dimension in Perception and Communication of Risk". In: Kasperson
18
and Stallen (Eds.): Communicating Risk to the Public. Dordrecht Boston: Kluwer Academic Publ.
19 Tegart McG., W.J., G.W., Sheldon, and D.C. Griffiths (Eds.), 1990: Climate Change. The
20
IPCC Impacts Assessment. WHO/UNEP. Report prepared for IPCC by Working Group II.
21
Published for the Department of the Arts, Sport, the Environment, Tourism and Territories by the
22
Australian Government Publishing Service.
23
Tol, R.S.J, 1994: "The Damage Costs of Climate Change: Towards More Comprehensive
24
Calculations. Environmental and Resource Economics (in press).
25
Tooley, M.J., 1994: "Sea-Level Changes and Impacts". Environmental Hazard: Liability or Opport-
26
unity? Report of the SANDOZ/INSEAD Colloquium held 29 June 1993. Paris: INSEAD Centre for
27
the Mangement of Environmental Resources, pp. 10-16.
28
Ulph A, and D. Ulph, 1994: "Global Warming: Why Irreversibility may not Require Lower Current
29
Emissions of Greenhouse Gases". University of Southampton, UK. Discussion Paper, 36 pp.
30
UN Framework Convention on Climate Change, 1992. Published by IUCC/UNEP
31
(Information Unit on Climate Change), Geneva.
32
Van den Berg, A., L. Hendrickx, and Ch. Vlek, 1993: Concern about Tomorrow? Temporal Dis-
33
counting in Judgemenis of Long-Term Risks. Mimeo.
34
Vercelli, A., 1991: Methodological Foundations of Macroeconomics: Keynes and Lucas. Cambridge:
35
Cambridge University Press.
36
Wald, A., 1950: Statistical Decision Functions. Wiley.
37 Walley, P., 1991: Statistical Reasoning with Imprecise Probabilities. London: Chapman and Hall.
38 Waterstone, 1993: Environmental Management 17(2): 141-152.
A p. 8
There are several ways of responding to any risk: take
actions which reduce the likelihood of the risk occurring or of its
magnitude when it occurs; take actions which reduce the cost
inflicted; or divesting the risk to others through insurance.
There is an analogous portfolio of actions in response to the
threat of global warming: Mitigation actions (reducing emissions)
reduce the chances of significant global warming. Other actions
facilitate the ability to adapt and reduce the magnitude of
impacts. The include redesigning sea walls and the choice of
location for new construction, both of which reduce the costs that
might be incurred, say, from an increase in sea levels; the
development of new seed strains, may reduce the impacts on
agriculture of small changes in weather. The potential adequacy of
insurance markets will be discussed below.
Within each of the broad categories of mitigation and adaption
there are again a wide range of actions which should be included in
a portfolio approach designed to minimize risks. These include
"policy" actions, such as phasing out subsidies to fossil fuels,
that reduce welfare and directly or indirectly increase gas
emissions; R & D, to develop technologies which lower the costs of
reducing emissions as well as which may make it easier to adapt to
any changes that do occur; and international actions, including
joint implementation (agreements for one country to fund emissions
reduction in another), technology transfer and other forms of
international cooperation; policies aimed at enhancing
IPCC, revisions, Chapter 1
February 16, 1995
1
sequestration as well as reducing gross emissions; and policies
aimed at methane and other greenhouse gases, as well as those that
focus on CO2.
IPCC, revisions, Chapter 1
February 16, 1995
2
B, p. 8
1.3.1.2. Risk aversion
There is a general consensus that individuals and societies are
risk averse, that is, they are willing to pay something to reduce
the risks which they face. The amount that they are willing to pay
is called the risk premium (see box) .
IPCC, revisions, Chapter 1
February 16, 1995
3
C. p. 9
Ascertaining the magnitude of the risks--or how they are
affected by any particular action--is often difficult in a dynamic
setting. As we note below, perhaps the most important aspect is
how the particular action affects one's options, the set of
actions, their costs and benefits, which are available in the
future.
We refer to risk reducing expenditures as precautionary
investments. We can think of these expenditures in part as the
purchase of "real" insurance.
IPCC, revisions, Chapter 1
February 16, 1995
4
D. p. 9
Traditional insurance is based on two principles: the pooling of
risks, and the transfer of risks from those that originally
confront it to those who are more willing or better able to bear
it.¹ Insurance markets face problems, though not insurmountable
ones, on both accounts.
Because the risks associated with climate change are
correlated, pooling is less effective. Still, the variety of
impacts (with some areas actually benefitting) noted earlier
implies that the risks are far from perfectly correlated, so that
some degree of risk pooling is possible.
By the same token, risk transfer through an insurance
mechanism is clearly feasible:
¹This "transfer" is facilitated by the dividing of the risk
into small parts, so that any individual only faces a small risk.
The magnitude of risk premium an individual needs to compensate him
for bearing a small risk is very small.
IPCC, revisions, Chapter 1
February 16, 1995
5
E
There are, however, three problems facing insurance markets.
First, ideally, one would like to have mechanisms by which the risk
is transferred from those likely to bear the risks--future
generations - to the current generation. The only way to effect
such intergenerational risk transfers is through governmental
actions. 2 Secondly, losses associated with climate change are
likely to be both correlated and large, compared to losses absorbed
in a single year by the commercial industry, which already has
found itself hard pressed to handle natural disasters. Thirdly,
²This is often used as a basis for arguing that government
should take "real" actions--such as undertaking mitigation
investments; but there are financial mechanisms for effecting
intergenerational risk redistributions. See Stiglitz 1984.
However, without physical investments, there may be problems
assuring that the government will actually undertake the financial
redistributions promised.
IPCC, revisions, Chapter 1
February 16, 1995
6
F
(There are a myriad of other actions, including those listed above
in section and those discussed below, in section
)
Precautionary investments may not only be directed at reducing
the magnitude of the risk of climate change (mitigation) but also
at enhancing the ability of future generations to respond, in two
different ways. The first can be understood by analogy to a "rainy
day fund. " By setting money aside (which grows with the rate of
interest) when things are going well, individuals are better able
to cope with crisis when they occur.
So too, countries by
investing more today in productive assets will be better able to
cope with climate change, should it occur. Precautionary
investment policies entail investing more than one would otherwise
have invested³ The second is the design of investment strategies
(including investment in R & D) which reduce the magnitude of the
economic impacts and which enhance the economy's adapt to those
impacts, should significant climate change occur.
³On the basis of standard intertemporal trade-offs in the
absence of the risk of climate change
IPCC, revisions, Chapter 1
February 16, 1995
7
G, p. 11
The articulation of inappropriate interim goals and commitments
can have a significant adverse impact on overall costs of
mitigation strategies, inducing a significant departure from
optimal sequential decision making strategies. For instance,
commitments to particular levels of emissions in particular years
ignores the important effect that temporary economic disturbances
(such as prices of oil and the level of GDP) may have on emission
levels, and therefore on the costs of attaining such a commitment.
Similarly, focusing on such "flow" commitments ignores the critical
role played by "stock" variables, and diverts attention away from
policies aimed at these variables which are likely to play a more
important role in the long run (including the development of
technologies which reduce emissions and increase energy efficiency.
IPCC, revisions, Chapter 1
February 16, 1995
8
H. p. 11
In a sense, any action today changes the options available in
the future, or more precisely, it changes the consequences to any
future action. Sequential decision making focuses on identifying
how those consequences are affected if one action is taken today
rather than another, or how those consequences are affected if an
action is taken together rather than postpone by a short period of
time.
IPCC, revisions, Chapter 1
February 16, 1995
9
good does not serve as a perfect subsitute for domestic production.²⁷ In Malaysia, some claim that the
local manufacturer of a car has had important spill-overs to the intermediate good firms producing parts,
and that these firms, in turn, have had further benefits to the production of other final goods firms. 28
Returns to scale: a problematic explanation. Not all the arguments put forward by governments as a
rational for industrial policies, however, are persuasive. One argument that particularly influenced the
Japanese economy was that with returns to scale, without government intervention, firms will be too small
scale. Government intervention is required to rationalize the industry. (Thus, the Japanese government
not only condoned the increased concentration in the steel industry in the 1960s {check}, but, in one of
its most famous "mistakes" tried to discourage Honda--at the time a successful manufacturer of
motorcycles-- from entering the automobile market.) The reason that this often cited argument is
unpersuasive is that if there truly were increasing returns to scale, then it would pay a single firm to
increase its production; as it increased its production, its costs would be lowered, and it would then be
able to undercut its rivals. Thus, natural economic forces lead to the rationalization of industries, in the
absence of government intervention.
Returns to scale combined with capital market imperfections. There is a slight variant of this argument
which may have some validity. Increasing returns combined with capital market imperfections may lead
to overly small firms. Firms cannot expand to take advantage of the increasing returns, either because
they cannot get access to capital, or because the only form of capital to which they can get easy access
is credit, and this imposes too high of a risk. Government intervention then can serve to lower costs of
production and increase economic efficiency.
²⁷Note that in traditional planning/Marxian models, there was no rationale for why domestic
production of say steel was required.
28Historically, these spill over effects have been extremely significant. The development of the
machinery and related industries in the Ohio valley in the late nineteenth and early twentieth century, in
conjunction with the development of the automobile industry, provided the intermediate goods which were
necessary inputs in the development of the airplane.
23
Returns to scale and strategic trade policy. Still another variant of this argument relates increasing
returns to strategic trade policy. Historically, the most common version of this argument focuses on
industries with learning by doing. If today's production lowers future marginal costs, then there is a form
of increasing returns much akin to the more familiar static increasing returns. A firm that expands its
production will find that its future production costs are lower. It can undercut its rivals.
Earlier versions of this argument led to the infant industry argument: protection is important so
that the firm can gain the experience required to lower its production costs, so that it can be viable.
More recent discussions discounted this argument, saying that if the firm really were to be profitable in
the long run, it would pay for it to encounter losses today. But again, this counterargument runs into a
fatal fallacy: it is based on the premise that capital markets are perfect. With imperfect capital markets,
firms may not be able to sustain the losses that would enable it to produce at a level at which it would
eventually become profitable. Moreover, if the firm is unable to appropriate all the returns to its
learning, then social returns to production will exceed private returns.
Beyond these arguments lies a third, fundamental argument: whenever there is learning by doing
(or increasing returns more generally) markets will be imperfectly competitive²⁹; and with imperfect
competition, social and private returns often differ. There may be imperfect competition rents earned
by dominant firms in the developed countries. Government policies may be directed at trying to
appropriate some of these rents. For instance, with even small sunk costs and fierce (Bertrand)
competition, the first entrant into a market may have a distinct advantage: it can have sustained profits,
without a threat of entry; since entrants know that, were they to enter, the profits will disappear. There
is an effective entry barrier. (Stiglitz, 1987a). Protection allows the development of a domestic industry,
removing this entry barrier.
Protection generates profits-which form the basis of future investment and risk taking. This
2⁹See Dasgupta and Stiglitz [1987].
24
provides a rationale for the (alleged) practice of the Japanese government in allowing or encouraging
dumping.30 Japanese firms may be effectively protected, by a variety of formal or informal trade
barriers, allowing them to charge high prices domestically. Prices can then be set in the international
(unprotected) market at long run marginal cost, 31 which may be substantially less than current marginal
cost, or average costs. The losses on international production are offset by profits in the protected
market. Competition in the international market ensures that the firm be efficient. This policy ensures
that the firm does not suffer from X inefficiency, at the same time that it ensures that the industry can
benefit from returns to scale, and capture the rents associated with being the technological leader.
Excessive competition. While imperfect competition thus provides one set of rationale for government
industrial policies, excessive competition provides another set, one to which typically economists are not
sympathetic. Since the theoretical ideal is only attained with perfect competition, there is no possibility,
within the standard theory, for excessive competition.
We have already discussed perhaps the two most important rationale for limiting competition.
Increasing the opportunity to avail the economies of scale; and in the absence of strong equity/capital
markets, the enhanced accumulation of equity capital. Government actions to stabilize the market
(recession cartels) lead to some short run inefficiency. But if the elasticity of demand is relatively small,
this inefficiency may be small. But the higher profits lead to greater investment (including in R & D)
and willingness to bear risks, and this effect may well outweigh the short run efficiency effects. 32
30 Again, there is a market failur basis for this policy: with perfect capital markets, firms would have
no trouble raising capital for worthwhile projects. With imperfect capital markets, in particular,
imperfect equity markets, generating capital and divesting risk become important problems. For
information theory explanations of these capital market imperfections, see, for instance, Stiglitz [1992].
³¹It is possible to show, under certain simplifying assumptions (see Spence [1981]) that optimal
production of the firm entails setting the marginal value of output equal to the long run marginal cost.
³²There is another reason of perhaps limited direct relevance to the East Asia economies. In the
absence of perfect competition, competition may not have the positive force widely ascribed to it, of
driving firms to produce better products at lower costs. Rather, it may take a destructive turn: firms
25
Thirdly, to a large extent, government policies were not directed at picking winners, in the
narrow sense of the term. Several of the governments decided to support export oriented industries. This
wining
is picking a develbpment strategy. It does not necessarily entail micro-managing. Even when the
X
government picked an industry, for instance, the banks seem to have had discretion about picking which
firms or projects within that industry to support.
Creating Winners and Making Sure that Those Picked Remain Winners The discussion of "picking
winners" focuses on one of the central problems facing any economic organization, the selection problem.
But it is only one of the central problems, and the emphasis on it has perhaps detracted from the other
major problems facing the developing economies, areas in which government policy may have been
particularly efficacious.
The first of these is "creating" (or finding) winners. The discussion of "picking winners" seems
to envisage a fixed pool of applicants for government support, and the government passively culling
among these for those with the highest long run social returns. At least within East Asia, governments
have taken a far more active role. (See discuss below on government as entrepreneur.)
The second of these is the monitoring problem, of ensuring that those to whom support is given
Seprevin
act "appropriately," e.g. due not siphon off funds for their private use. But again, this provides a too
equy
limited view of the role assumed by the East Asian countries, who not only undertook what might be
marker
viewed as the passive role of monitoring, but also undertook a more active role of trying to ensure the
success of the favored industries and enterprises.
GRWOTH WITH EQUALITY
While industrial policies seek to direct the economies' resource allocations in ways which
maximize growth, income distribution policies seek to promote greater equality. Historically, the
development process has been characterized by marked increased in inequality (the Kuznets curve).
Several theoretical arguments were put forward for this: the massive amounts of capital accumulation
28
equality, but also yielded reasonable economic returns.
(viii) In Malaysia, policies which would in modern parlance be referred to as affirmative
action policies not only were able to draw upon a reservoir of human talent that had not been well used
before, but welded together a nation which had already demonstrated a potential for ethnic strife.
A careful reader may have noticed seemingly contradictory statements concerning the effects of
increased wages on growth. On the one hand, increased wages lead to increased efficiency (efficiency
wage effects) and greater political stability, promoting growth; on the other hand, increased wages
decrease profits, and thus capital accumulation in general and equity accumulation in particular. So long
as wages are below the efficiency wage, there is no contradiction: increases in wages both increase
profits and increase productivity. Whether these countries paid wages substantially higher than efficiency
wages is not clear.
There are, of course, positive feedback relations between growth and equality. The high rates
of growth provided resources which could be used to promote equality; just as the high degree of equality
helped sustain the high rates of growth.
While all of this may seem to be little more than common sense, we need to recall that until the
experience of East Asia, "common sense" suggested quite the contrary: growth produced inequality and
inequality was necessary for growth.
31
I
1.1 INTRODUCTION
I
on the economy, particularly the economy of certain countries or regions.
2
2
3
In recent years, atmospheric emissions of greenhouse gases have risen significantly.
3
Even if the first view is adopted, economics has much to contribute to the discussion, for
4
Concentrations are currently about 25 percent greater than at the beginning of the Industrial
4
the question of cost-effective emissions reductions must still be addressed. If the second
5
Revolution. If current trends continue, concentrations will double from pre-industrial levels
5
view is adopted, economics and cost-benefit analysis will clearly be relevant, both in
6
before the end of the next century and, if unchecked, continue to rise thereafter (IPCC
6
deciding how much mitigation to undertake, and in designing the measures.
7
1990).
7
8
8
This chapter, and others in this Assessment Report, draw on the findings of Working
9
The scientific community has noted the potentially serious effects of increased
9
Groups I and II, and follow the guidelines provided by the Framework Convention. That
10
concentrations. These climatic effects could, in turn, have further effects on the biosphere,
10
Convention makes clear that important questions remain to be addressed by subsequent
11
including an increase in mean global temperature, an increase in sea level, changes in
II
negotiations, including the adequacy of national commitments. This chapter takes the
12
agricultural yields, forest cover and water resources, and a possible increase in storm
12
Framework convention as a political document, recognizing that particular terms of the
13
damage.
13
agreements reached by international negotiation may or may not accord with accepted
14
14
scientific research; that for instance, the agreements may or may not have provided the
15
Increased concentrations of greenhouse gases are the result of fossil fuel burning,
15
basis of a cost-effective approach to mitigating climate change. It is hoped that the findings
16
livestock raising, and other human activities. Concerted action on the part of individuals
16
of this chapter and the assessment report more broadly will enhance an understanding of the
17
and governments will be required to slow the increase in concentrations. Changes in
17
costs and consequences of alternative actions, and will provide the scientific basis for
18
greenhouse gases concentrations and the analysis of the climatic and other physical
18
ongoing negotiations.
19
consequences of those changes lie within the purview of the physical sciences. The role of
19
20
human activity in generating greenhouse gases, the consequences of those changes for
20
1.2 FEATURES OF CLIMATE CHANGE
21
humans, and possible responses, lie within the purview of the social sciences.
21
22
22
Climate change could impose a variety of impacts on society. IPCC Working Group II
23
Climate change impacts are likely to vary dramatically from country to country. A
23
analyses these impacts in detail. They include effects on agriculture¹, forests', water
24
warmer climate could benefit sectors of the economies of some mid- and high-latitude
24
resources', the costs of heating and cooling⁴, the impact of a rising sea level rise on small
25
countries. At the same time, a rising sea level and the possibility of increased storm surges
25
island states and low-lying coastal areas⁵, and a possible increase in storm damage.
26
could threaten the survival of some small island states and coastal areas, and could increase
26
Although most attention to date has focused on negative impacts, some impacts will be
27
the risk of midcontinent drought and desertification for inland areas on the periphery of
27
positive. Beyond these tangible impacts are a variety of intangible impacts including
28
deserts.
28
damages to existing ecosystems and the threat of species losses'.
29
29
30
Within the past decade, a consensus has emerged on some key issues in the economics
30
Climate change presents the analyst with a set of formidable complications: large
31
of climate change. This report describes areas of consensus, as well as areas of
31
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
32
disagreement, the sources of disagreement, and further research that could narrow the range
32
long time lags between emissions and effects, an irreducibly global problem, wide regional
33
of disagreement. This chapter frames the issue of climate change largely from the point of
33
variations, and multiple greenhouse gases of concern.
34
view of economics but also from that of other social sciences, introducing the more detailed
34
35
discussions in the chapters to follow.
35
Large uncertainties. While natural scientists agree that greenhouse gas emissions are
36
36
rising, they do not agree about the mechanism linking concentrations, and temperature.
37
The commitment of resources to mitigate climate change may rest on one of two
37
Further, although natural scientists agree that some warming will occur, they do not agree
38
arguments. The first arises from fundamental values, the second from decision analysis:
38
on the speed of change, or the ultimate amount of change (IPCC 1992, 1994). In addition,
39
39
social scientists do not agree on the size of the behavioral responses or economic effects
40
1. We have only one planet. Some changes are largely irreversible, and may occur
40
that would follow, or on the effect of these changes on well-being.
41
rapidly. Prudence calls for avoiding a large-scale experiment with the planet. Thus,
41
42
avoiding climate change lies beyond normal economic calculus.
42
Nonlinearities and irreversibilities. Nonlinearities occur when changes in one variable
43
43
cause a more than proportionate impact on another variable. For example, some have
44
2. The potential exists for sudden, largely irreversible non-linearities with major effects
44
suggested that even a modest increase in atmospheric greenhouse gas concentrations could,
3
4
I
SUMMARY
I
Issues of efficiency and equity can largely be separated. The Framework Convention on
2
2
Climate Change calls on all parties to implement cost-effective measures for abatement,
3
Climate change presents the analyst with a set of formidable complications: large
3
enhancement of sinks, and adaptation. The Framework Convention also explicitly requires
4
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
4
an equitable sharing of the burdens of response, recognizing the common but differentiated
5
long time lags between emissions and effects, an irreducibly global problem, wide regional
5
responsibilities of the parties. Different countries will be affected differently by climate
6
variation, and multiple greenhouse gases of concern. The risks of climate change are
6
change and by policy responses to it. The South is more likely to be adversely affected
7
highly asymmetrical, with a large probability of a small loss, and a small probability of a
7
than the North; moreover, developing countries often lack the financial and technical
8
large loss. Even in the presence of significant uncertainty, this asymmetry, plus the
8
resources to respond. The Framework Convention does not, however, include a formula for
9
principles of risk aversion and portfolio balancing provide the rationale for going beyond
9
sharing the costs of addressing climate change.
10
no-regrets policies to those that incur net costs.
10
11
11
Efficiency requires that emission reductions occur where their cost is lowest,
12
The atmosphere is an international public good, in that all countries benefit from each
12
irrespective of who bears the financial responsibility. Efficiency calls for removing energy
13
country's reduction in greenhouse emissions; greenhouse gases are an international
13
subsidies, reforming and clarifying property rights that affect energy use and carbon
14
externality, in that emissions by one country affect all other countries to the same extent.
14
storage, and reducing non-greenhouse externalities that have the side benefit of reducing
15
15
greenhouse emissions. Efficiency may also be promoted, and greenhouse emissions
16
Both public goods and externalities require a legal framework within which the
16
reduced, by better information dissemination and by addressing capital market imperfections
17
problems they pose can be addressed. No such legal framework now exists for global
17
that inhibit the adoption of energy-efficient technology. Dynamic analysis indicates large
18
climate change. Mechanisms for control of international public goods may include the
18
potential gains from flexibility in timing of greenhouse reductions to allow for the
19
definition of property rights, the definition of limits to emissions and a consensus for
19
economical turnover of capital stock, and to allow time for the development of low-cost
20
distributing the same in a fair and equitable manner. If, on the other hand, each agent acts
20
substitutes. Policies that promote efficiency by requiring nations to face the full costs of
21
in its individual interest, the result will be too little of the public good and too much of the
21
their actions will also address equity concerns.
22
externality.
22
23
23
Efficiency also calls for international mechanisms such as joint implementation and
24
Climate change demands a decision process that is sequential, can respond to new
24
coordinated economic instruments. Coordinated carbon taxes and tradable carbon emission
25
information with mid-course corrections, and can include insurance, hedging, and the option
25
permits can correct the market failure associated with greenhouse emissions.
26
value of alternative courses of action. The challenge today is to identify short-term
27
strategies in the face of long-term uncertainty. The question is not, what is the best course
28
over the next 100 years, but rather, what is the best course for the next few years, knowing
29
that a prudent hedging strategy sill allow time to learn and change course.
30
31
Policy measures to reduce risks to future generations include 1) immediate reductions
32
in emissions; 2) R&D on new supply and conservation technologies; 3) continued research
33
on how much change is likely and what its effects will be; and 4) investments to assist in
34
adaptation if significant climate change occurs. A well-chosen portfolio of policies will
35
yield greater benefits for a given cost than any one option undertaken by itself. Striking the
36
appropriate balance requires taking into account costs, benefits, and risks.
37
38
In an interrelated global economic system, an attempt to reduce greenhouse gas
39
emissions in one region or one sector of the economy may be offset by increases in other
40
regions or sectors. This may occur through a) the loss of comparative advantage in the
41
carbon-intensive sectors of the regions that limit emissions; b) the relocation of industries;
42
or c) changes in world energy prices and the resulting shift in consumption. Any control
43
strategy must account for these global effects.
44
1
2
Chapter 1
Introduction: Scope of the Assessment
Lead Authors:
J. Goldemberg, R. Squitieri, J. Stiglitz, A. Amano, X. Shaoxiong, R. Saha
January, 1995
I
INTRODUCTION: SCOPE OF THE ASSESSMENT
I
1.5.3 Market Failures and Government Responses
29
2
2
1.5.3.1
Revising national accounts
31
3
SUMMARY
I
3
1.5.4
Innovation
32
4
1.1
INTRODUCTION
3
4
1.5.5
Carbon Taxes and Tradable Permits.
33
5
5
1.5.5.1
A Double Dividend?
35
6
1.2
FEATURES OF CLIMATE CHANGE
4
6
1.5.5.2
Energy Taxes
35
7
7
1.5.5.3
Tradeable Permit Markets.
36
8
1.3
CONTRIBUTION OF ECONOMICS
7
8
1.5.5.4
Combining taxes with tradeable permits
36
9
1.3.1
Risk
7
9
1.5.5.5
Intertemporal patterns of taxation
36
10
1.3.1.1
Portfolio Theory
8
10
1.5.6 Regulatory Approaches
37
11
1.3.1.2
Insurance
8
11
12
1.3.1.3
Precautionary investments
10
12
1.6
SUSTAINABLE DEVELOPMENT
38
13
1.3.2 Sequential Decision Making
11
13
1.6.1 The Economic Concept of Sustainable Development
38
14
1.3.2.1
Value of information
11
14
1.6.2 Implications of Sustainable Development for Developing
15
1.3.2.2.
Option value
11
15
Countries
39
16
1.3.3
Dynamics
12
16
17
1.3.3.1
Kaya Identity
12
17
1.7
CONCLUSIONS
40
18
1.3.3.2
Non-renewable resources, backstop technologies, and
18
19
emission reduction strategies
13
19
1.8
ENDNOTES
42
20
1.3.4 International Public Goods
14
20
21
1.3.4.1
Property rights
15
21
57
22
1.3.4.2
Paying For An International Public Good: Principles
22
23
and Approaches
15
23
1.9
REFERENCES
58
24
1.3.4.4
Enforcement
17
24
25
1.3.4.5
Knowledge
18
25
26
1.3.5
Efficiency
18
26
27
1.3.5.1
Bankable permits
19
28
1.3.5.2
Exchange/risk efficiency.
19
29
1.3.5.3
Comprehensiveness
19
30
1.3.6 General Equilibrium
20
31
1.3.6.1
Intertemporal Substitution
20
32
33
1.4
EQUITY
21
34
1.4.1
General Issues
21
35
1.4.1
Intergenerational Equity
24
36
1.4.2
Within-country Equity
25
37
38
1.5
ECONOMICS OF POLICY ACTIONS
25
39
1.5.1
Zero-cost Options
26
40
1.5.2
Government Reform
26
41
1.5.2.1
Removing Energy Subsidies
26
42
1.5.2.2
Property Rights Reform
26
43
1.5.2.3
Administrative Reforms
27
44
1.5.2.4
Regulating Non-greenhouse Externalities
27
45
1.5.2.5
Special Problems of Economies in Transition
27
46
1.5.2.6
Examples of Efficiency-Increasing Policies
28
I
parties always arrive at efficient bargaining solutions, critics point out that the outcomes of bargaining models
I
simpler emissions permit system (EPS) issues permits on the basis of source emissions and ignores what effects
2
with incomplete information often entail large inefficiencies.
2
those emissions have on the receptor points. Within a given region or zone, the polluter would have only one
3
market to deal with and one price. Finally, there is the pollution offset (PO) system wherein the permits are
3
86. Matters are more complicated, since the patent does not reward the innovator with his marginal contribution--
4
defined in terms of emissions, trade takes place within a defined zone. However, the standard has to be met at
4
the increase in the present discounted value of benefits as a result of the innovation occurring earlier than it
5
all receptor points. The exchange value of the permits is then determined by the effects of the pollutants at the
5
otherwise would have occurred. For a fuller discussion, see Stiglitz [1994], Dasgupta and Stiglitz [1980a, 1980b],
6
receptor points. The PO system thus combines characteristics of the EPS and the APS. (Pearce and Turner
6
Barzel [ ].
7
1990) These distinctions are of limited relevance for greenhouse gases, where what is of concern is global
8
emission levels. The specific location of the emissions is of no concern.
7
87.
If less developed countries fail to implement fully a set of corrective taxes or tradeable permits, or if less
9
8
developed countries fail to adopt and enforce effectively intellectual property rights, there will be insufficient
9
incentives to produce energy and emission savings innovations, particularly those appropriate for the level of
10
90. The choice between taxes and tradeable permits depends on the objectives of the policy maker and nature of
10
technological knowledge, human capital, and factor prices in those countries. If less developed
11
the uncertainty about the marginal cost and marginal benefit curves for carbon emission reductions (Weitzman,
II
countries do take these actions, there is concern that they will result in higher prices of innovations, and thus the
12
1974). Theory tells us that if the nature of the curves is known with very little certainty, but the marginal cost
12
pace of adoption will be retarded. An effective form of aid, targeted to reducing greenhouse gas emissions, may
13
curve is known to be relatively steeper (i.e. a change in the level of pollution allowed brings about a greater
13
be subsidies directed at producing appropriate energy saving and emission reducing technologies for LDCs.
14
change in the marginal costs of mitigation compared to the marginal benefits) then taxes should be the policy of
14
15
choice. This is because, in this case, an erroneous estimation of the optimal tax rate will lead to a relatively small
16
deviation from the optimal pollution level. On the other hand, an erroneous estimation of the optimal level of
15
88. Edmonds et al. (1994) has studied the importance of available advanced energy technologies such as those
17
total emissions in a permit scheme will lead to a relatively large deviation from the optimal cost of the permits.
16
proposed by Johannsen (1993). Edmonds et al. use the Edmonds-Reilly-Barnes model for energy related
18
17
greenhouse gas emissions; the MAGICC model for atmospheric composition, climate response, and sea level rise;
19
If the marginal benefit curve is known to be relatively steeper than the marginal cost curve, however, tradable
18
the IPCC scenario IS92a (IPCC, 1992) as the reference base case and five alternate energy scenarios that are far
20
permits are the better option. Here, an erroneous estimation of the optimal tax rate will lead to a relatively large
19
more advanced over today's energy supply and transformation technologies. The five energy scenarios are:
21
deviation from the optimal level of emissions while an erroneous estimation of the optimal level of emissions in
20
22
a permit strategy will lead to a relatively small deviation from the optimal cost of the permits.
21
a. advanced fossil fuel technologies
23
22
b. advanced liquified hydrogen fuel cells
24
In the case of greenhouse gas emissions, the time horizon for adjustment is sufficiently long that many of these
23
C. advanced hydrogen fuel cells without liquified hydrogen
25
uncertainties become less important. If the tax rate initially chosen yields too high a level of emissions in one
24
d. low cost biomass
26
year, it can be increased, and the net impact of the erroneous initial estimate on global warming (or the total cost
25
e. accelerated rate of exogenous end-use energy intensity improvement
27
of achieving a given level of atmospheric concentration) will be negligible. In any case, as the earlier discussion
26
28
on sequential decision making has emphasized, there is likely to have to be continued revisions in either tax rates
27
Combined, the energy technologies reduce annual emissions from fossil fuel use to levels that stabilize
29
or permit levels.
28
atmospheric concentrations below 550 ppmv (i.e. double the concentration prior to the Industrial Revolution).
30
29
The tax rate, assumed to apply globally, used was the marginal cost of stabilizing fossil fuel carbon emissions
31
30
in the reference case. With values reflected for only carbon dioxide emission reductions, the cost of global
32
Still, there is some argument that the required adjustments under a permit scheme may be less burdensome.
31
emission reductions grow from approximately $35 (US) Billion in 2005 to $230 (US) billion per year in the year
33
(Tietenberg 1992) For instance, if the authority feels that the old standard needs some tightening they may enter
32
2095. With advanced fossil fuels, low cost solar electric power, low cost fuel cell vehicles, the present discounted
34
the market themselves and buy some of the permits, holding them out of the market.
33
value of adding low cost biomass fuel to the energy technology bundle is almost half a trillion dollars (US). The
35
34
present discounted value of the advanced energy technologies taken together is $1.8 trillion (US).
36
There must be effective, competitive markets in tradeable permits, if such schemes are to achieve efficient
35
37
outcomes. There are transactions costs of running such schemes, just as there are transactions costs associated
36
The introduction of advanced biomass energy production technology was found to play a key role in reducing
38
with collecting tax revenues. Whether transactions costs gives one system a decided advantage over the other
37
emissions. Biomass energy at $2.00/GJ growing to become the core energy supply technology by 2050 could
39
is not clear. There seems to be no compelling reason to believe that good markets in tradeable permits would
38
significantly reduce emissions highlighting the potential role of technology development and deployment relative
40
not develop.
39
to that of fiscal and regulatory intervention.
41
40
42
41
These results should be viewed as illustrative rather than predictive. In this analysis, the gains from introduction
42
and deployment of advanced energy technologies is dependent on the order of technologies evaluated in the study.
43
91 is possible to design allocations of trading permits which (i) on average, impose no net burden on developing
43
44
countries (thus conforming to the ability to pay principle); (ii) provide those economies which are growing faster
44
45
per capita with commensurately greater permits, thus imposing no net drag on economic growth, provided the
46
economy exhibits an increase in fuel efficiency at least equal to the average of fast growing LDCs; and (iii)
45
89. The literature has identified three types of permit systems. The ambient permit system (APS) works on the
47
rewards those economies which are able to reduce greenhouse gas emissions faster than benchmark, either through
46
basis of permits defined according to exposure at the receptor points. Each polluter, then, may face quite
48
greater control of population growth, through larger increases in energy efficiency, or through switching from
47
complex markets different permit markets according to different receptor points, and hence different prices. The
49
higher to lower carbon fuels.
50
53
54
I
The extent to which individual circumstances of countries should be taken into account in setting benchmarks
I
For a 45% reduction in baseline emissions by 2020, the required tax would be in the range of $150-$325 per ton
2
remains a question for international negotiations; to the extent that high emissions is due to natural endowments
2
of carbon and the cost might be in the range of 1.5%-2.9% of world GDP. A 70% reduction in baseline
3
(e.g. the availability of coal rather than natural gas as a source of energy), a persuasive case can be made for
3
emissions by 2050 could require a tax between $230 and $880 and a loss in world GDP of 2.4%-3.8% (Dean
4
benchmarks to reflect initial emission levels. To the extent that high emissions are due to inappropriate energy
4
1994).
5
pricing policies, the case that benchmarks should reflect initial emission levels is far more tenuous.
5
6
The required carbon taxes and associated costs vary significantly across regions in all of the models. This
6
92. Similarly, some developing countries have asked, should the North be given higher levels of permits, simply
7
indicates that the same proportional reductions in emissions across all regions would give rise to very different
7
because it has, in the past, been the chief source of greenhouse gases?
8
costs in different regions and would thus be globally inefficient with great potential for savings in the global
9
cost of reducing emissions through the use of emission trading between countries or regions (see section 1.5.2.1)
8
93. For a discussion of the polluter pays principle, see above.
10
or a global carbon tax.
11
9
94. Standard tradeable permit schemes essentially take the revenue from a carbon tax, and distributes it to current
12
Three insights emerging from the OECD study are (Dean, 1994):
10
user emitters, rather than using the revenue to reduce other taxes. An alternative to these standard schemes is
13
a. Small amounts of emissions reduction can probably be achieved with low taxes;
11
for the government to auction off the tradeable permits.
14
b. Large reductions can only be achieved at high tax rates (i.e. marginal reduction costs rise with emissions
12
15
reductions).
13
If taxing carbon leads to reduced labor supply or reduced savings, then government revenues from wage or capital
16
C. Carbon-free backstop technologies are likely to slow the rise of the carbon tax, or halt it altogether, if they
14
taxes may be reduced, more than offsetting the direct revenue gain from the carbon tax. Cross elasticities of this
17
are available at constant marginal cost.
15
magnitude are unlikely, though any such cross elasticity will reduce the net gain from the carbon tax. The
18
16
magnitude of the double dividend has been the subject of some dispute, with Goulder [ ] and taking opposite
19
Energy Modeling Forum Project 12 (Impact of Carbon Emission Control Strategies)
17
views.
20
A recent study at Stanford, the Energy Modelling Forum Project 12, examines the cost of reducing CO2
18
21
emissions (Energy Modeling Forum, 1993). A diverse group of economic models, employing common
22
assumptions for selected numerical inputs, were used to analyze a standardized set of emission reduction
19
95. Two main studies provide insights into the root of the variance in estimates of the economic effects of carbon
23
scenarios. In all, 14 top-down models participated in the study.
20
taxes: The Energy Modeling Forum Study (12) and the OECD comparison project. In each case, sophisticated
24
21
sensitivity analyses are run by standardizing key economic assumptions along with use of common reference case
25
The EMF model comparison provides the most comprehensive application of top-down methodologies to date.
22
scenarios of reductions. The magnitude of the effect on economic growth will depend both on assumptions
26
The study addresses a wide range of policy questions. How large are emissions likely to grow in the absence
23
concerning the effect of carbon taxes on savings and labor supply, and the induced investment to offset the higher
27
of controls? How much market intervention will be required to meet alternative targets? What will be the price
24
energy prices. If higher energy prices do not lead to much capital substitution and if the cross elasticity with
28
tag? In exploring economic costs, the modelers were asked to examine the impacts of timing, research and
25
savings and labor are low then the likely effect on economic growth will be small.
29
development, and revenue recycling.
26
30
27
31
The EMF exercise provides a wealth of useful information for policy making. Although the focus was primarily
28
OECD Project on Economy-Wide Cost Estimates of Carbon Taxes:
32
on the U.S., many of the insights are applicable to developed countries in general.
29
An OECD model comparisons project was conducted to compare economy wide estimates of the effects of carbon
33
30
taxes. Time horizons as well as the key economic assumptions on growth, population and resource prices, and
34
In selecting parameters for standardization, the EMF study focused on what were felt to be the most influential
31
the reduction scenarios for six global models were standardized. The global models compared were the GREEN
35
determinants of mitigation costs. These included: GDP, population, the fossil-fuel resource base, and the cost
32
model, the IEA model and four North American models [Edmonds-Reilly Model (ERM), Global 2100 of Alan
36
and availability of long-term supply options. In addition, although the EMF models differed considerably in their
33
Manne and Rich Richels (MR), the Carbon Rights Trade Model (CRTM) of Tom Rutherford, and the Whalley-
37
technology representation, the study attempted to impose uniformity with regard to world oil prices, the oil and
34
Wigle model] (Dean 1994).
38
gas resource base, and the cost of backstop technologies. For its reference case, EMF adopted the average of the
35
39
IPCC high and low economic growth cases (IPCC, 1990c). Also for consistency with the IPCC, the study
36
There is significant variation in tax rates and costs for the same amount of emissions reduction among models
40
adopted the population growth projections based on Zachariah and Vu.
37
due to differing assumptions regarding several key considerations. Several factors explain the differences between
41
38
model results. The most important factors are:
42
The modelers generally used taxes based on the carbon content of the fossil fuels in order to achieve a prescribed
39
43
emissions reduction. The magnitude of the tax provides a rough estimate of the degree of market intervention
40
a. The degree of substitution between fuels the ease with which producers and consumers can switch from
44
that would be required to achieve the carbon emissions target. Estimates range from $20 to $140 per ton for the
41
high-carbon content fuels to low-carbon content fuels;
45
carbon taxes required to hold emissions at 1990 levels in 2010. Estimates of the carbon taxes required to reduce
42
b. Expectations about future energy prices and taxes;
46
emissions by 20% below 1990 levels in 2010 range from $50 to $330 per ton.
43
C. The speed of emissions reduction;
47
44
d. The way in which revenue is recycled;
48
Two parameters are particularly important in explaining the differences in tax projections: the price elasticity of
45
e. The treatment of the removal of energy subsidies; and
49
energy demand and the speed with which the capital stock adjusts to higher energy prices. Neither were
46
f. Assumptions regarding backstop technology and a host of other technical and economic factors.
50
controlled in the EMF experiments. Those models using lower price elasticities required higher taxes to achieve
47
51
the same emissions goal. Those models which assumed greater malleability of capital required lower taxes.
48
Because of the varying approach to these questions, the range of estimated tax rates and costs are quite wide.
55
56
I
I
1.9
REFERENCES
2
2
96. Rapid capital stock retirement may add to the cost of immediate CO2 reductions. Also, reductions may be
3
Abreu (**date**)
3
cheaper later because of technical changes in the intervening years.
4
4
97. Indeed, the "ratchet effect," a commonly observed phenomena in the command and control economies
5
Arrow (**date)
5
(Stiglitz, 1975, Weitzman, 19 ) leads firms to just satisfy the target.
6
7
Ausubel, J.H. 1990, Conventional Wisdom About Impacts of Global Change. Rockefeller
6
98. Two examples from the United States Clean Air Act: 1) the requirement that coal-burning power plants
8
University, New York. ?pp.
7
install stack-gas scrubbers even if they burn low-sulfur Western coal; 2) the mandates for ethanol in reformulated
9
8
gasoline.
10
Baumol (**date**)
9
World Commission on Environment and Development, Our Common Future, Oxford: Oxford University Press,
II
10
1987.
12
Barzel (**date**)
13
11
100.lt has been argued that new species and varieties produced by genetic engineering may be able to more than
14
Bongaarts, J. 1994. "Population Policy Options in the Developing World," Science (263),
12
offset the loss in genetic variability from climate impacts, particularly because of the increasing capacity to direct
15
11 February 1994, pp. 771-776.
13
those mutations toward socially desirable objectives.
16
14
101. Other views represented in the debate on sustainable development but not generally accepted in the
17
World Commission on Environment and Development 1987. Our Common Future.
15
economics profession include those of Daly (1977, 1980) and Daly and Cobb (1991), and those of the new field
18
London: Oxford University Press.
16
of ecological economics (Costanza 1991).
19
20
Burniaux, J.M., Martin, J.P. and Oliviera Martins, J. 1992. 'The effects of existing
17
102.Tariq Osman Hyder, "Climate Negotiations: the North/South Perspective," in Mintzer, ed. Hyder has headed
21
distortions in energy markets on the costs of policies to reduce CO2 emissions: evidence
18
the Pakistani delegation at meetings of the INC, and GEF (Global Environment Facility).
22
from GREEN.' OECD Economic Studies No. 19
19
103. "Principles of the G-77 and China on the Climate Convention," cited in Tariq Osman Hyder, "Climate
23
20
Negotiations, the North/South Perspective," in I. Mintzer, ed., Confronting Climate Change: Risks, Implications,
24
Chao, Hung-po, 1992. "Managing the Risk of Global Climate Catastrophe," Electric Power
21
and Responses, Cambridge: Cambridge University Press, 1992.
25
Research Insitute.
26
27
Cline, W.R. 1992. The Economics of Global Warming. Institute for International
28
Economics, Washington, D.C.
29
30
Coase (1990)
31
32
Coase (1960)
33
34
Dasgupta, Gilbert, andStiglitz (**date**)
35
36
Dasgupta, and Stiglitz (1980a, 1980b)
37
38
Dasgupta, P. and K.G. Maler 1990. The Environment and Emerging Developing Issues,"
39
Proceedings of the World Bank Annual Conference on Development Economics, pp.
40
101-131.
41
42
Dean, A. 1994. 'The Effectiveness of Carbon Taxes at the International Level.' In:
43
Climate Change: Policy Instruments and their Implications: Proceedings of the Tsukuba
44
Workshop of IPCC Working Group III, 17-20 January, 1994. pp.46-59.
57
58
I
Dixon et al. (1994)
I
Hannaway (**date**)
2
2
3
Edgeworth, Mathematical Physics,
3
Heal and Chichinisky (**date**)
4
4
5
Edmonds, J., Wise, M. and MacCracken, C. 1994. 'Advanced Energy Technologies and
5
Hoffman, J.S., Keyes, D. and Titus, J.G. 1983. Projecting Future Sea Level Rise:
6
Climate Change: An Analysis Using the Global Change Assessment Model (GCAM).'
6
Methodology, Estimates to the Year 2100, and Research Needs. US GPO 055-000-0236-3.
7
Global Environmental Change Program, Pacific Northwest Laboratory. Wahington D.C.
7
GPO, Washington, D.C.
8
Draft.
8
9
9
Hoffman, J.S., Wells, J.B. and Titus J.G. 1986. 'Future global warming and sea-level rise.'
10
Energy Modeling Forum 1993. EMF 12 Global Climate Change: Impacts of Greenhouse
10
In: Iceland Coastal and River Symposium. G. Sigbjarnason (ed) National Energy Authority,
11
Gas Control Strategies. Palo Alto, Ca.
II
Reykjavik.
12
12
13
Environmental Protection Agency (EPA) 1989. 'Sea level rise.' In Report to Congress
13
Hotelling, H. 1931. "The Economics of Exhaustible Resources," Journal of Political
14
on the Potential Effects of Global Climate Change on the United States.
14
Economy.
15
15
16
Fair (1970)
16
Interamerican Development Bank (**date**)
17
17
18
Fankhauser (**date**)
18
Intergovernmental Panel on Climate Change (1994)
19
19
20
Farrell (**date**)
20
Intergovernmental Panel on Climate Change 1992. The Supplemental Report to the
21
21
IPCC Scientific Assessment. Cambridge University Press, Cambridge. 200 pp.
22
Gleick, P.H. 1987, 'Regional Hydrologic Consequences of Increases in Atmospheric CO2
22
23
and Other Trace Gases'. Climatic Change 10, pp. 137-61.
23
Intergovernmental Panel on Climate Change 1990a, Climate Change: The IPCC
24
24
Scientific Assessment. Report Prepared for IPCC by Working Group I, World
25
Goulder (**date**)
25
Meteorological Organization and United Nations Environment Programme, New York.
26
26
27
Greenwald and Stiglitz (1986, 1988)
27
Intergovernmental Panel on Climate Change 1990b, Climate Change: The IPCC
28
28
Impacts Assessment. Report Prepared for IPCC by Working Group 1, World Meteorological
29
Greenwald, Stiglitz, and Weiss (1984)
29
Organization and United Nations Environment Programme, New York.
30
30
31
Grossman (1981)
31
Jaffe and Stiglitz (**date**)
32
32
33
Grubb, M. et al. 1993. The Costs of Limiting Fossil-Fuel CO₂ Emissions: A Survey and
33
Johannsen (1993)
34
Analysis.' Annual Review of Energy and the Environment 18. pp.397-478.
34
35
35
36
Grubb, M., Minh Ha Duong and Thierry Chapuis 1993. "Optimising Climate Change
36
Kalkstein, L.S. and Davis, R.E. 1989, 'Weather and Human Mortality: an Evaluation of
37
Abatement Responses: on inertia and induced technology development," IIASA Workshop
37
demographic and interregional responses in the United States'. Annals of the Association of
38
on Integrative Assessment of Mitigation, Impacts, and Adaptation to Climate Change,
38
American Geographers 79, pp. 44-64.
39
IIASA, October 1993.
39
40
40
Kane, S., Reilly, J. and Tobey, J. 1992, 'An Empirical Study of the Economic Effects of
41
Grubb, M., Thierry Chapuis and Minh Ha Duong 1995. "The Economics of Climate
41
Climate Change on World Agriculture'. Climatic Change 21, pp. 17-35.
42
Change: Implications of adaptability and inertia for optimal climate policy," forthcoming.
42
43
43
Kaya (1989)
44
Hall et al. (1993)
44
59
60
I
Kononov (1993)
I
Mendelsohn, R. 1993. Global Climate Change Impacts on Coastal Resources. Summary
2
2
Workshop Report to EPRI and NOAA.
3
Kopp, R.J. and Smith, V.K. 1993. Valuing Natural Assets: The Economics of Natural
3
4
Resource Damage Assessment. Resources for the Future, Washington D.C.
4
Meyers and Sathaye, J. 1988. Electricity in Developing Countries: Trends in Supply and
5
5
Use Since 1970. Lawrence Berkeley Laboratory, LBL-2166.
6
Larson, E.D., Ross, M.H., and Williams R.H., "Beyond the Era of Materials," Scientific
6
7
American, vol. 254, no. 6, pp.34-41, June 1986
7
Mirrlees (1971)
B
8
9
Leemans, Rik and Solomon, A. 'Modeling the Potential Change in Yield and Distribution
9
Myers and Maljuf (1984)
10
of the Earth's Crops Under a Warmed Climate'. Climate Research 3, pp. 79-96.
10
11
11
Nash (**date**)
12
Leemans, Rik 1992, 'Modelling Ecological and Agricultural Impacts of Global Change on
12
13
a Global Scale'. Journal of Scientific & Industrial Research 51, pp. 709-724.
13
Nordhaus, W.D. 1991. 'To Slow or Not to Slow: The Economics of the Greenhouse
14
14
Effect.' The Economic Journal 101, pp. 920-937.
15
Leibenstein (**date**)
15
16
16
Nordhaus, W.D. 1993, Managing the Global Commons: The Economics of Climate
17
Lind, R.C. et al. 1982. Discounting for Time and Risk in Energy Policy. Resources for
17
Change. MIT Press, place? ??pp.
18
the Future, Washington D.C.
18
19
19
Orians, G., G. Brown, W. Kunin, and J. Swierzbinski. The Preservation and Valuation of
20
Lind, Robert C. 1994, "Intergenerational Equity, Discounting, and the Role of Cost-Benefit
20
Biological Resources. Seattle: University of Washington Press, 1985.
21
Analysis in Evaluating Global Climate Policy," in N. Nakicenovic, W.D. Nordhaus, R.
21
22
Richels, and F.L. Toth, eds., Interactive Assessment of Mitigation, Impacts, and Adaptation
22
Pachauri and Khanna (1993)
23
to Climate Change: Proceedings of a Workshop held on 13-15 October 1993 at the Institute
23
24
for Applied Systems Analysis, Laxenburg, Austria.
24
Pearce, D.W. and R.K. Turner 1990. Economics of Natural Resources and the
25
25
Environment, Baltimore: Johns Hopkins University Press.
26
Lovelock, J. 1979. Gaia: A New Look at Life on Earth London: Oxford University Press.
26
27
27
Pearce, D.W., E. Barbier, and A. Markandya 1989. Economics and Environment in the
28
Maler, M.G. 1991. "National Accounts and Environmental Resources," Environmental and
28
Third World, London: Edward Elgar.
29
Resource Economics 1, pp. 1-15.
29
30
30
Peck, S., and Teisberg, T., 1991. "Global Warming Uncertainties and the Value of
31
Manne, Alan S., Richels, Richard G. 1992, Buying Greenhouse Insurance: The Economic
31
Information: An Analysis using CETA," Electric Power Research Insitute.
32
Costs of CO2 Emission Limits Massachusetts Institute of Technonlogy
32
33
33
Peck and Teisberg (1993)
34
March, James, and Simon, Herbert. 1958, Organizations. New York: Wiley.
34
35
35
Polasky, S. and Solow, A.R. 1994. 'The value of a collection of species.' Journal of
36
Marris (**date**)
36
Environmental Economics and Management, forthcoming.
37
37
38
McKibben and Wilcoxen (1992)
38
Polasky, S., Solow, A.R. and Broadus, J.M. 1993. 'Searching for uncertain benefits and
39
39
the conservation of biological diversity.' Environmental and Resource Economics 3, pp.
40
Mearns et al. (1984)
40
171-181.
41
41
42
Mendelsohn, R. and Nordhaus, W. 1994, 'The Impacts of Climate on Agriculture: A
42
Radner (**date**)
43
Ricardian Approach'. American Economic Review, (forthcoming).
43
44
44
Ramsey, F.P. 1928, 'A Mathematical Theory of Saving'. The Economic Journal Vol??, pp.
61
62
I
543-559.
I
Schneider and Chen (1980)
2
2
3
Raper et al. (1990)
3
Sedjo, R.A. and Solomon, A.M. 1989, 'Climate and Forests'.
4
4
5
Rawls, John. 1971, Theory of Justice, Belknap Press
5
Sen, A.K. 1993, 'The Economics of Life and Death'. Scientific American Vol?, pp. 40-47
6
6
7
Reddy and Goldemberg J., "Energy for the Developing World," Scinetific American,
7
Sen, A.K. 1982, 'Approaches to the Choice of Discount Rates for Social Benefit Cost
8
September 1990, pp. 111-118.
8
Analysis'. In: Discounting for Time and Risk in Energy Policy, R. Lind et al.(eds).
9
9
Resources for the Future, Washington, DC, pp. 325-53.
10
Reilly (1993)
10
II
II
12
Reilly, J., Hohman, N. and Kane, S., 'Climate Change and Agricultural Trade: Who
Shah, A. and Larsen, B. 1991. 'Carbon Taxes, the greenhouse effect and developing
12
countries.' World Bank, Washington.
13
Benefits, Who Loses?' Global Environmental Change, forthcoming.
13
14
14
15
Richels, R, and J. Edmonds 1993. "The Economics of Stabilizing
Solow, A.R., Polasky, S. and Broadus, J.M. 1993. 'On the measurement of biological
15
diversity.' Journal of Environmental Economics and Management 24, pp. 60-68.
16
Atmospheric CO2 Contentrations," draft, 11/2/93.
16
17
17
Solow, A.R. 1993. 'The response of sea level to global warming.' In: The World at Risk:
IS
Robey et al. (1993)
18
Natural Hazards and Climate Change, R. Bras (ed). AIP Conference Proceedings 277, pp.
19
19
38-42. American Institute of Physics, New York.
20
Romer (**date**)
20
21
21
22
Rosenthal, D.H., Gruenspecht, H.K. and Moran, E.A. 1994. 'Effects of Global Warming
Solow, Andrew (1992)
22
23
on Energy Use for Space Heating and Cooling in the United States.' Paper presented at the
23
24
Annual Meeting of the Western Economics Association, July 1994, Vancouver, British
Solow, R.M. 1974. 'Intergenerational Equity and Exhaustible Resources' Review of
24
Economic Studies 1970. pp. 29-45.
25
Columbia, Canada.
25
26
26
Rosenzweig, Cynthia and Martin Parry 1994, 'Potential Impact of Climate Change on
Steer, A. and E. Lutz 1993, "Measuring Environmentally Sustainable Development",
27
27
28
World Food Supply'. Nature 367, pp. 133-138.
Finance and Development, pp. 20-23.
28
29
29
Stewart, R. and Wiener, J.B. 1990, 'A Comprehensive Approach to Climate Change: Using
30
Ross (1973)
30
the Market to Protect the Environment'. In: The American Enterprise, Vol.1, ???
31
31
Rothenberg, J. 1993. Time Comparisons in Public Policy Analysis of Climate Change:
Schneider, S.H. and Chen, R.S. 1980. 'Carbon dioxide warming and coastline flooding:
32
32
33
An Economic Analysis. MIT.
physical factors and climate impact.' Amer. Rev. Energy 5, pp. 107-140.
33
34
34
Stiglitz (1975, 1977, 1979, 1984, 1988, 1994)
35
Salop (1971)
35
36
36
37
Salop and Stiglitz (1977, 1982)
Stiglitz, J.E. 1982. "The Efficiency of the Stock Market Equilibrium," Review of Economic
37
Studies 49 pp. 241-61.
38
Sandler, T. 1981 "Tropical Deforestation: Markets and Market Failures," Land
38
39
39
40
Economics, August, 225-33.
Stiglitz, J.E. 1974. "Wage Determinism and Unemployment in LDC's. I: The Labor
40
Turnover Model," Quarterly Journal of Economics 88, pp. 194-227.
41
41
42
Schmalensee, R. 1993, 'Comparing greenhouse gases for policy purposes'. The Energy
42
Stiglitz, J.E. and P. Dasgupta 1971. "Differential Taxation, Public Goods, and Economic
43
Journal 14, no. 1, pp.??
43
Efficiency," The Review of Economic Studies 37, pp. 151-74.
44
44
63
64
I
Stiglitz and Weiss (1981)
I
World Bank 1992, World Development Report 1992: Development and the Environment.
2
2
Washington, D.C.
3
Tietenberg (1992)
3
4
4
5
Titus, J.G. 1987. 'Causes and effects of sea level rise.' In: Preparing for Climate
6
Change. Proceedings of the First North American Conference on Preparing for Climate
7
Change. Government Institutes, Inc., Rockville, MD.
8
9
Tobin, J., and W. Nordhaus 1972, "Is Growth Obsolete?" ***
10
11
Hartwick 1990,
12
13
United Nations Environment and Development Branch, U.N. Center for Transnational
14
Corporations, "Options to Increase Transfer of Environmentally Sound Technologies to
15
Developing countries on Favorable Terms," New York, 1991, 89pp.
16
17
United Nations 1989, 'African Alternative Framework to Structural Adjustment
18
Programmes for Socio-Economic Recovery and Transformation'. UNECA/OAU,
19
E/ECA/cm. 15/6/Rev 3.
20
21
Unterwurzacher and Weiss (1991)
22
23
U.S. Action Plan (1993)
24
25
U.S. Congress, Office of Technology Assessment. Technologies to Maintain Biological
26
Diversity. Government Printing Office: Washington, D.C., 1987.
27
28
Warrick, R., Oerlemans, J. et al. 1990. 'Sea level rise.' In: Climate Change: The IPCC
29
Scientific Assessment. J.T. Houghton, G.J. Jenkins and J.J. Ephrams (eds). Cambridge
30
University Press, Cambridge.
31
32
Weitzman, M.L. 1993. 'What to preserve? An application of diversity theory to crane
33
conservation.' Quarterly Journal of Economics 108, pp. 157-184.
34
35
Weitzman, M.L. 1991, 'On Diversity'. Discussion Paper 1553, Harvard University Institute
36
of Economic Research, Cambridge, MA. 54 pp.
37
38
Weitzman M.L., Newey, W., and Rabin, M. 1981. "Sequential R&D Strategy for
39
Synfuels," The Bell Journal of Economics 12(2), 574-590.
40
41
Weitzman (1974)
42
43
Wilson, E.O. 1988. The Current State of Biological Diversity,' in E.O. Wilson (ed.)
44
Biodiversity, National Academy Press, Washington, pp. 3-18.
65
66
I
83. Analysts now use two methods to estimate stocks. The first assumes a fixed stock of a natural resources such
I
focus.
2
as oil. Consumption of oil then depletes the stock by the amount of consumption. The second begins by treating
2
3
discovered reserves as the asset. Thus, additions to reserves increase the asset, while consumption reduces it.
3
(c) Variation across individual consumers: Engineering models generally assume an average consumer; actual
4
If in any given year, new discoveries match resource utilization, then according to this method no net depletion
4
consumers may display a wide range of characteristics and usage patterns. Except when demand functions are
5
has occurred.
5
linear in the relevant variables, the consumption of the "average" individual is not equal to the average
6
consumption; and what is optimal for the average person may not be optimal for a significant fraction for the
6
84. Many researchers have noted deficiencies in standard national income accounts. First, national income
7
population.
7
accounts do not, in general, provide an adequate measure of welfare; second, they do not provide the correct
8
information relevant for making policies relevant to sustainable development. Sustainable development is
8
74.For a survey, see Jaffee and Stiglitz [ ]. The basic theory of credit rationing was developed in Stiglitz and
9
concerned with society's resources; an economy is growing when its resource base (capital stock combined with
9
Weiss [1981] and the theory of credit rationing is developed in Greenwald, Stiglitz, and Weiss [1984] and Myers
10
natural resources) is growing. GDP does not, and is not intended to be, a measure of resource availability. Firms
10
and Maljuf [1984].
11
have two sets of accounts-cash flow (income) statements and balance sheet statements. GDP is a statement of
12
11
75.This generally accepted methodology is, for instance, reflected in the guidelines issued by the Office of
the former type.
13
12
Management and Budget in the United States for the evaluation of projects and regulations. The applied literature
14
Standard accounting procedures require that firms, in an attempt to present an accurate account of "true income,"
13
does not address the question of whether this procedure is appropriate in the presence of certain types of time and
15
take account of depreciation. GDP measures gross output; it does not take into account depreciation, either of
14
risk non-separabilities.
16
natural or physical capital stocks. The reason is simply that it is hard to get accurate measures of depreciation.
17
15
76.Though if the variance of the net benefits is increasing over time in a particular manner, the differences in the
18
Net national product does take into account depreciation, the change in capital stock. And it is this account which
16
two methodologies may not be large.
19
should be most subject to criticism, since it accounts for changes in the physical capital stock, but not in other
20
17
77.Again, under certain restrictive conditions, where the shadow value of a capital constraint is changing
capital assets, in particular, environmental assets and natural resources.
21
18
systematically over time, the differences in the two methodologies may not be great.
22
A number of difficult conceptual problems face the analyst defining levels and changes in levels of these assets.
23
19
78.See Wilson Though the original discussion of winners curse focused around bidding in auctions, it has
24
20
subsequently come to be applied to a range of other market phenomena. See, e.g. Stiglitz [ ].
25
First, how should the "stock" of natural resources be defined? Coal poses perhaps the easiest situation. The
26
location of coal reserves is known. Costs of extraction are high, so the rents (the value of coal in situ) is low.
21
79.For a discussion of the role of the state in capital markets, see Stiglitz [1994].
27
The depletion can be measured not by the coal used times the market price, but the coal used times the in situ
28
value. But for oil and other minerals, information about where reserves are located is vital. Two models have
22
80.In some industrialized countries, energy efficient home mortgage lending may help correct the problem.
29
23
Lenders generally set criteria for the maximum loan amount based on the borrowers' ability to repay, which in
been proposed. One sees the world as having a fixed stock of natural resources (say oil). When one uses oil,
30
24
turn depends on income and wealth. The fact that a particular expenditure which would enhance efficiency and
one is depleting this stock. Thus, to calculate the value of depletion, one does not need to know the entire stock;
31
25
reduce utility bills is not given special attention. Energy efficient mortgages provide funds to households to
the flow (the amount of oil consumed) provides an accurate measure of the change in stock.
26
make energy efficiency enhancing investments intended to pay for themselves, i.e. result in reduced utility bills
32
27
33
The alternative model looks at the size of discovered reserves. Reserves are treated as the asset. Additions to
equal to or greater than the interest payments. With capital constraints, builders may have an incentive to trade
28
off initial capital costs for higher maintenance costs (lower energy efficiency). Building codes specifying minimal
34
reserves thus are viewed as increasing the resource base. If in any given year, new discoveries match resource
29
levels of energy efficiency and full disclosure of expected life cycle energy costs may help address these market
35
utilization, then there is no net depletion. This is the approach being taken by the U.S. Department of Commerce.
30
distortions.
36
This accounting framework would be correct if there were an infinite supply of the resource (reflected in zero
37
31
rents): the essential "capital" good is information about where the resource is located.
38
32
39
Environmental assets-such as air quality-present another set of problems, because there are not market prices
33
81. A country that rapidly depletes its natural resources may show a high rate of growth under conventional
40
to value the asset. Dynamic aptimization problems of the kind described earlier can be used to calculate shadow
34
41
income accounting, but a lower rate of growth when resource depletion is taken into account. Repetto (1991,
prices. How sensitive these shadow prices are to specific assumptions remains to be investigated.
35
1992) calculated resource-adjusted GDP for several countries rapidly harvesting their stocks of hardwoods and
42
36
other resources, arguing that conventional measures sharply overstated GDP.
43
Accounting systems do not, however, have to aggregate all information together. Just as information about
44
longevity and other indicators of well being (see below) serve to complement information from national income
37
82.Cobb and Daley (1989) have even claimed that U.S. per capita GDP, when adjusted for environmental damage,
45
accounts concerning standards of living, so too can information about physical environmental measures be used
38
was stagnant between 1950 and 1986. This assertion is hard to reconcile with the steady improvement in most
46
to complement information from the extended national income accounts.
39
measures of environmental quality since 1970, when measurements standards were established.
47
85. There is some concern that excessively broad and long patents may actually impede innovation. When
48
technological progress occurs by building on previous innovations, later innovator require permission of earlier
49
innovators to realize the returns on their innovation. While advocates of broad patent coverage argue that the
52
51
I
revenues can be used. These include: reducing budget deficits; reducing marginal rates of income, payroll,
I
67.This issue has been addressed in several papers. See, in particular, Stiglitz [1975, 1984] and Grossman [1981].
2
corporate or other taxes; granting tax incentives to preferred activities; or increasing the level of government
3
expenditures. The costs of the tax will vary widely depending on how the revenues are recycled.
2
68.This is both because those who would, under "true" disclosure, be at 8 competitive disadvantage have an
3
incentive to add "noise" and because there are strong market forces for product differentiation; in markets with
4
60.Top-down models estimate that for developing countries there exist low-cost options to reduce emissions in
4
homogeneous commodities, profits, even with a limited number of suppliers, will be driven to zero (in Bertrand
5
the near term, but eventually costs would exceed I to 2% of GDP (EMF 1993). For economies in transition,
5
competition.) For a discussion of these and related issues, see Salop [1977], Salop and Stiglitz [1977, 1982],
6
because of historical inefficiencies and energy subsidies, there exist large opportunities to reduce emissions at
6
Stiglitz [1977, 1988].
7
little or no cost. For developing countries, problems of informal economies make hard estimates difficult, but
8
the cost of stabilizing emissions would likely be large enough to cut into economic growth.
7
The standard reference in the organizational literature is March and Simon Economic theories emphasizing
8
the non-value maximizing behavior of managers include the works of Baumol 1 ]. Marris I ]. and Leibenstein
9
51.Recent comparisons indicate that the most important differences between top-down and bottom-up models arise
9
I
]. The principal agent literature [Ross, 1973, Stiglitz, 1974] provided the informational micro-foundations for
10
from differences in input parameters, rather than from differences in model structure.
10
understanding the divergence of interests. See Stiglitz [1988]. A more recent overview is provided by Stiglitz
11
I ] and the symposium in the Journal of Economic Perspectives.
11
62.Government institutions and regulations often hinder the efficient use of energy. The developing countries
12
are least able to absorb the costs of these inefficiencies. Thus, while some developing countries argue that they
12
70.This key point, while noted in March and Simon's [1958] original work, was elaborated upon by Radner [
13
cannot afford to reduce greenhouse emissions the same countries have the most to gain from reforming
13
} and Hannaway I ].
14
government-caused inefficiencies. At least in the short run, international agreements committing countries to
15
eliminate at least the most egregious of these practices might go a long way to addressing the problem of
14
71.The facts that time is a scarce commodity and that decision making in large organizations is decentralization
16
emission reductions,
15
do not in themselves constitute a market failure; they do not prove that resources are not efficiently allocated
17
16
given the real constraints facing society, which include time. However, Greenwald and Stiglitz [1986, 1988] have
17
established a very general theorem showing that when information is imperfect and costly, market equilibrium
18
63.This may also be a problem with electricity generated by the private sector, as regulation has historically set
18
is, in general, not (constrained) Pareto efficient. Thus, there is no presumption concerning the efficiency of the
19
price equal to average cost, rather than allowing it to match the competitive price. In many countries, the increase
19
market economy, even in the absence of the kinds of externality and public goods problems that are associated
20
of competitive pressures has moved electricity prices closer to the marginal cost of production.
20
with greenhouse gases. For a more extended discussion, see Stiglitz [1994].
21
21
64.Full utilization of non-fossil fuel energy sources, taking account of other environmental impacts.) For example,
22
One of the main insights of recent advances in the economics of information is to provide a sound
22
when hydroelectric power generation, which does not increase greenhouse emissions, can be cost-effectively
23
micro-foundations for these theories of the firm. And indeed, the importance of the limitations on the availability
23
expanded without other environmental effects, it should be done.
24
of information, and the consequent importance of attention directing efforts applies to individuals as well as to
24
25
organizations. Some studies have suggested that the limited success of the special tax provisions in the United
25
Eliminating Regulations Impeding Efficient Energy Utilization Many, perhaps most, countries have a host of
26
States designed to encourage savings (IRA accounts) was primarily due to the competitive efforts of banks to
26
regulations which increase energy use as they impede economic efficiency. For instance, the United States has
27
recruit these accounts, and the attention which savings got as a result.
27
had a policy of restricting oil exports to Alaska. Whatever the merits of that policy, it has forced Japan to import
28
28
oil from Indonesia and Saudi Arabia. World oil transportation costs have thus been greatly increased, at the
29
expense of the American economy. Another example of government reform, included in the US' Action Plan
29
72.Network externalities are manifested in other ways: builders fail to install energy efficient light bulbs, because
30
(1993), encourages efforts to expand and improve natural gas markets through continued regulatory reform. These
30
customers dislike them, because stores do not carry replacements; and stores do not carry them because the
31
reform efforts include guidelines to allow greater natural gas use in the summer in coal- and oil- fired power
31
demand for them is too low.
32
plants.
32
33
33
When there are important network externalities, market equilibria are frequently inefficient. The economy
34
Other regulation Unintended effects of many tax, expenditure and other policies have contributing further to
34
might, for instance, get "stuck" in the wrong equilibrium. Government action can, in these instances, "force" the
35
inefficiencies in land use. Among the unfortunate effects of the U.S. Superfund program for the management of
35
economy to move from one equilibrium to another.
36
hazardous wastes has been the creation of large unoccupied holes in the center of major cities.
36
37
65.Consider the following thought experiment: compare an optimally designed road system which only carried
37
73.This is not the only explanation of differences between bottom-up and top-down models. There are several
38
cars; and contrast that with an optimally design road system which also carries trucks. The incremental cost of
38
other features of market behavior that bottom-up models often ignore.
39
carrying trucks is, in most countries, much larger than the proportionate share of the cost they bear in gasoline
39
40
taxes and other fees.
40
(a) Hidden Costs: Consumers value a range of attributes difficult to include in an engineering model. For
41
example, auto buyers value not only initial costs and fuel economy (which computer models can easily calculate),
41
66.For example, in many countries, governments have taken an active role in the dissemination of information
42
but also performance, safety, and durability, which they typically do not.
42
to the agriculture sector. These programs are in some measure responsible for the large increase in agricultural
43
43
productivity in countries with agricultural extension services.
44
(b) Divergence between laboratory and in-use performance: Especially for new technologies, actual energy use
44
45
often differs significantly from energy use calculated in the laboratory. It is the latter upon which purchasers
49
50
I
43.Using either an egalitarian social welfare function approach, or a Rawlsian analysis "behind the veil of
I
52. Only a few models take into account international capital flows. Thus, most models do not address issues
2
ignorance" (Rawls 1971) leads to the rejection of the "polluter pays principle." Since at the time the relevant
2
of industry relocation (McKibben and Wilcoxen, 1992). Chapter 11 provides a more complete discussion on
3
actions are taken, the polluter is not cognizant of the effects, such fees have no incentive effects, but rather appear
3
leakages.
4
as random taxes, lowering each person's expected utility, and in particular the expected utility of the worst-off
5
individual.
4
53. In the case of production of highly substitutable commodities, carbon leakage will, of course, be much greater.
6
5
7
There is a further ethical issue: ascertaining who the true beneficiary of escaping paying for the pollution is
$
generally difficult. It need not be the individual, firm, or country actually engaging in the externality generating
6
54. Whether taxes in fact have this effect depends in part on the shape of the demand curves. With intertemporal
9
activity; in competitive markets, when firms are not charged the full social costs of production, product prices
7
separability in demand curves and constant elasticity, with no backstop technology, a constant ad valorem tax has
10
will fall, giving consumers a substantial fraction of the benefits.
8
no effect on the pattern of consumption.
9
II
44. Manne and Richels (1992) show that, under the IPCC emissions scenarios, even the most drastic controls on
10
Coal presents markedly different issues, not so much because of its greater emissions per unit energy, but because
12
emissions from developed countries would be insufficient to stabilize greenhouse concentrations without some
11
of it higher cost of extraction to price ratio. Lowering producer prices may result in less coal being consumed,
13
means of controlling emissions from developing countries.
12
provided alternative energy sources become available. Thus, taxes on coal are likely to have significant general
13
equilibrium as well as partial equilibrium effects; the increase in the price of coal will lead to a substitution of
14
45.Public goods exist when property rights are not or cannot be clearly assigned. The atmosphere is an
14
gas and oil. If alternative energy sources are not available, such policies will only affect the intertemporal timing
15
international public good because assigning property rights to the atmosphere is difficult for one nation acting
15
of coal consumption (given the much more limited resources of gas and oil). But even that might be of some
16
alone, and particularly difficult when many sovereign states must agree among themselves. Tradable greenhouse
16
value in reducing long run greenhouse gas emissions, as the ability to extract energy from coal may increase
17
emission permits, discussed below, attempt to resolve the problem by explicitly assigning property rights to
17
significantly over time. Analyzing the optimal intertemporal structure of taxes, to minimize long run ambient
18
greenhouse emissions.
18
levels of greenhouse gases, taking into account both intertemporal substitution and substitution across energy
19
19
sources, is a complicated technical issue, that to date has not been adequately analyzed.
20
46.See Abreu I
].
20
55. United Nations Conference on Environment and Development, Framework Convention on Climate Change,
21
May 9, 1992.
21
47. In some cases, equity considerations may prevent Coase (1960), in his discussion of externalities, emphasized
22
the separability between efficiency and equity issues. Though there have been several important qualifications
22
56. The most difficult problem is posed by those investors who invested in these resources, on the basis of one
23
to Coase's conjecture, emphasizing the importance of public goods, imperfect information, and transaction costs
23
regime (where these resources were not taxed). Do they have any special claim to compensation for a "change
24
(see Farrell Stiglitz [1988]). still, the basic insight remains applicable here.
24
in regime." Changes in demands and supplies occur for virtually all resources, and are an inevitable part of the
25
risks in investing. While most economists would argue that arbitrary and capricious changes in policies contribute
25
48. Chapter 7 discusses this issue at greater length.
26
to business uncertainty, and therefore have an adverse effect on economic growth, reasoned changes in policies
27
in response to changes in information are an inevitable part of the business risk.
26
49. These concerns are not just theoretical possibilities, as the following two examples illustrate. Assume that
27
the North imposes high energy taxes, but the South fails to do so. Energy intensive industries, such as aluminum,
28
57. For example, some U.S.A. electric utilities are already making decisions in anticipation of some future policies
28
migrate from North to South. But energy efficiency in the south is much less than in the north, so that the total
29
to limit greenhouse emissions.
29
energy used to produce a ton of aluminum could increase substantially. While economic efficiency would call
30
for locating energy intensive industries where energy efficiency is greatest, a system of partial controls would
30
58. These issues also arise among countries: countries with large coal deposits will find the value of their natural
31
results in energy intensive industries being located where energy efficiency is lowest. Similarly, the reduced
31
wealth eroded, and quite naturally will be less enthusiastic about international agreements having that
32
energy consumption by the North will result in lower producer prices of oil and gas, leading to increased
32
consequence.
33
consumption of energy in the South, partially offsetting any energy conservation induced in the North.
33
59. The studies show a marked variation in GDP losses across models. For example, stabilizing emissions at their
34
50.The precise manner in which this should be done is a technical matter, treated in the literature on Global
34
1990 levels is estimated to reduce U.S. GDP by 2% to 8% in the year 2010 roughly a $20 billion to $80 billion
35
Warming Potential (see, e.g., IPCC 1990). To the extent that there are large differences in atmospheric lifetimes,
35
loss for that year. Estimate of the costs of reducing emissions by 20% below 1990 levels in the year 2010 range
36
then the relative weighting of different greenhouse gases should change over time, since the "shadow price"
36
from .9% to 1.7% of GDP. Aggregated models (top-down) have generally reported higher costs, while
37
associated with effects on relative concentrations at different dates will differ.
37
disaggregated models (bottom-up) have shown lower costs. Chapter 9 contains a more complete discussion.
38
38
51. Similarly, if the developed countries restrict forest cutting, the price of lumber may rise, inducing the
39
These GDP losses occur when the carbon taxes lead to investments that are more expensive than those that would
39
developing countries to cut down more of their own trees. While total carbon sequestration may not increase,
40
take place in the absence of the taxes. The higher the carbon taxes, the greater the investment in price-induced
40
environmental and economic efficiency will decrease if, as some researchers have concluded, hardwood forests
41
conservation and the more fuel switching toward the less carbon-intensive substitutes.
41
in the less developed countries may be the least desirable ones to cut down from an ecological or economic
42
42
perspective (Edmonds and Reilly, 1983).
43
The overall impact of a carbon tax will depend not only on the size of the tax but also on the uses to which the
44
revenues are put. In the standard EMF scenarios, it was assumed that tax revenues will be redistributed in a
45
neutral manner (i.e., without affecting the marginal tax rates). There are, of course, numerous ways in which tax
47
48
I
percent emissions reduction scenario (EMF 1993).
I
34.Although for counter examples to the received wisdom see Coase (1990).
2
22. For example, low-lying states may suffer permanent damage.
2
35.Formally, if A measures the quality of the atmosphere, then each individual's or country's welfare, U, is a
3
function of its own consumption, C, and the shared public good, A: U' (C', A). This does not mean that all
3
23. Weitzman et al. (1981), cited in Lind (1993), make these points in formulating a sequential decision strategy
4
individual's (country's) value changes in A the same; that is may differ in magnitude, and
4
for developing synthetic fuels.
5
even in sign.
5
24.Manne and Richels, Nordhaus, and Peck and Teisberg all report a high value of information on and
6
36.Although there may also be trade-offs: reducing those gases that contribute most to local pollution may
6
in their computer simulations. In addition, *** also reports a high value to information on ***.
7
sometimes be at the expense of increased emissions of greenhouse gases.
7
25.Richels and Edmonds (1993) provide a demonstration of this proposition; they calculate relatively low costs
8
37.This kind of optimization problem was first studied by Edgeworth, Mathematical Psychics ( ). The
8
for stabilizing CO2 concentrations if flexibility in timing is allowed, compared to capping and stabilizing
9
importance of incentive effects for the analysis of distributional issues was first emphasized by Mirriees (1971)
9
emissions to achieve the same atmospheric concentration.
10
and Fair (1970). There are, obviously, important incentive effects: if the LDCs were able to classify any
11
expenditure that had some effect on mitigation as a mitigation expenditure, with the cost borne by the developed
10
26. Alternatively, the Kaya identity may be written
12
countries, they would have an incentive to undertake excess expenditures of this type. The GEF (Global
11
13
Environmental Facility) directly addresses this issue, by only providing funds for incremental costs, that is those
12
Growth rate
=
growth rate
decline in energy
emissions
14
costs that go beyond what would have been the efficient level of expenditures ignoring the public good benefits
13
of CO2 emissions
of output
per unit output
per unit energy use
15
of greenhouse gas mitigation.
14
15
That is, CO2 emissions will not rise as long as output grows no faster than the combined decline in energy
16
38. The importance of the assumptions of the absence of transactions costs, including the presence of perfect
16
intensity per unit of production and CO2 emissions per unit of energy use. This formulation applies most usefully
17
information, has only gradually come to be recognized. See Farrell I ]. or Stiglitz [1988] for an elementary
17
to the developed countries.
18
textbook treatment.
18
27. Chapter 8, "Evaluating the Costs of Mitigation," treats the important issue of inertia and technology.
19
That is, for instance, it makes no difference whether smokers or non-smokers are given the property rights to
20
air. Whether smokers value smoking more or less than non-smokers value clean air will determine whether
19
Note that energy efficient development paths for developing countries have been proposed (Goldemberg and
21
smoking occurs. How property rights are assigned does make an important difference for the distribution of
20
Reddy 1988).
22
welfare. Coase's result that outcomes are unrelated to the initial assignment of property rights obviously ignore
21
23
potentially important income effects.
24
22
29. That is, for conventional exhaustible resources, there is a stock, S. Welfare depends on flows out of the stock
25
A slight extension of this perspective says that social scientists should simply describe the outcome of the
23
each year:
26
bargaining process by which property rights are assigned. Beginning with the important work of Nash [ ). a
24
U(S, So, S2- S₁,........S₁,₁ - S₁,.....), where S, is the stock at the end of period 1.
27
variety of bargaining theories have been developed, most of which emphasize the importance of "threat
25
28
points"--the outcomes which arise in the absence of a bargaining agreement-- to the determination of the eventual
26
Here, welfare depends directly only on the stocks, though indirectly, through the effects on consumption, on
29
outcome. In this case, the fact that the net losses of many developed countries may be limited relative to those
27
emissions, which affect the change in stock:
30
of many of the less developed countries suggests a bargaining solution in which much more of the costs of
28
31
mitigation are borne by the less developed countries than under the "social welfare function" allocations described
29
U(S,,C,(S, - S,),S₁,C₂(S₂- S₁),.......S,, C,(S,,, S,),.....).
32
earlier.
30
30. Even when the flow exceeds the long-run sustainable level, it will not be optimal to reduce the flow
33
40.For small taxes, these are "compensated" taxes, and have no welfare effect, though they have a substitution
31
instantaneously, unless -- or even when - there are zero costs of adjustment.
34
effect, and therefore do reduce pollution.
32
33
For the atmosphere, a sustainable stock of greenhouse gases means stable concentrations. Current emissions are
35
41. The loss in welfare (ignoring the benefits from reduced greenhouse gas warming) are the Harberger triangles,
34
estimated to be about twice emissions consistent with stable concentrations.
36
and can thus be shown to be proportional to the product of the elasticity of demand for energy and the share of
37
energy in national output. Since poorer countries are likely to have less access to alternatives which increase the
35
31. Postponing action may lead to some irreversible damages, for example the flooding of low-lying states.
38
elasticity of demand, and since the share of energy is larger in richer countries, the burden of the tax is
39
36
progressive.
32. This is not quite correct, since price affects incentives for exploration, and some marginal wells would not
37
be drilled if oil prices fall too low.
40
42. The polluter pays principle endorsed by the OECD is exclusively prospective.
38
33. For analyses of market and optimal responses to uncertainty about the arrival of backstop technologies, see
39
Dasgupta, Gilbert, and Stiglitz [ ] and Dasgupta and Stiglitz I ]l.
45
46
PALGRAVES atc for CITES
Chapter 1
Introduction: Scope of the Assessment
Lead Authors:
J. Goldemberg. R. Squitieri, J. Stiglitz. A. Amano, X. Shaoxiong. R. Saha
January, 1995
I
INTRODUCTION: SCOPE OF THE ASSESSMENT
1
1.5.3 Market Failures and Government Responses
29
2
2
1.5.3.1
Revising national accounts
31
3
SUMMARY
I
)
1.5.4 Innovation
32
4
1.1
INTRODUCTION
3
4
1.5.5 Carbon Taxes and Tradable Permits.
33
5
3
1.5.5.1
A Double Dividend?
35
6
1.2
FEATURES OF CLIMATE CHANGE
4
6
1.5.5.2
Energy Taxes
35
7
7
1.5.5.3
Tradeable Permit Markets.
36
$
1.3
CONTRIBUTION OF ECONOMICS
7
8
1.5.5.4
Combining taxes with tradeable permits
36
9
1.3.1 Risk
7
9
1.5.5.5
Intertemporal patterns of taxation
36
37
10
1.3.1.1
Portfolio Theory
8
10
1.5.6 Regulatory Approaches
"
1.3.1.2
Insurance
8
11
12
1.3.1.3
Precautionary Investments
10
12
1.6
SUSTAINABLE DEVELOPMENT
38
13
1.3.2 Sequential Decision Making
11
1.6.1 The Economic Concept of Sustainable Development
38
13
14
1.3.2.1
Value of information
II
14
1.6.2 Implications of Sustainable Development for Developing
15
1.3.2.2.
Option value
11
15
Countries
39
16
1.3.3 Dynamics
12
16
17
1.3.3.1
Kaye Identity
12
17
1.7
CONCLUSIONS
40
18
1.3.3.2
Non-renewable resources, backstop technologies, and
18
1.8
ENDNOTES
42
19
emission reduction strategies
13
19
20
1.3.4 International Public Goods
14
20
21
1.3.4.1
Property rights
IS
57
21
22
1.3.4.2
Paying For An International Public Good: Principles
22
23
and Approaches
15
23
1.9
REFERENCES
58
24
1.3.4.4
Enforcement
17
24
25
1.3.4.5
Knowledge
18
25
26
1.3.5
Efficiency
18
26
27
1.3.5.1
Bankable permits
19
28
1.3.5.2
Exchange/risk efficiency.
19
29
1.3.5.3
Comprehensiveness
19
30
13.6 General Equilibrium
20
31
1.3.6.1
Intertemporal Substitution
20
32
33
1.4
EQUITY
21
34
1.4.1
General Issues
21
35
1.4.1
Intergenerational Equity
24
36
1.4.2
Within-country Equity
25
37
38
1.5
ECONOMICS OF POLICY ACTIONS
25
39
1.5.1
Zero-cost Options
26
40
1.5.2
Government Reform
26
41
1.5.2.1
Removing Energy Subsidies
26
42
1.5.2.2
Property Rights Reform
26
43
1.5.2.3
Administrative Reforms
27
44
1.5.2.4
Regulating Non-greenhouse Externalities
27
45
1.5.2.5
Special Problems of Economies in Transition
27
46
1.5.2.6
Examples of Efficiency-Increasing Policies
28
I
Issues of efficiency and equity can largely be separated. The Framework Convention on
I
SUMMARY
2
Climate Change calls on all parties to implement cost-effective measures for abatement,
2
3
enhancement of sinks, and adaptation. The Framework Convention also explicitly requires
3
Climate change presents the analyst with a set of formidable complications: large
4
an equitable sharing of the burdens of response, recognizing the common but differentiated
4
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
s
long time lags between emissions and effects, an irreducibly global problem, wide regional
responsibilities of the parties. Different countries will be affected differently by climate
5
6
change and by policy responses to it. The South is more likely to be adversely affected
6
variation, and multiple greenhouse gases of concern. The risks of climate change are
7
than the North; moreover, developing countries often lack the financial and technical
7
highly asymmetrical, with a large probability of a small loss, and a small probability of a
8
resources to respond. The Framework Convention does not, however, include a formula for
8
large loss. Even in the presence of significant uncertainty, this asymmetry, plus the
9
sharing the costs of addressing climate change.
9
principles of risk aversion and portfolio balancing provide the rationale for going beyond
10
10
no-regrets policies to those that incur net costs.
11
Efficiency requires that emission reductions occur where their cost is lowest,
11
The atmosphere is an international public good, in that all countries benefit from each
12
irrespective of who bears the financial responsibility. Efficiency calls for removing energy
12
13
subsidies, reforming and clarifying property rights that affect energy use and carbon
13
country's reduction in greenhouse emissions; greenhouse gases are an international
externality, in that emissions by one country affect all other countries to the same extent.
14
storage, and reducing non-greenhouse externalities that have the side benefit of reducing
14
15
greenhouse emissions. Efficiency may also be promoted, and greenhouse emissions
15
16
reduced, by better information dissemination and by addressing capital market imperfections
16
Both public goods and externalities require a legal framework within which the
17
that inhibit the adoption of energy-efficient technology. Dynamic analysis indicates large
17
problems they pose can be addressed. No such legal framework now exists for global
18
potential gains from flexibility in timing of greenhouse reductions to allow for the
IS
climate change. Mechanisms for control of international public goods may include the
definition of property rights, the definition of limits to emissions and a consensus for
19
economical turnover of capital stock, and to allow time for the development of low-cost
19
distributing the same in a fair and equitable manner. If, on the other hand, each agent acts
20
substitutes. Policies that promote efficiency by requiring nations to face the full costs of
20
21
their actions will also address equity concerns.
21
in its individual interest, the result will be too little of the public good and too much of the
22
22
externality.
23
Efficiency also calls for international mechanisms such as joint implementation and
23
Climate change demands a decision process that is sequential, can respond to new
24
coordinated economic instruments. Coordinated carbon taxes and tradable carbon emission
24
25
information with mid-course corrections, and can include insurance, hedging. and the option
25
permits can correct the market failure associated with greenhouse emissions.
26
value of alternative courses of action. The challenge today is to identify short-term
27
strategies in the face of long-term uncertainty. The question is not, what is the best course
28
over the next 100 years, but rather, what is the best course for the next few years, knowing
29
that a prudent hedging strategy sill allow time to learn and change course.
30
31
Policy measures to reduce risks to future generations include 1) immediate reductions
32
in emissions; 2) R&D on new supply and conservation technologies; 3) continued research
33
on how much change is likely and what its effects will be: and 4) investments to assist in
34
adaptation if significant climate change occurs. A well-chosen portfolio of policies will
35
yield greater benefits for a given cost than any one option undertaken by itself. Striking the
36
appropriate balance requires taking into account costs, benefits, and risks.
37
38
In an interrelated global economic system, an attempt to reduce greenhouse gas
39
emissions in one region or one sector of the economy may be offset by increases in other
40
regions or sectors. This may occur through a) the loss of comparative advantage in the
41
carbon-intensive sectors of the regions that limit emissions; b) the relocation of industries;
42
or c) changes in world energy prices and the resulting shift in consumption. Any control
43
strategy must account for these global effects.
44
2
I
1.1 INTRODUCTION
I
on the economy, particularly the economy of certain countries or regions.
2
2
3
In recent years, atmospheric emissions of greenhouse gases have risen significantly
3
Even if the first view is adopted, economics has much to contribute to the discussion, for
4
Concentrations are currently about 25 percent greater than at the beginning of the Industrial
4
the question of cost-effective emissions reductions must still be addressed. If the second
5
Revolution. If current trends continue, concentrations will double from pre-industrial levels
5
view is adopted, economics and cost-benefit analysis will clearly be relevant, both in
6
before the end of the next century and, if unchecked, continue to rise thereafter (IPCC
6
deciding how much mitigation to undertake, and in designing the measures.
7
1990).
7
I
8
This chapter, and others in this Assessment Report, draw on the findings of Working
9
The scientific community has noted the potentially serious effects of increased
9
Groups I and II, and follow the guidelines provided by the Framework Convention. That
10
concentrations. These climatic effects could, in turn, have further effects on the biosphere,
10
Convention makes clear that important questions remain to be addressed by subsequent
"
including an increase in mean global temperature, an increase in sea level, changes in
"
negotiations, including the adequacy of national commitments. This chapter takes the
12
agricultural yields, forest cover and water resources, and a possible increase in storm
12
Framework convention as a political document, recognizing that particular terms of the
13
damage.
13
agreements reached by international negotiation may or may not accord with accepted
14
14
scientific research; that for instance, the agreements may or may not have provided the
15
Increased concentrations of greenhouse gases are the result of fossil fuel burning.
13
basis of a cost-effective approach to mitigating climate change. It is hoped that the findings
16
livestock raising. and other human activities. Concerted action on the part of individuals
16
of this chapter and the assessment report more broadly will enhance an understanding of the
17
and governments will be required to slow the increase in concentrations Changes in
17
costs and consequences of alternative actions, and will provide the scientific basis for
18
greenhouse gases concentrations and the analysis of the climatic and other physical
IS
ongoing negotiations.
19
consequences of those changes lie within the purview of the physical sciences. The role of
19
20
human activity in generating greenhouse gases, the consequences of those changes for
20
1.2 FEATURES OF CLIMATE CHANGE
21
humans, and possible responses, lie within the purview of the social sciences.
21
22
22
Climate change could impose a variety of impacts on society. IPCC Working Group II
23
Climate change impacts are likely to vary dramatically from country to country A
23
analyses these impacts in detail. They include effects on agriculture¹, forests², water
24
warmer climate could benefit sectors of the economies of some mid- and high-latitude
24
resources', the costs of heating and cooling⁴, the impact of a rising sea level rise on small
25
countries. At the same time, a rising sea level and the possibility of increased storm surges
25
island states and low-lying coastal areas', and a possible increase in storm damage.
26
could threaten the survival of some small island states and coastal areas, and could increase
26
Although most attention to date has focused on negative impacts, some impacts will be
27
the risk of midcontinent drought and desertification for inland areas on the periphery of
27
positive. Beyond these tangible impacts are a variety of intangible impacts, including
28
deserts.
28
damages to existing ecosystems and the threat of species losses'.
29
29
30
Within the past decade, a consensus has emerged on some key issues in the economics
30
Climate change presents the analyst with a set of formidable complications: large
31
of climate change. This report describes areas of consensus, as well as areas of
31
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
32
disagreement, the sources of disagreement, and further research that could narrow the range
32
long time lags between emissions and effects, an irreducibly global problem, wide regional
33
of disagreement. This chapter frames the issue of climate change largely from the point of
33
variations, and multiple greenhouse gases of concern.
34
view of economics but also from that of other social sciences, introducing the more detailed
34
35
discussions in the chapters to follow.
35
Large uncertainties. While natural scientists agree that greenhouse gas emissions are
36
36
rising, they do not agree about the mechanism linking concentrations, and temperature.
37
The commitment of resources to mitigate climate change may rest on one of two
37
Further, although natural scientists agree that some warming will occur, they do not agree
38
arguments. The first arises from fundamental values, the second from decision analysis:
38
on the speed of change, or the ultimate amount of change (IPCC 1992, 1994). In addition,
39
39
social scientists do not agree on the size of the behavioral responses or economic effects
40
1. We have only one planet. Some changes are largely irreversible, and may occur
40
that would follow, or on the effect of these changes on well-being.
:41
rapidly. Prudence calls for avoiding a large-scale experiment with the planet. Thus,
41
42
avoiding climate change lies beyond normal economic calculus.
42
Nonlinearities and irreversibilities. Nonlinearities occur when changes in one variable
43
43
cause a more than proportionate impact on another variable. For example, some have
44
2. The potential exists for sudden, largely irreversible non-linearities with major effects
44
suggested that even a modest increase in atmospheric greenhouse gas concentrations could,
3
4
I
beyond a certain point, trigger a substantial increase in temperature. Alternatively, even a
I
world economy, not only are the actions of a single country, or group of countries, not
2
modest increase in average temperature might significantly increase weather-related
2
likely to be sufficient to address the problem, they are likely to be largely offset by actions
3
agricultural losses because, for many crops, extra days of extreme heat severely limit yields
3
of other countries. If, for instance, one group of countries reduces timber cutting to
4
In Figure 1. if the threshold temperature for crop damage is T. then even a small increase
4
increase carbon absorption, then the price of humber will rise, which may induce other
5
in the mean temperature, from ml to m2, may greatly increase the number of days above
,
countries to increase cutting in their forests.
6
the threshold, represented by the area under the curves to the right of T,'. Irreversibilities
6
7
are changes that once set in motion cannot be reversed, at least on human time scales.
7
Moreover, whereas economic analysis generally takes the point of view of a single
8
"Runaway warming" that some have hypothesized is an example. Some oceanographers
$
decision maker or government, the important decisions on climate change will of necessity
9
have expressed concern that warming might disrupt the deep water formation and the North
9
be made by many sovereign governments. Economic and decision sciences are not yet able
10
Atlantic Ocean circulation, responsible for much of the oceans' carbon uptake. Once
10
to predict the outcome of bargaining problems of this type".
=
disrupted, this current might not revert to its original position for centuries
II
12
12
Regional variation. Impacts are likely to vary greatly both within and among countries.
13
Long planning horizon. Greenhouse gas concentration changes occur over a long
13
Some countries and regions will suffer from warming: others will benefit, at least in some
14
period of time, with the full consequences of actions taken over the coming decades being
14
sectors. Some cold countries will benefit from a reduction in heating costs and an increase
15
felt in future centuries. The truly long term nature of the problem is one of the distinctive
15
in the length of growing seasons; some warm countries will see a drop in yields from
16
aspects of greenhouse gas warming. Seldom has the world consciously faced a set of
16
agriculture and forestry; low-lying states are likely to suffer from increased storm surges
17
decisions likely to affect our descendants one, two, or three centuries from now".
17
and flooding. For the world as a whole, the net effect at any time will be the sum of local
10
Because the costs of taking action today are born by the current generation, while the
18
effects at many points on the globe, some positive and some negative; analysts have no way
19
benefits accrue possibly hundreds of years in the future, the world community is now faced
19
to estimate this sum without detailed local calculations (summarized in Chapter 6).
20
with issues of intergenerational equity on an unprecedented scale. While society has
20
21
addressed similar problems concerning trade-offs over periods of fifty or even a hundred
21
Aggregation. On a priori grounds, there is no reason to believe that greenhouse
22
years, the long planning horizon for climate change puts the analytic questions at issue in a
22
warming should have in the aggregate, either positive or negative long-run effects (although
23
new light. The length of time involved has one further implication changes in technology.
23
adjustment implies costs in the short run). Indeed, studies twenty years ago focused on the
24
as well as population and consumption patterns, become of paramount importance.
24
dire consequences of global cooling. It seems unlikely that the earth is somehow endowed
25
25
at this moment with the optimal temperature; the Earth has nonetheless adjusted to its
26
Long life of capital stock. Every country has made large capital investments in its
26
current climate, and readjustments in either direction may be costly.
27
cities, farms, ports, etc. Some of this investment cannot be changed without large costs:
27
28
low-lying port cities, for example, cannot easily be rebuilt. For other investments, the cost
28
Multiple gases of interest. Increases in radiative forcing (heat trapping ability) depend
29
of change will be small; for diversified agricultural economies, the cost of switching from
29
on the concentration of all greenhouse gases, not just CO2, even though most economic
30
one annual crop to another will be small if the temperature increase is on the order of
30
modeling to date has limited itself to co, Because greenhouse gases differ in radiative
31
0.2°C. to 0.3°C. per decade, as climate scientists now expect (IPCC 1994). For less
31
forcing and atmospheric lifetime, analysts have devised measures of global warming
32
diversified economies, the costs may be larger.
32
potential (IPCC 1990), making the different gases commensurable".
33
33
34
Inertia in the climate system. Atmospheric concentrations, rather than emissions,
34
Importance of net emissions. Because greenhouse gas concentrations depend on net
35
determine the amount of warming that occurs. Concentrations change much more slowly
35
rather than gross emissions, changes in forests and other greenhouse gas sinks must be
36
than emissions, meaning that affected nations might not have enough time to prevent
36
taken into account".
37
warming damage or to mitigate economic impacts after the effects of climate change
37
38
become evident. In this respect, the risks of climate change are unlike those of earthquake's
38
Efficiency vs. Equity. How much to reduce emissions is a matter of efficiency. Who
39
or floods. The combination of time lags in the climate system and uncertainty about effects
39
pays is a matter of equity. Economics has much to say about the former, but much less
40
requires taking action before unambiguous information emerges about the timing and
40
about the latter. Nonetheless, equity considerations will drive many of the policy decisions
41
magnitude of climate effects.
41
to be made under the Framework Convention on Climate Change.
42
42
43
Global scope. Climate change is a global challenge, which cannot be answered by a
43
44
single country acting by itself. Mitigation must be coordinated globally. In an interlinked
5
6
I
I
13 CONTRIBUTION OF ECONOMICS
1.3.1.1 Portfolio Theory
2
2
Risk and Uncertainty
3
3
Economics and the social sciences offer perspectives on climate change not provided by
A portfolio manager attempts to get the
4
best return for a given level of risk. One
Uncertainty arises when a decision can lead
4
the physical sciences. In the classic definition, economics is the study of the allocation of
5
scarce resources that have alternative uses. Economics emphasizes the importance of
important approach is to buy several types
to a range of outcomes.
5
6
6
tradeoffs between different uses of resources, and the foregone value of other uses of a
of assets, whose returns are not correlated
Expected return or expected value of a
7
7
resource, called the opportunity cost. In the context of climate change. this means that I)
or are negatively correlated (that is, whose
decision is the mean of the distribution of
8
costs and benefits matter; 2) resources are not free; and 3) resources used for one purpose
prices move either independently or in
8
returns, the amount a person would on
9
,
are no longer available for other purposes.
opposite directions). In this respect, climate
average receive as a consequence of the
10
10
change policy decisions can be compared
decision.
If
II
This chapter sets out the logic of cost-benefit analysis as applied to climate change.
to investment portfolio decisions.
Risk aversion measures an individual's
12
12
Standard cost-benefit analysis requires I) a valuing of costs and benefits over time, using
unwillingness to take risks.
13
13
willingness to pay as a measure of value and 2) a criterion for accepting or rejecting
Several responses can reduce the impact
14
proposals". The standard criterion is the compensation principle (Kaldor date, Hicks
14
of climate change: insurance and
Risk premium is the amount an individual
date). which says that if the project yields positive net benefits, then those made better off
15
15
precautionary investments". Precautionary
would pay to replace the uncertain
16
16
investments may take the form of
distribution of outcomes with the expected
could compensate those made worse off with something extra left over. The result is an
17
unambiguous gain in welfare. without the necessity of weighing effects on different
mitigation or in actions that enhance the
value.
17
IS
individuals.
IS
ability of future generations to adapt.
Certainty equivalent is the amount that
19
Other choices include agreements for one
19
makes an individual indifferent between it
Climate change raises difficulties with both requirements. Valuation is difficult because
20
20
country to fund emissions reductions in
and a risky proposition; for a risk-averse
another country (joint implementation),
person, the certainty equivalent is higher
21
of the difficulty in valuing environmental amenities, which are generally not traded in the
21
22
technology transfer and other forms of
than the expected return; the difference is
22
market. And the compensation principle will not apply if mechanisms for affecting
23
23
international cooperation; and phasing out
the risk premium.
transfers do not exist, either between countries or regions in one generation. or -- especially
24
24
-- between generations. If transfers are not feamble, then the analysis must assign weights
existing policies, such as subsidies to fossil
to different individuals (for example, the utilitarianium gives equal weight to each person)
29
fuels, that reduce welfare and directly or
25
26
26
Only then can conclusions be drawn about net benefits for society as a whole. This issue
indirectly increase greenhouse gas emissions.
27
27
will arise again in the discussion of equity in section I 4 below.
28
28
A well-chosen portfolio of climate change investments will yield greater benefit for
Beyond these fundamental concepts are ideas, originally from other areas of economics,
29
29
a given cost than any one option undertaken by itself. For an individual country, the
30
30
that may be applied directly to the study of climate change; this includes work on risk,
issue is how to choose the portfolio of policy measures best suited to its circumstances, and
31
31
dynamics, sequential decision making, public goods and externalities, taxation, and general
to adjust the portfolio over time in response to new developments. Governments will be
equilibrium.
32
32
making climate change decisions for several decades at least. This means that they will
33
33
have many opportunities to adjust the size (total resources) and mix (choice of measures)
34
1.3.1 Risk
34
of their portfolio of responses.
35
35
The uncertainties associated with climate change are large. Moreover, results available
36
1.3.1.2
Insurance
36
37
to date from integrated assessment models indicate that climate change risks are highly
37
38
asymmetrical, with a high probability of a small loss and a low probability of a large loss
38
Several concepts from insurance have important applications to climate change: risk
39
(Manne and Richels 1992, Peck and Teisberg 1993, Nordhaus 1994). In the past thirty
39
aversion, risk premium, and certainty equivalent.
40
years, much new economic research has focused on rational responses to risk". including
40
41
That individuals and societies are risk averse means that average utility is increased by
41
three areas important to a systematic examination and rational response to climate change:
42
portfolio theory, insurance, and decision analysis.
42
pooling risks, or equivalently that people are willing to pay to reduce the risks they face. If
43
43
society as a whole is risk averse, then some investments with a negative expected return,
44
44
for example in climate change mitigation, should be undertaken if they reduce the
8
7
probability of a loss, or the costs of future adaptation.
I
I
is global by nature, and therefore undiversifiable, traditional (intragenerational) insurance
2
cannot fully insure against climate change risks. Rather, some form of intergenerational
2
3
The magnitude of those expenditures depends on (a) society's degree of risk aversion;
3
insurance is likely to arise as a way to redistribute more of the risk.
and (b) the magnitude of the risk. The risk premium-the extra amount that society is
4
4
$
willing to pay to reduce a risk--is small if the stakes (say, the maximum loss) are small, and
,
Losses associated with climate change are likely to be both correlated and large,
6
large if the stakes are large. An investment of a dollar is justified if it reduces the loss of
6
compared to losses absorbed in a single year by the commercial insurance industry. The
,
expected utility by more than a dollar, and not if it reduces the loss by less than a dollar.
,
long term nature of climate change insurance also raises the problem of contract
8
Thus, results reported below focusing on the expected loss of GDP from climate change do
I
enforcement: will contracts signed today be enforceable tomorrow? Will the insurers be
,
not directly address the risk premium. If a possible outcome is a loss of 10%, even though
9
around to pay claims 50 or 100 years from now? (Even in the industrialized countries,
10
the expected loss is only 3%, then the certainty equivalent loss will exceed 3%. A dollar
10
private markets may be inadequate to insure against losses from a major national disaster
11
investment that reduces this certainty equivalent loss by more than a dollar should be
II
today.) These considerations suggest that private markets may not be able to provide
12
undertaken; such an investment could either reduce the average loss, e.g. by reducing the
12
insurance against climate change, and argue for international action to establish insurance
13
probability of the loss occurring (through mitigation actions) or reduce the variance of the
13
markets, perhaps with government reinsurance. Should such an insurance market be
14
loss. For example, some actions the reduce extreme losses will have more than
14
established, careful attention will have to be given to ensuring that the insured parties
15
proportionate returns.
15
engage in appropriate adaptation actions, mitigating the losses that might be associated with
16
16
any greenhouse gas warming.
17
The insurance expenditures associated with mitigation actions and investments are. in a
17
IS
sense, only the differences between the actual expenditures and the no-regreis benefits (the
IS
19
benefits other than those associated with greenhouse gas emissions). Thus, investments in
19
1.3.1.3 Precautionary investments
20
fuel efficient cars may have a direct benefit in reducing the cost of running a car. The
20
21
mitigation investment is only the additional investment for climate purposes.
21
A business makes precautionary investments to reduce the total risk of its portfolio.
22
22
Numerous policy measures are available to reduce risks to future generations from climate
23
The economic principles of risk aversion and portfolio theory provide the rationale
23
change. Four have dominated discussions in recent years: I) immediate reductions in
24
for going beyond no-regrets policies. Naturally. actions that reduce greenhouse gas
24
emissions to slow climate change"; 2) R&D on new supply and conservation technology
25
emissions at zero or negative costs should be undertaken But a prudent response to the risk
23
to reduce future abatement costs; 3) continued research to reduce uncertainties about how
26
of greenhouse gas warming goes beyond no-regrets policies to some policies incurring net
26
much change will occur and what effects it will have; and 4) investments in actions to
27
costs.
27
assist human and natural systems to adapt to climate change if it occurs. Other choices
28
28
include agreements for one country to fund emissions reductions in another country (joint
29
Traditional insurance involves pooling a large number of diversifiable risks. Insurance
29
implementation), technology transfer and other forms of international cooperation; and
30
exist when people differ in their ability to bear a given risk, or would be affected
30
phasing out existing policies, such as subsidies to fossil fuels, that reduce welfare and
31
differently by the risk. Insurance markets transfer risks from those who are less able to bear
31
directly or indirectly increase greenhouse gas emissions.
32
the risk to those who are more able to bear the risk, and by spreading the risk among a
32
33
large number of individuals, reduce the aggregate impact of the risk". By transferring
33
Precautionary investments may also enhance the ability of future generations to react.
34
risk, traditional insurance can address one aspect of the risk from climate change: the large
34
An important reason that people establish savings accounts is to reduce the impact of
35
differences in expected regional impacts. This holds irrespective of who pays the cost.
35
unfavorable events in the future. Similarly, a society may elect to accumulate capital
36
because wealthier individuals and countries are better able to bear risk". and because
36
against the possibility of a large loss from climate change. This is one thread of the debate
37
many countries likely to be most adversely affected will be developing countries of the
37
over discount rates in Chapter 3. Those who argue for a discount rate close to the
38
South, while many of the countries least affected (or positively affected) will be the
38
opportunity cost of capital point out that society may choose between immediate
39
industrialized countries of the North, who could provide insurance for effects of climate
39
greenhouse gas mitigation, at a cost, and delayed mitigation, with some of the money saved
40
change that might fall harder on the South".
40
put aside as a savings account for our grandchildren in the event of large climate-induced
41
41
damages.
42
Traditional insurance will not, however, address all risks from climate change.
42
43
Shipowners can buy insurance against storm loss because shipping risks are diversified: one
43
44
storm is unlikely to sink all the ships covered by the insurer. But because climate change
44
9
10
I
1.3.2 Sequential Decision Making
I
1.3.3 Dynamics
2
2
3
As a policy question. global climate change is often posed as a choice between a) doing
3
The problem of greenhouse gas warming involves additions to concentrations resulting from
4
nothing at all, or b) committing to all-out effort. Given the large current uncertainties about
4
net emissions over extended periods of time. Thus the analysis must focus on dynamics.
$
costs and benefits of greenhouse mitigation, this is the wrong way to frame the issue. as it
5
Dynamic analysis involves three stages: the dynamic processes involved, the trade-offs,
6
obscures the choices that should be evaluated. Moreover, because option b) may be
6
and judgments concerning those trade-offs.
7
perceived as too expensive to get political support, policy paralysis often results.
,
8
$
Dynamic analyses have lead to important insights. For example, atmospheric
9
A more useful formulation is: "Given current knowledge and concerns, what actions
,
concentrations and therefore temperatures depend on the total amount of greenhouse gases
10
should we take over the next one or two decades to position ourselves to act on new
10
emitted over a period of years. A given concentration target can be achieved by a variety
"
information that will become available?" (Lind 1993) For example, decisionmakers would
11
of emissions time paths. Time paths that provide for the economical turnover of existing
12
like to know if the possibility of irreversible damages" justifies immediately undertaking
12
capital stock and time to develop low-cost substitutes are likely to be less costly. This
13
an aggressive abatement program".
13
suggests large potential gains from flexibility in timing of emission reductions".
14
14
15
Climate change demands a decision process that is sequential and can incorporate new
15
16
information. Timing will be a key element, and the date of resolution of uncertainty an
16
1.3.3.1
Kaya Identity
17
important element of the analysis. Figure I shows schematically the progression from a
17
IS
simple decision to a sequence of linked decisions The simple decision is sometimes
18
The driving forces in emissions of any greenhouse gas can be seen in the following identity
19
referred to as the "leam, then act" model, the second as "act, then learn"; or, in the case of
19
for carbon dioxide emissions, (Kaya 1989):
20
the decision depicted, "act. then learn. then act."
20
21
20
CO2
CO2/E
E/Q
Q/L
L
22
A sequential decision making strategy aims to identify short-term strategies in the face
22
carbon dioxide
carbon dioxide
energy
output population
23
of long-term uncertainty. The next several decades will offer opportunities for learning and
23
emissions
emissions per
per unit
per
24
mid-course corrections. The relevant question is not what is the best course for the next
24
unit energy output
capita
25
100 years, but rather: what is the best course for the next few years, knowing that a prudent
25
26
hedging strategy will allow time to learn and change course.
26
or, expressed in rates of change:
27
27
28
For example, the choices might be 1) immediate investment in new plant and equipment
28
d In CO2/dt - d InCO2/E /dt + d InE/Q /dt + d InQ/L /dt + d InL /dt
29
2) aggressive R&D on greenhouse abatement technology or 3) deferring large investment
29
30
for 10 years, until the nature and size of the threat is better understood, and when costs will
30
that is, the percentage rate of change in carbon dioxide emissions is equal to the rate of
31
presumably have dropped and the job can be done more efficiently.
31
change in carbon dioxide emissions per unit energy plus the rate of change in energy
32
32
requirements per unit output plus the rate of change in output per capita plus the rate of
33
Because of the high cost of being wrong in either direction, the value of information
33
change in population
34
about climate change is likely to be great. In particular, the value of information about the
34
35
sensitivity of temperature to CO2 increases, the temperature damage function. the gdp
35
This identity clarifies different approaches to reducing emissions. For a developed
36
growth rate, and the rate of energy efficiency improvement is likely to be high (Chao 199
36
country with a stable or slowly growing population, as long as increases in emissions/output
37
Peck and Teisberg 1991, Manne and Richels 1992, Nordhaus 1994)
37
efficiency (emissions per unit energy times energy per unit GDP) keep pace with labor
38
38
productivity, CO2 emissions will not increase. Because of the substantial potential for
39
The presence of uncertainty along a dynamic path creates an option value. the value to
39
energy efficiency improvements, this seems feasible for most developed countries. Many
40
preserving choices for the future. In climate change, the term has been used in two
40
opportunities exist for increasing end-use efficiency, represented by the second term on the
41
different ways. One stresses the irreversibilities of climate change; mitigation expenditures
41
right-hand side, for example from cars to public transportation, from less to more fuel-
42
now preserve the option of avoiding adaptation expenditure later. The other stresses
42
efficient cars and homes". Fuel switching (for example, from coal-based electricity to oil,
43
irreversibilities in investment, and the cost of premature turnover of capital.
43
gas, hydroelectric, wind, and geothermal), represented by the first term, also offers the
44
44
potential to limit CO2 emissions in many countries.
11
12
I
For many developing countries, emissions will increase unless energy efficiency and
I
sources --gas and oil are exhaustible; this limits total emissions from gas and oil. With
greenhouse gas emissions per unit of energy change to offset growth in per capita output
2
exhaustible natural resources, the question is not how much will be consumed, only
2
3
and population. For many developing countries with rapidly growing populations, pressures
3
when".
for economic development will make it difficult to direct capital from investments with
4
4
higher greenhouse gas emissions to those with lower greenhouse gas emissions"
$
Figure 2 depicts fuel use in an economy, in three phases. In the first phase, the
5
On two essential issues evidence is currently limited: (a) to what extent will
6
economy relies on non-renewable resources; in the second, on coal; in the third, on
6
improvements in energy efficiency require net increases in investment beyond the resources
7
backstop technologies-- e.g., biomass combined with non-carbon sources. The switch
7
$
saved from reduced energy usage, i.e., how much does aggressive emission reduction
8
points depend on rising energy prices and improving technology, which lowers the costs of
9
depress economic growth? Order-of-magnitude calculations suggest the presence of only
9
the backstops. The figure shows alternative long run scenarios. In panel A, the price of
limited trade offs, at least for the near term; (b) do developing countries have the
10
the backstop technology falls sufficiently slowly that for a time, the economy relies on coal.
10
institutional capacity to achieve the desired increases in emissions efficiency?
II
In panel B, the price of the backstop technology falls fast enough to eliminate the
II
12
intervening stage of primary reliance on coal.
12
13
1.3.3.2 Non-renewable resources, backstop technologies, and emission reduction
13
14
Since, to a first-order approximation, the total carbon load from oil and gas is fixed (and
14
strategies.
15
limited), the total carbon load on the atmosphere is primarily related to coal usage. From
15
16
Though in principle the atmosphere is a renewable natural resource, the longevity of the
16
this perspective, an important uncertainty is the pace at which the cost of the backstop
greenhouse gases and the relationships between stocks and flows mean that for practical
17
decreases. If it decreases fast enough, the intermediate stage of coal dependence will be
"
IS
purposes, it may be better treated as an exhaustible natural resource, but one where welfare
18
short, and the total carbon load low, while if the price decreases slowly, the carbon load
19
depends not just on the flow out of the stock. but 00 the stock itself".
19
can be much larger.
20
20
The central problem with natural resources, whether renewable or not, is timing. Many
21
From this perspective, conservation of gas and oil is an important part of a risk strategy.
21
renewable resources possess a maximum sustainable rate of exploitation. The harvest
22
for it provides insurance against the possibility of delayed arrival of backstop energy
22
cannot long exceed this maximum rate without depleting the stock, and ultimately reducing
"
sources." This, in turn, has important implications for the "leakage" debate, which asks
23
the harvest. Corresponding to the sustainable flow rate is a meady-state stock If, initially.
24
whether, in the event that the North imposes carbon taxes but the South does not, the
24
the actual stock exceeds the steady-state stock. then. for a while. the flow can exceed the
25
South's response will largely offset emission reductions made in the North. It is also
25
26
maximum sustainable flow. The question is how to distribute this excess over time. Even
26
possible that lower prices for gas and oil will induce coal-rich countries to decrease their
27
when the stock is at the sustainable level, in times of emergency it may be desirable to
27
reliance on coal, thus possibly producing negative leakages. But to the extent that occurs,
exceed the maximum sustainable flow. For a renewable resource, this can be done, though
28
the insurance provided by greater conservation of oil and gas is eliminated, and it is this
28
29
only at the expense of decreased flows later".
29
insurance which should be an essential part of the dynamic strategy. If the backstop arrival
30
is delayed, then the earlier fuel switching by China and India will yield no long-run
30
31
Timing of reductions in greenhouse gas emissions should reflect differences in costs,
31
benefits; while if the backstop arrival is early, the whole issue is largely moot. Leakage
discounting (to evaluate those costs). and risk. If technological change will make future
32
needs to be looked at not from a static perspective, of what occurs in a single year, but
32
emissions reductions much less costly. some reductions should be postponed".
33
from a dynamic perspective, corresponding to the long run nature of the global climate
33
Conversely, research on "learning effects" shows that if actions taken today will lower costs
34
34
problem.
faced tomorrow. then these dynamic benefits should be included in the calculus (Anow
35
35
1962, Atkinson and Stiglitz 1969).) In the context of climate change. if emissions
36
36
constraints stimulate technical of other developments that help to lower the costs of
37
1.3.4 International Public Goods
37
38
continuing or additional emission abatement, then reductions should be accelerated (Grubb
38
b
1993, 1995). If discount rates are high, the costs borne by future generations will carry less
39
The atmosphere is an International public good in that atmospheric concentrations are
39
A
weight than if discount rates are low. In the presence of risk, non-linearities, or
40
the result of combined actions by all countries. A pure public good Samuelson, 1954) has
40
41
irreversibilities, the principle of risk aversion suggests a strategy of early mitigation.
41
two properties: the marginal cost of an additional individual using the good is zero
42
(non-rivalry), and the marginal cost of exclusion--of stopping an individual from enjoying
42
The theory of non-renewable resources has a second set of lessons for climate change
43
the good--is prohibitive (non-excludability.)" The atmosphere has both characteristics.
43
policy. Among the four energy categories -- coal, gas and oil, biomass and non-carbon
44
One country's greenhouse gas reductions affect all other countries as well. If this thwarts
44
14
13
I
climate change, all countries benefit".
I
Second, it is inappropriate to redress all equity issues through climate change initiatives,
2
Public Goods and Externalities
2
although climate change should not aggravate disparities between one region and another.
3
Different countries will be affected
)
4
differently, so that the benefits of avoiding
Externality An externality, or spillover, anses
4
5
greenhouse gas warming will differ from
when the private costs or benefits of
Economists agree less about the framework for deciding the burden of financing
production differ from the social costs or
5
mitigation and adaptation. At least four approaches have been proposed to determine how
6
country to country. Further, some actions
benefits Because the social costs or benefits
6
the burdens of taxation should be shared. One approach looks at benefits: just as those who
7
will affect simultaneously local
are external to the private costs that firms face,
7
benefit from private goods must pay for them, those who benefit from a public good should
8
atmospheric conditions (providing local
private markets alone the economy will tend to
I
be made to pay for them. The principle has some force when large differences in
9
public goods) and greenhouse
produce too little of a public good (ke
9
10
concentrations. For example, actions that
education) and too much of 8 public bad (lake
preferences exist within any income class; providing a particular public good benefits some
10
pollution).
of those individuals more than others, creating inequalities in the absence of benefit taxes.
II
reduce urban driving improve local air
II
A major problem frequently encountered, particularly for pure public goods, is that it may
12
quality and at the same time reduce
Public good: A public good has two
12
be difficult to determine who benefits. It is in general possible to ascertain the economic
13
greenhouse gas emissions".
properties. the marginal cost of an additional
13
benefits of mitigation, and these are likely to be quite unequally distributed. But this
14
individual using the good is zero, and the
14
principle, by itself, does not fully determine who should bear the costs: appropriately
13
1.3.4.1
Property rights
marginal cost of exclusion-of stopping an
15
designed mitigation strategies will produce a surplus of benefits over costs, which must
16
individual from enjoying the good-rs
16
somehow be divided.
17
An important strand of economic
prohibitive Lighthouses are public goods
thought associates externalities with a
when lighthouse services are provided to one
17
IS
person, others may enjoy the same services
18
A second approach looks at ability to pay. It is often held that richer countries (or
19
failure to assign property rights (Coase
without cost
19
individuals) should pay more than poorer ones. This approach sometimes rests on the claim
20
1990). Assigning property rights to the
20
that all people are entitled to a certain minimum consumption (Dasgupta 1982) But this
21
atmosphere is particularly difficult, since
Market Failure: Private markets may
21
22
sometimes tall to provide a good at the most
principle does not answer the question of how much extra the richer countries should pay.
this would require the agreement of many
22
23
sovereign states. Tradeable greenhouse
desirable level private markets alone will likely
24
emission permits, discussed below. can be
provide too kew lighthouses and too much
23
A third approach is based on contribution to the problem. Because the North has
pollution
24
contributed more than two-thirds of the stock of greenhouse gases in the atmosphere today,
25
thought of as an attempt to resolve the
25
this approach seems to suggest a large responsibility for the North. On the other hand, by
26
problem by explicitly assigning property
26
the time greenhouse gas concentrations double from preindustrial levels, the developing
27
rights to greenhouse emissions.
27
countries are projected to be contributing more than half of annual emissions, and roughly
28
28
half of the total stock in the atmosphere (IPCC, 1990; Cline 1992). Thus under this
29
1.3.4.2 Paying For An International Public Good
29
criterion, the developing countries might eventually pay far more of the mitigation costs
30
30
than under the other principles described earlier.
31
Who should pay for a global public good? every country faces this question internally
31
32
in determining who should pay for the public goods it provides. Who should. for instance.
32
Economists have turned to a fourth approach the social welfare function to answer
33
pay for pollution control within a country?
33
the question of how much extra different parties should pay, as well as the question of how
34
34
to distribute the surplus. The discussion of equity below differentiates between the
35
Economists generally agree on several principles:
35
Rawlsian and utilitarian approaches. In this case, both approaches yield similar results:
36
36
In the absence of incentive problems, both imply that all of the surplus should be allocated
37
First, concerns about equity should be separated from concerns about efficiency. though
37
to the poorer countries, or that the burden of effort be borne by the richer countries".
38
ultimately they are intertwined. An implication of this principle is that because pollution is
38
39
a social cost of production (and consumption). everyone should be made to pay the full
39
Yet a different approach holds that social scientists as such have nothing to say about
40
social costs of the pollution they generate. Thus, if there is a social cost to a unit of
40
these ethical issues. Coase [1960], for instance, approaches the problem of externalities by
41
emissions of greenhouse gases, that cost is the same no matter who produces the emissions
41
emphasizing (a) the importance of assigning property rights, so that, in the absence of
42
All should pay the full social costs of their actions, whether rich or poor. In this
42
bargaining costs, an efficient solution can be obtained and (b) that the outcome is
43
perspective, corrective (Pigouvian) taxes should be imposed uniformly.
43
independent of how property rights are assigned". Coase also emphasizes the importance
44
44
of transactions costs, which will often influence the choice of policies.
15
16
I
A simple approach that yields efficiency but does not require redistribution (and is thus
I
1.3.4.4
Knowledge
consistent with the two principles enunciated above) requires coordinated tax rates so that
2
2
all countries face the same energy prices. This approach makes the cost of emitting an extra
3
A key element in addressing the problem of climate change is an increase in
3
4
ton of carbon equal across all countries, and each country retaining the revenues thus
4
knowledge--knowledge about climate science as well as about the economic and social
generated The net cost of such a tax (ignoring the benefits from reduced greenhouse
,
aspects of impacts, mitigation, and adaptation. Much of this knowledge is in the nature of
5
gas emissions) will, in general, be smaller for poorer countries, as a percentage of their
6
an international public good. Developing ways of increasing energy efficiency will benefit
6
national output: the burden of the tax is progressive, even though the tax is levied at the
7
all countries. While in some cases, those engaged in the research will be able to
,
8
same rate in all countries".
I
appropriate for themselves a significant fraction of the private benefits (mostly in reduced
9
energy costs), they will not be able to appropriate the broader social benefits, unless there is
9
Accounting for past damages. The Framework Convention on Climate Change directs
10
a sufficiently high energy tax (or permit fee) that the price fully internalizes the emissions
10
the Annex I countries to take the lead in responding to the threat of climate change. Some
II
externality. But even then, the social benefits of innovation tend to far exceed the private
II
have argued in addition that because the North has been the major contributor to current
12
benefits.
12
levels of greenhouse gases, it should bear most of the costs of mitigation. This view says
13
13
that costs should be borne not in proportion to benefits expected, but in proportion to
14
This suggests the need for an international agreement to fund basic research and
14
15
contribution to pollution. This argument, however, is not based on efficiency, which
15
subsidize applied research, particularly in energy-related technologies for the developing
requires that incentives be prospective (forward looking). not retrospective No incentive
16
countries. There need not be a central funding agency, nor a central directorate
16
effects attach to imposing charges based on past actions. Whether to charge nations that
17
determining which research should be undertaken. But some mechanism, possibly
"
contributed co, to the atmosphere is an issue of ethics, not efficiency. Moreover, many
IS
including joint implementation, must exist for sharing research plans results, and for
18
have challenged the ethical basis for assigning responsibility based on past damages".
19
ensuring that the fruits of this research are made freely available.
19
20
20
The controversial issue of population growth. while central to the question of economic
21
1.3.5 Efficiency
21
development, bears on climate change largely through its effect on emissions. This chapter
22
22
will limit its treatment of population to this one effect
23
With some exceptions noted below, issues of efficiency and equity can be separated
23
24
Analysts agree that any actions responding to climate change should be cost-effective: no
24
25
matter who bears the cost of emission reductions, reductions should occur where their cost
25
1.3.4.3
Enforcement
26
is lowest". Because of the low energy efficiency in many developing countries, many
26
Both externalities and public goods need a legal framework within which the problems
27
have proposed that the major effort at the emission reductions be concentrated there.
27
they pose can be addressed. Without compulsory taxation, there is an incentive for each
28
28
individual to be a free rider, though there is some empirical evidence that the free rider
29
Mechanisms for reducing emissions equitably and efficiently, including joint
29
effect may not be as significant as economists have previously assumed (Bohm [date]). In
30
implementation, tradeable permits, and coordinated tax policies, are discussed below and in
30
31
the absence of compulsory taxation, externalities can only be addressed with well defined
31
chapter 11. All of these approaches, however, attempt to confront all individuals and
32
property rights (Coase 1990)45 and a legal system that enforces compensation for
32
producers in all countries with the same cost of emissions. Emission control is an
33
international public good in that greenhouse emission reductions have the same effect,
33
externalities.
34
wherever they occur. Just as efficiency in the production of steel or any other commodity
34
Though in many areas, great strides have been made in the development of an
35
requires that all consumers and producers face the same price, so too with emissions.
35
international rule of law, problems are likely to face any system for enforcing climate
36
Uniform prices can be achieved either through coordinated energy taxes or through
36
change agreements. For example, the incentive for free riding is great, whether through
37
tradeable permit requirements; but unless the rules are applied in a systematic way to both
37
delay, nonagreement, or noncompliance. Several alternatives have been proposed, each
38
developed and developing countries, emission reductions will be inefficient.
38
raising political or economic concerns. No consensus now exists about the relative
39
39
desirability of these alternatives, though it is clear that some sert of international judicial
40
Partial participation in an international emissions reduction program, however, will
40
41
significantly reduce its effectiveness. The growth of international trade has resulted in
41
process will be required to determine and enforce compliance Abren.
42
important links between the developed and developing countries, and the total effects of any
42
43
policy undertaken in the North can only be evaluated in terms taking into account the
43
44
systemic responses, including responses from the South, as discussed under "General
44
18
17
I
Equilibrium," below".
I
1.3.6 General Equilibrium
2
2
3
1.3.5.1
Bankable permits
)
General equilibrium theory, a cornerstone of economic research over the past century,
4
4
demonstrates the advantage of looking beyond first-stage effects. It offers two important
5
Efficiency imposes several requirements. One, just described, is that at any moment, the
3
insights for climate change analysis. First, the various parts of an economic system are
6
costs of reducing emissions should be minimized. The second is intertemporal efficiency.
6
interrelated; perturbations to one part have ramifications for other parts, which may be quite
7
the marginal cost of reducing emissions at two points in time must be the same. If it will
,
distant. Second, when all of the reverberations are taken into account, the net effect of an
8
cost less to reduce emissions at some future date, adjusting for time discounting. option
8
action may be markedly different from the initial (and intended) effect.
9
values (risk), and impacts on atmospheric concentrations, then the reduction schedule
9
10
should be adjusted accordingly.
10
One implication of general equilibrium theory has already been noted: taxes imposed on
II
II
one part of the global economy may have little if any effect on global emissions; they may
12
Intertemporal efficiency would be promoted by allowing banking of permits (allowing a
12
simply result in a relocation of economic activity. Increasingly, the world's economic
13
source to use fewer permits in one year and more in another), and by the development of
13
system must be viewed from a global perspective. Location of economic activities is
14
futures and options markets. A bankable permit system would address some North-South
14
determined primarily by relative factor prices, taking into account certain specific locational
15
equity issues, including the concern in the South that delays in mitigation now by the North
IS
advantages and specialized competencies. If, for example, the OECD countries impose
16
would leave 8 greater burden for the future.
16
carbon taxes on energy-intensive industries, those industries may relocate outside the
17
17
OECD. Further, if greenhouse mitigation puts an economic drag on the developed
IS
1.3.5.2
Exchange/rish efficiency.
II
countries, developing countries would be affected through trade.
19
19
20
When different parties to an agreement hold widely differing views about risk and the
20
If different countries have different obligations to reduce greenhouse emissions, different
21
probability of loss, significant efficiency gains can result from transferring risk among
21
implicit tax rates will result. This will interfere with world economic efficiency--decreasing
22
parties. An earlier section discussed the importance of establishing an international
"
world real output--possiNly with little effect on total greenhouse gas emissions. For
23
insurance market for those facing the threat of losses under climate change. and noted the
"
example, the most energy intensive activities-such as aluminum production--may well
24
advantages of establishing a market within which countries who are less concerned shout
24
relocate to developing countries".
25
the economic risks associated with climate change can assume more of the insurance
25
26
burden. Any efficient international system for addressing the problems of global climate
26
While many of the energy-economy-carbon models described in subsequent chapters
27
change must include both mitigation and insurance obligations. Governments that believe
27
attempt to estimate the magnitude of carbon leaks, most are based on standard international
28
they have a comparative advantage in assuming climate nsks can assume a larger share of
28
trade models. in which the location of production of various goods and machines is fixed
29
those risks, trading off other obligations, and substantially reducing overall costs of
29
(Whalley and Wigle, 1992) Thus, estimates of carbon "leakages" are based only on
30
responding to climate change.
30
commodity substitution". But allowing for relocation of industries -- as would occur in
31
31
the very long run, the time span of interest for an analysis of climate change-- leakages
32
1.3.5.3
Comprehensiveness
32
may well be higher. No consensus now exists on the magnitude of long-run leakages.
33
33
34
Efficiency also requires that the cost of reducing all greenhouse emissions be minimized.
34
1.3.6.1 Intertemporal Substitution
35
This principle implies that any mitigation program must include all greenhouse gases
35
36
(taking into consideration their heat-trapping potentials and atmospheric lifetimes"), and
36
General equilibrium issues also arise when production can be shifted from one period to
37
must include carbon sinks.
37
another. For example, in a partial equilibrium analysis, a tax on gas or oil raises the price
38
38
of the fuel taxed, thereby reducing its consumption, and thus associated emissions. But an
39
Finally, and perhaps most controversially, it implies that mitigation strategies focus on
39
exhaustible natural resource like gas or oil has, over the long run, by definition an inelastic
40
all elements of the Kaya identity above, ensuring that the marginal cost of reductions be the
40
supply (ignoring for the moment extraction costs, which, in the case of gas and oil, are
41
same for each of the possible strategies. Thus, population control may be an element in a
41
small relative to the price). The general theory of incidence argues that when a commodity
42
long term mitigation strategy no less than a shift in the composition of production or an
42
is in inclastic supply, a tax affects the price, but not the level of consumption. That is,
43
increase in the energy efficiency of the economy.
43
when all countries impose a tax on gas or oil, the price of oil falls. by an amount just equal
44
44
to the tax. The full incidence falls on producers; as a first approximation, the level of
19
20
I
consumption--and thus the level of emissions--remains unchanged. This provides a first
I
insist that the gainers pay the costs up front. An alternative explanation addresses these
2
approximation. Obviously. if the tax is large enough, the price will fall below the cost of
2
equity concerns by assigning different weights, perhaps based on economic status, to
3
extraction, and supply will be reduced. Moreover, in the case of exhaustible natural
)
changes in consumption of different individuals (Atkinson date)
resources, taxes may affect the timing of consumption of the oil and gae (Stiglitz, 1977,
MC
4
4
5
1979,
1993)
,
No consensus exists among either economists or philosophers about the appropriate
6
6
ethical responses to the changes that would come with climate change. Should, for
7
,
instance, owners of resources be compensated for the losses that they incur as a result of
8
1.4
EQUITY
8
mitigation actions and the consequent change in prices or values? Economists have often
9
,
argued no, for several reasons. First, some wealth is not a reward for productive activity,
10
Which nations will be allowed to increase their greenhouse emissions, and who pays for
10
but merely an accident. It is not because country X did something that $200 billion worth
II
greenhouse gas abatement and adaptation are among the most contentious issues in climate
II
of minerals or oil was discovered to lie beneath its territory. Many would say that these
12
change. This has immediate implications for policy as well, because the initial allocation of
12
random allocations of wealth have actually contributed to world inequality, and that
13
emission rights and emission constraints will largely determine the distribution of costs.
13
eliminating a part of these windfall gains would, from the perspective of an egalitarian
14
The Framework Convention on Climate Change explicitly directs the parties to consider
14
social welfare function, be welfare-increasing. Another view denies that government
15
equity:
15
policies would be taking away value from assets that by right should be there; until now,
16
16
according to this view, these resources have simply been overpriced, not fully reflecting the
17
The Parties should protect the climate system for the benefit of present and future
17
social costs imposed by their use. In this view, actions to discourage overuse would simply
IS
generations of humankind, on the basis of equity and in accordance with their common
18
rectify a previous mistake."
19
but differentiated responsibilities and respective capabilities. Accordingly, the developed
19
20
country Parties should take the lead in combating climate change and the adverse effects
20
Whether investors or countries should be compensated for the adverse effects on their
21
thereof
21
market values remains controversial. There is, however, reasonable consensus on three
22
22
general principles: First, workers in adversely affected sectors may need assistance to
23
[Account must be taken of the differences in their starting points and approaches,
23
switch occupations (The market failure here is that workers cannot purchase insurance to
24
economic structures and resource bases, the need to maintain strong and sustainable
24
insure themselves against these kinds of adverse shocks.) Second, gradual transitions may
25
economic growth, available technologies and other individual circumstances, as well as
25
significantly lower the absolute cost of the transition. For instance, workers leave jobs
26
the need for equitable and appropriate contributions by each of these Parties to the
26
through natural attrition; if those leaving are not replaced, the industry will be scaled down,
27
global effort regarding [the Convention's) objective"
27
with no transition cost to any individual worker. Third, the magnitudes of the
28
28
uncompensated redistributions associated with any change in policy are often correlated
29
29
with the magnitude of the political opposition.
30
1.4.1 General Issues
30
31
31
Most policy changes produce winners as well as losers. If the relative price of natural
32
Within the limits of cost-benefit analysis, equity arises because of the principle of
32
gas increases, owners of natural gas deposits may actually be better off. In economies with
33
compensation, discussed in section 1.3, above. For example, suppose it could be shown
33
progressive taxation, either of capital gains or consumption, then implicitly some part of
34
that a business-as-usual path produced higher total benefits than a path with lower
34
those gains are shared more broadly.
35
greenhouse emissions. We could not from this evidence conclude from this that the world
35
36
as a whole would be better off. Suppose that costs of warming fall predominantly on one
36
Virtually all policies discussed below also have different effects on different groups.
37
group or one generation, while benefits accrue to another. (For example, some have
37
Residents of very hot and very cold climates consume more energy for heating and cooling,
38
speculated that the costs of damages from warming would fall largely on the South, without
38
and thus would be worse off, relative to those in more moderate climates. In some
39
a compensating increase in benefits (Parikh date).) Unless the gainers actually compensate
39
countries, city dwellers can choose less energy-intensive modes of transport than those in
40
the losers for their losses, cost-benefit analysis cannot conclude that the change has been on
40
the countryside.
41
balance good for society. Compensation is particularly difficult if future generations would
41
42
bear most of the costs, because no "Fund for Future Greenhouse Victims" exists (Steer
42
Many of these impacts will be reflected in capital costs. Thus, the value of land in
43
1992). Thus, some have argued that in the absence of mechanisms to make these transfers,
43
temperate climates is likely to rise, and the value in extreme climates to fall. (Similar
44
we should not rely on possible future transfers from gainers to losers, but should instead
44
points can be raised, of course, about the costs imposed by climate change itself.) This
21
22
I
capitalization effect has both favorable and unfavorable implications. In the long run.
2
residents of colder climates are likely to consume less heating fuel, but perhaps at a higher
I
might affect the implied tax rate. For instance, a country which, during the preceding ten
3
price, which would leave them on balance about where they started. Energy prices are
2
years, had made every effort to increase energy efficiency and switch consumption to less
4
likely to rise, and land rents to fall, in an almost offsetting way. On the other hand. current
3
energy-intensive commodities would face the burden of reducing its emissions still further.
5
owners of land would bear the full brunt of the present discounted value of all future
4
Because the marginal cost curve for emissions reductions rises steeply beyond a certain
6
increases in taxes (or the tax-equivalent cost of regulations designed to reduce energy
5
point, the implied tax rate would be considerably higher than for a country that had
7
utilization.) As a result, unless policy changes are introduced gradually, dramatic changes
6
previously encouraged overconsumption of energy, for example by energy subsidies. For
8
in land values may occur, with possibly large effects on financial institutions and the
7
the second country, achieving the emission targets might require only eliminating the
9
economy as a whole. Anticipation of these policies would partially offset these dramatic
8
energy subsidies, a policy with an implied negative tax rate. For a similar reason, countries
10
changes in land values".
9
with large endowments of hydroelectric power may find it relatively difficult to meet an
11
10
emission target of this form.
12
National Security. Climate change itself may affect the national security of many
11
13
countries. At the same time, policies to reduce greenhouse emissions may affect export
12
Public finance theory has focused extensively on "second-best" policies, recognizing the
14
earnings and therefore national security of the energy exporting countries.
13
difficulty of achieving first-best objectives of either economic efficiency or distributive
15
14
justice. In this context, the central issue is whether there exist alternatives to
16
While countries may be willing (or forced) to accept changes in national wealth as a
15
benchmarking, setting target emission reductions in relation to past emissions. When the
"
result of changes in world prices induced by the response to the threat of climate change,
16
U.S. government recently issued tradeable emission permits for sulfur dioxide, it took
IS
countries are less likely to be willing or able to accept what they may perceive as implied
17
account of emission reductions already achieved. Benchmarking reflects information that
19
threats to national security over which they have some control. Thus, a country with 8
IS
would not be reflected in a simple criteria, such as a particular emission level per unit
20
large endowment of coal becomes more vulnerable if it comes to rely on imported oil or
19
population or GDP. Thus, a strong case can be made for including benchmarking. if not in
21
gas-supplies of which could be cut off in times of war. These countries may feel it
20
the final allocation of permits (obligations), at least in the transition rules.
22.
imprudent to switch, even if private economic gains were to be had; and they are
21
23
particularly unlikely to switch if the benefits take the form of an international public good
22
1.4.1 Intergenerational Equity
24
23
25
Increasing world political stability would clearly address these concerns. But even were
24
Efforts to control greenhouse emissions will provide benefits primarily for our
26
that successful, it would not suffice. To increase national security, policies will need to
25
grandchildren and their descendants. We face a difficult task in estimating and judging
27
focus on increasing energy efficiency and reducing energy demand. This is an example of
26
what aspects of climate and environment they will value and how best to preserve those
28
the necessity in designing an international structure for greenhouse gas emissions for
27
aspects for them. If we take aggressive action to limit climate change, they may regret that
29
allowing sufficient flexibility that the particular circumstances of each country can be
28
we did not use the funds to push ahead development in Africa, to better protect the species
30
appropriately taken into account.
29
against the next retrovirus, or to dispose of nuclear materials safely. Chapter 3 addresses
31
30
directly the most important issue in intergenerational equity: choice of an appropriate
32
Benchmarks. The earlier discussion of equitable distribution of the burdens of
31
discount rate.
33
responding to greenhouse gases, employing generally accepted principles of public finance,
32
34
avoided the concept of benchmarks, which have played an important role in international
33
This argument also applies to actions with differential impacts on different regions. If
35
negotiations. Indeed, the "obligations" assumed under the Framework Convention obligates
34
greenhouse warming turns out to be a major threat to countries of the South, and if the
36
the developed countries to return emissions to 1990 levels by the year 2000. One
35
North fails to reduce emissions aggressively now, countries of the South might suffer
37
justification for this form of obligation is that it gives some sense of equity. The difficulty
36
additional damage later. Alternatively, if the North chooses to embark on an aggressive
38
of the task for each country was gauged by its 1990 emissions; requiring each to keep
37
control regime now, and if this cuts into Northern growth rates, the result would shrink
39
within that target appeared to impose a proportionate burden on each. It was a rough sense
38
exports markets for the South, and thus reduce growth there; in addition, if the North views
40
of equity, because no attention was paid to the relative incomes of developed countries and
39
its greenhouse efforts as, in effect, development aid for the South, it might cut back on
41
therefore no attention was paid to the implied tax rate.
40
other programs (sanitation, water, education for women, etc.) with a greater impact on life
42
41
expectancy, health, and well-being.
43
But in a more fundamental sense this criterion can be criticized as inequitable. It pays
42
44
no attention to past efforts at achieving energy efficiency, nor to other circumstances that
43
44
23
24
I
1.5.1 Zero-cost Options
I
1.4.2 Within-country Equity
2
2
While most discussions of equity and climate change have so far focused on North/South
3
A variety of inefficiencies in the energy sector-many of them government
3
4
induced--would, if eliminated, increase economic efficiency and reduce greenhouse gas
4
issues or on issues between one country and another, issues of equity within a country are
5
also important, and indeed play a central role in the political debates about appropriate
5
emissions at the same time. How large is the reservoir of conservation opportunities?
responses to climate change. Most policy recommendations involve large within-country
6
Proponents of the two major approaches to the question have debated this point for more
6
losses for certain groups. For instance, any policy leading to less use of coal, and lower
1
than a decade. Top-down models extrapolate observed behavior into the future. Bottom-up
7
producer prices for coal, will lead to large losses for coal mine owners and workers".
$
models, combine engineering analyses to estimate costs with economic models of individual
8
9
choice. Top-down models generally show significant costs to reducing greenhouse
9
10
The net efficiency gains (in reduced emissions) relative to the distributive effects may
10
emissions in the future Bottom-up, or technology-specified models, have been used to
differ markedly across resources. Thus, if elasticity of world oil supply is small, a tax on
If
show the existence of significant reductions in the cost of energy as new low emissions
"
oil will be reflected in the prices received by producers, and have little effect on cumulative
12
technologies are adopted. Some proponents of bottom-up models argue that emission
12
13
reductions can be achieved at essentially no cost, seemingly repealing the long standing
13
consumption of oil, though it may result in some short-run substitution against coal.
14
Policies aimed at increasing the date of arrival of substitutes for fossil fuels could lead to an
14
economists' presumption that there is no free lunch".
increase in current emissions, though long-run effects on atmospheric concentrations would
15
IS
16
Much of the disagreement turns on empirical estimates. Economists have catalogued the
16
be positive.
"
unintended consequences of government regulation. Many have also identified important
17
18
market failures that could give rise to inefficiencies within the private sector itself. The
IS
19
next two sections will examine each of these effects.
19
1.5
ECONOMICS OF POLICY ACTIONS
20
20
Earlier sections set forth a basic framework for analyzing policies related to global
21
1.5.2 Policy Reform
21
climate change, including a combination of mitigation, adaptation, and possibly climate
22
22
23
A variety of government reforms could enhance energy efficiency, including removing
23
engineering. Striking the appropriate balance requires taking into account the costs,
24
benefits, and risks associated with each strategy. For instance, if risk were irrelevant.
24
energy subsidies, reforming or clarifying property rights, reducing non-greenhouse gas
governments should reduce emissions to the point at which a dollar of extra spending
25
externalities, and administrative reforms.
25
would yield a dollar of expected savings in damages imposed by climate change. or save an
26
26
27
extra dollar of expected costs of adaptation. Adding risk and sequential decision making
27
1.5.2.1
Removing Energy Subsidies
complicates the analysis but leaves unchanged the basic principles. Because of the lasting
28
28
impact of climate change, and the magnitude of the resulting economic uncertainties, most
29
Energy subsidies induce inefficient energy use, reducing the total output of the economy
29
30
as well as increasing CO2 emissions. These subsidies are especially prevalent in
30
policy analysis has focused on a narrower set of questions:
31
developing countries". Shah and Larsen (1991) estimated world energy subsidies in 1990
31
32
1. What actions would improve economic efficiency (including the social costs of
32
to have been $230 billion, calculating that their elimination would reduce global carbon
implementing the policy) and reduce net greenhouse gas emissions? How much
33
emissions by 9 1/2% in addition to improving allocative efficiency thereby generating a
33
emission reductions could be achieved by these means?
34
welfare gain in subsidizing countries. Burniaux et al. (1992) obtained similar results using
34
35
the GREEN model, concluding that the elimination of all existing distortions on energy
35
2. Beyond these zero-cost options, what are the least-cost methods of reducing
36
markets would yield an increase in world real income of 0.7% per year in addition to
36
greenhouse gas emissions? What does the cost curve look like"? What are the
37
cutting world emissions by 18% in 2050. (Dean 1994) Likewise Unterwurzecher and Wirl
37
38
alternative policy measures, and how do they compare?
38
(1991) estimated that raising energy prices to world levels in Poland, Hungary, and
39
Czechoslovakia would reduce carbon emissions from those countries by 30%. Agricultural
39
40
3. What are the essential ingredients of an adaptation strategy, and to what extent will
40
subsidies also distort the outcome, especially affecting the size of forests.
market forces, on their own, provide the appropriate adaptive responses?
41
41
42
1.5.2.2 Property Rights Reform
42
43
43
44
One responsibility of governments is to define property rights and enforce contracts.
44
26
25
I
III-defined property rights encourage overconsumption of resources. A clearer definition of
I
While these problems are largely of governments' making, the remedy is likely to rely
2
property rights could be particularly important in helping to decrease deforestation, for
2
on a combination of public and private actions: effective environmental regulation,
3
example, while improving economic efficiency. Uncertainties about future property rights
3
elimination of government-caused price distortions, and an economic environment in which
4
may also contribute to economic inefficiency. Thus, for example, in developing countries
4
foreign and domestic investment can enhance the efficiency (including energy efficiency) of
5
where large forests are owned by a few large landowners, excessive deforestation may
5
6
result from landowners' fear that their tenure will be limited.
the economy. For example, many analysts believe that cutting methane leakage from gas
6
7
pipelines will yield both high economic and cost-effective reductions in greenhouse
,
emissions.
$
1.5.2.3
Administrative Reforms
8
9
10
9
1.5.2.6
Defining property rights and eliminating energy subsidies are two important actions
Examples of Efficiency-Increasing Policies
II
10
governments can take to reduce greenhouse emissions. At the same time, many less
Policies that cause individuals not to take into account the full social costs of their
12
11
sweeping reforms can improve economic efficiency and simultaneously reduce greenhouse
13
12
actions often result in greater energy use and greenhouse emissions. The National Action
gas emissions. For example:
14
13
Plans of many countries have revealed examples of such policies, and have suggested
15
Pricing of government produced electricity. Many governments price electricity not at the
14
remedies, including:
16
market price but at the cost of production. Economists generally recommend that
IS
17
electricity, like any good, be priced not at its cost of production, but at the competitive
16
Unit pricing of waste disposal to encourage recycling: The life-cycle social cost of
18
price. In countries with a mix of plants, this means that electricity from all sources should
17
consuming a good includes its costs of production plus disposal. Most consumers, and
19
be priced the same--at the highest marginal cost of production".
18
many businesses, pay a flat fee for trasi, disposal; with a flat fee, the marginal cost of
20
19
throwing away an extra pound of trash is zero. By moving from flat fees to unit pricing,
21
Land use and other regulation Changes in land use policy can also reduce energy
20
the actual price consumers pay to buy and dispose of a good will more closely match its
22
consumption (especially for transportation and space conditioning). and thus greenhouse gas
21
full life-cycle social cost.
23
emissions".
22
24
23
Pay-at-the-pump automobile Insurance: In most countries, drivers pay automobile
25
1.5.2.4 Regulating Non-greenhouse Externalities
24
insurance yearly or monthly. Once the premium is paid, the marginal insurance cost of
26
25
driving an extra mile is zero, even though driving more does increase the chance of being
26
in an accident. As an alternative, drivers could be required to pay a portion of their
27
Many activities producing greenhouse emissions also generate pollution of other types
28
For example, fossil fuel combustion releases conventional air pollutants: rush-hour auto use
27
insurance bill at refueling. With pay-at-the-pump insurance, a tax would be levied on
29
contributes to road congestion. In the presence of these spillover effects or externalities,
28
gasoline and earmarked to pay for insurance premiums. This would raise the cost of
30
market solutions will not properly reflect the externalities generated, leading to the
29
gasoline at the pump, but lower auto insurance premiums.
31
overconsumption of environmental resources. Energy taxes, congestion pricing. or tradable
30
32
permits can correct these market signals, resulting in lower emissions of both greenhouse
31
Subsidized parking. In some industrialized countries, many employers provide parking to
33
gases and other pollutants. Some reforms, such as congestion pricing, also reduce the need
32
employees at no cost or lower-than-market cost, thus lowering the relative price of
34
for roads and other physical capacity.
33
commuting by car relative to public transportation. A distortion arises when governments
35
34
tax income spent on public transportation, but not income implicit in the parking subsidy.
36
1.5.2.5 Special Problems of Economies In Transition
35
37
36
Subsidized housing In some industrialized countries, home morigage interest is
38
The economies in transition provide special opportunities for mitigating greenhouse
37
tax-deductible. In all but a few countries, the implicit income on owner-occupied housing
39
emissions. In the old Soviet bloc, high energy subsidies and other price distortions affected
38
is not taxed. These tax provisions encourage individuals to consume more housing than
40
energy usage directly, as well as indirectly through the composition of output. Spotty
39
they otherwise would. In cold or hot climates, where more housing space requires more
41
environmental regulation meant that Eastern Bloc nations lacked the environmental controls
40
energy, this tax treatment increases CO2 emissions.
42
common in the OECD. The capital shortage of the past decade has contributed to the
41
43
problem, through a general deterioration of physical capital stock.
42
Subsidized trucking. Studies suggest that virtually all road damage is caused by heavy
44
43
trucks, which pay only a portion of the expense of building and maintaining the road
44
system. Many countries thus subsidize trucking compared with rail or barge transport,
27
28
1
probably increasing greenhouse emissions".
I
efficiency. The marginal managerial time required to make efficient energy decisions may
2
2
be small and focusing attention on this issue--when information is being freely provided
,
1.5.3 Market Failures and Government Responses
3
through government and other sources-may thus yield private returns well beyond these
4
4
slight marginal costs.
5
Economists agree that governments can adopt policies to increase economic efficiency at
$
6
the same time that they reduce greenhouse emissions. For example, in some countries, fuel
6
Returns to Scale and System Effects (network externalities): Some technologies might be
1
prices do not reflect the full social cost of fuel burning: taxes can correct this market
7
economically attractive at a large scale of production, but not on the much smaller scale on
8
failure. There is less agreement about whether, given market prices, firms fail to take
8
which they might initially be adopted. Other technologies exhibit dynamic scale
9
advantage of all of the energy efficiency opportunities available to them. This controversy
9
economies: unit cost falls over time as a function of the cumulative output of firms or
10
underlies the bottom-up versus top-down controversy treated at greater length in Chapter 8.
10
industries. Technology "networks" may also affect diffusion rates. For example, cars and
II
Engineers have identified a host of seemingly profitable actions that would also save
"
trucks fueled with electricity, natural gas, methanol, etc., require a refueling infrastructure,
12
energy. Many economists, however, view this as evidence that the engineering analysis has
12
which itself competes for resources with the conventional fuel infrastructure already in
13
omitted characteristics important to consumers.
13
place".
14
14
IS
The substantial differences in practices both within and between countries suggests scope
15
Building codes can be justified both in terms of these network externalities effects and in
16
for substantially increasing energy efficiency. Moreover, even best practices within a
16
terms of information failures. Consumers often have limited information concerning the
17
country may not put it at the technological frontier. In deciding whether to adopt a new
17
construction of their houses, and obtaining the information after the house is completed is
IS
production process, businessmen look only at the private costs and benefits. Many
18
often difficult. Even were they to be provided with construction details, they would have
19
technologists, however, conclude that even considering private costs only. firms should he
19
difficulty interpreting the implications.
20
undertaking many energy-efficiency improvements. This section attempts to reconcile the
20
21
different schools of thought by reference to information-based market imperfections, as well
21
Capital market imperfections. A major explanation of the difference between best-practice
22
as by the criteria by which businesses make decisions
,
22
and actual-practice technology is that bottom-up models often compute cost-effectiveness
23
23
using a discount rate substantially lower than the cost of capital calculated by firms."
24
Information dissemination Acquiring information is costly. providing and disseminating
information has many features of a public gond Stiglits 1988 Romer 19 190 Arrow In
24
Studies of implicit discount rates consistently show that households and firms use discount
25
25
rates substantially above market rate for long-term government bonds.
26
the absence of government intervention, there will be too little production and dissemination
26
27
of information. This is particularly true for information with widely dispersed impacts, as
27
Two explanations have been offered: I) Risk: Interest rates facing firms and households
28
opposed to information about, for example, the production of certain chemicals, which is
28
reflect the risk premium that lenders require to compensate them for the probability of
29
primarily of value to a few companies
29
default. Firms often use discount rates that include a risk premium to reflect the riskiness
30
Moreover, both theory and evidence support the view that markets. on their own, do not
30
of projects; 2) Capital constraints: Individuals and firms often face rationing in capital
31
provide an efficient level of disclosure of information". Indeed. some evidence indicates
31
markets, both for credit and equity. Recent research has provided a rationale for this
32
that markets may try to obfuscate relevant information." This provides the rationale for
32
rationing based on the fact that information is imperfect and costly.
33
government provision of information, or laws that in many countries require disclosure of
33
34
interest rates and other consumer-relevant information, including appliance energy
34
These capital market problems have one important implication: Models analyzing
35
consumption.
35
best-practice cost-effective technologies using discount rates lower than those typically
36
36
employed by firms will overestimate the rate of dissemination of these technologies and
37
Bureaucratic structure, limited scope of attention In recent years, economic and
37
underestimate the perceived costs (to the firms and households adopting these technologies)
38
organizational theory" has emphasized that large organizations are not, in general. run by
38
of mitigation strategies.
39
owners; that the managers, even with the best designed incentives, do not in general
39
40
maximize the firm's market value; and that one of the principal scarce factors within an
40
But these capital market problems raise three other questions: (1) Are firms rational in
41
organization is time and attention. How managers direct their attention has much to do
41
using such high discount rates? (2) Does the use of such high discount rates imply a
42
with what the firm does." The information disclosures noted in the previous subsection,
42
market failure? (3) If so, will government intervention improve on the market outcome?
43
as well as a number of other government programs focusing on energy efficiency in
43
44
consumer products, electric lights, and motors, help focus management attention on energy
44
Economists emphasize that an analysis of the costs and benefits of a project must
Anaw KJ 1971
29
Essays in the They 1 Rish Bearing
30
N-H
I
separate four issues: timing, risk, capital constraints, and information. Discount rates are
I
of natural resources has no obvious definition. While most geologists would agree on the
2
only to be used for timing. Risk should be treated by converting costs and benefits into
2
size of coal stocks - their location is known and their in situ value can be estimated - this
3
certainty equivalents, then discounting costs and benefits for each year at the relevant
)
cannot be said for oil or minerals". Second, environmental assets-such as air quality--
4
discount rate." Higher risks should not result in higher discount rates." Similarly,
4
present another set of problems, because no market prices exist to value the asset.
5
capital constraints should be reflected in the shadow price of capital, not in the discount
5
6
rate." Because of limited information (and a version of the "winners' curse"") firms
6
Four approaches are commonly used to calculate changes in the natural environment
,
often require threshold rates of return significantly greater than the market rate of interest.
,
(Peskin and Lutz, 1990):
$
In doing so, they may confuse time and information risk: that is, the rules of thumb firms
I
9
use to evaluate investments may sometimes lead to market inefficiencies, including perhaps
9
a) The environmental expenditure approach, used recently in the United States, subtracts
10
in the area of energy-efficient technologies.
10
pollution abatement expenditures from GDP;
11
11
b) The physical accounting approach, used in Norway and France, establishes satellite
12
Even were firms to follow the economist's guidelines, in the presence of capital
12
accounts using physical units of measurement to account for flows and stocks of
13
constraints, market outcomes would not, in general, be socially efficient (i.e., they are not
13
resources;
see box
14
constrained Pareto optimal.) There may be significant discrepancies between social and
14
c) The depreciation approach adjusts gross and net product by-subtracting out the value of
15
private returns to investment (even apart from the externalities associated with greenhouse
15
natural resource depletion Repello 1949; Serafy and
16
gases or technological diffusion.) This provides part of the rationale for possible
16
d) The comprehensive approach Uses both physical measures and value (Statistical Office
17
government interventions in capital markets." Though these capital market imperfections
17
the United Nations, 1992).
IS
imply that there is no presumption that market allocations are efficient. there is no
IS
19
consensus that they lead to significant underinvestment in energy-efficient technologies in
19
Another measure of broad-based welfare, although not including environmental amenities,is
20
particular".
20
the UN's human development index or HDI (UNDP, 1990). The HDI gives a composite
21
21
measure of human development by combining three key indicators: longevity (measured by
22
1.5.3.1
Revising national accounts
22
life expectancy at birth), education (measured by adult literacy and mean years of
23
23
schooling), and income (real GDP per capita adjusted for purchasing power). Although the
24
Some have suggested revising the conventional systems of national accounts to
24
HDI is not directly related to global environmental issues, both climate change and
25
incorporate full social pricing of resources. An early contribution suggested a new
25
abatement policies may affect it".
26
measure of economic welfare based on consumption that increases quality of life (Tohin
26
27
and Nordhaus, 1972). These authors and others recognized that national income
27
1.5.4 Innovation
28
accounting, widely adopted after World War II, measures aggregate income and expenditure
28
29
flows, but does not incorporate environmental costs and benefits.
29
Standard competitive analysis argues that given all required Information and technology
30
30
market economies produce efficient outcomes. But recent economic analyses have shown
31
Conventional national income accounting does not fully report three categories of
31
that, in general, market economies need not result in the efficient allocation of resources to
32
resource expenditures: a) defensive expenditures, either for pollution prevention before the
32
information production and dissemination, and to innovation. The first of these issues was
33
fact or for cleanup after the fact (although these expenditures are not separately reported.
33
discussed earlier. The second is more complex.
34
they are counted in GDP); b) consumption of environmental goods (such as exhaustible
34
35
resources)"; and c) conflicting uses of environmental services (such as the atmosphere,
35
In the absence of intellectual property rights, firms would have less incentive to
36
used by producers as an input into production, and by households as a consumption good).
36
innovate. With standard patent terms, firms are not able to appropriate all the returns to
37
37
their innovative activity. Setting the optimal patent life involves balancing off the
38
One proposal would include in GDP the effect of changes in quality of the environment.
38
inefficiencies resulting from the exercise of monopoly power during the duration of the
39
In Eastern Europe and the former Soviet Union, steady increases in reported post-war GDP
39
patent (static inefficiencies) with the increased incentives for innovation. Largely
40
masked the effects of decades of environmental degradation; for part of that period,
40
because innovators seldom appropriate all of the returns to their innovation, there is a
41
environment-adjusted GDP almost certainly declined".
41
general consensus that markets provide insufficient incentives for R & D; and the greater
42
42
the spill-overs, the greater the undersupply." Since the spill-overs are likely to be greater
43
However, important conceptual problems in defining levels and changes of
43
the more basic the research, this suggests a role for government in subsidizing basic and
44
environmental assets, complicate the task of modifying national accounts. First, the stock
44
near-basic research. In the same way, the high cost of establishing intellectual property
31
32
I
rights impedes the transfer of technology to developing countries.
I
can sell excess permits on the market and generate revenue."
2
2
Still, there is a general consensus among economists that the patent system provides a
3
While permits thus create a marginal cost of production related to the marginal
3
better basis of financing applied research than do government grants, largely because of
4
emissions, carbon taxes impose a tax directly on the marginal emissions. Both systems thus
4
difficulties government has in picking those most likely to produce high returns. The
,
force producers and households to face the true social costs of their actions". Permits are
3
question is, is there any reason to believe that any market failures, in terms of insufficient
6
better for regulating large sources; carbon taxes are better for small sources. In principle,
6
7
level of innovation, are worse in this area than in other areas, i.e. are there any special
1
both could be adjusted to achieve the desired level of emissions desired, although
8
grounds for arguing for government R & D subsidies, provided the government has
$
adjustments of this sort are likely to be difficult in practice.
9
corrected energy prices to reflect the externalities generated? Obviously, in the absence of
9
such corrections, market incentives to provide energy saving innovations will be distorted,
10
The initial allocation of permits will largely determine the distribution of costs of
10
just as market incentives to adopt energy saving technologies are reduced". (Tradeable
II
abstement (Chapter 3 discusses these issues of equity at greater length), and at the same
II
permits have similar effects to corrective taxes, since firms will value reductions in
12
time influence the growth path of participants' economies. For example, an allocation
12
13
emissions from new technologies, since such reductions will require them to purchase fewer
13
based on population at a given date would provide an incentive for population control".
permits.)
14
14
IS
Imposing carbon taxes can have large distributive consequences. While a system of
IS
16
Innovation is important, because it provides perhaps the best opportunity for low-cost
16
grants can largely offset these distributive consequences, such offsetting grants might well
methods of reducing emissions. Several studies have confirmed the impact of accelerated
17
not be made. Providing tradeable permits equal to existing levels of emissions seemingly
17
deployment of advanced energy technologies on the future rate and timing of anthropogenic
IS
makes no firm or household a loser. But granting permits in that way represents effectively
18
19
climate change".
19
a grant of money (such permits have monetary value) in a way which may not well accord
20
with standard ethical principles. For instance, by embarking on an ambitious program to
20
21
1.5.5 Carbon Taxes and Tradable Permits.
21
reduce emissions, a firm may qualify for fewer permits than it would otherwise. Not only
22
does this violate ordinary notions of fairness, but anticipation of granting permits in this
22
Economic efficiency requires all agents in the economy to pay the full marginal social
23
way would, accordingly, have strong adverse effects on emissions".
23
costs of their actions. But firms and households are not charged for the additional warming
24
24
25
potential they add to the atmosphere, and so do not pay the full social costs they impose.
25
That a tax has large distributive consequences, while presenting a political impediment to
26
Two economic instruments can correct this market failure: carbon taxes and tradeable
26
its introduction, is not necessarily an argument against it. Some argue that those who failed
27
to pay the full social costs of their actions earlier are not therefore entitled to special
27
permits.
28
allotments now".
28
29
A tradeable permit scheme involves a determination of the total level of permits and a
29
distribution of the initial allocation, with emission levels for any firm limited to the number
30
Once one recognizes that the distribution of permits across countries will, inevitably, be
30
31
of permits held. The initial distribution may be made by an auction or allocation
31
decided by some principle other than current levels of emissions, then it becomes clear that
32
according to benchmarks (e.g. per capita as of a given date). or by historical emission levels
32
both taxes and tradeable permits will have distributive consequences. An agreement among
33
("grandfathering"); alternatively, emission rights could be grandfathered in at current levels.
33
countries to impose uniform corrective taxes, with each country retaining its own revenue,
and gradually shifted over to a per capita allocation as of a given date.
34
would have little redistributive consequences across countries and the burden of the tax
34
35
would, as noted earlier, likely be progressive.
35
36
Once permits are distributed among the regulated entities, a market is set up allowing
36
37
companies to buy and sell permits according to their plants' planned emissions. The cost of
37
Governments in the North might decide to use some of the revenues so generated to
production then includes not only the costs of conventional inputs. but also the costs of
38
encourage activities (such as R & D directed at technology appropriate for developing
38
39
additional permits to offset additional emissions. Plants whose cost of mitigation is low
39
countries) that benefit the South, or to provide other forms of assistance. Decisions about
will find it relatively easier to abate pollution rather than to buy permits. Plants with higher
40
this could be made bilaterally, or collectively. By contrast, decisions about how tradeable
40
41
costs of mitigation will have a greater preference for buying permits than for abating
41
permits would be allocated across countries would have to be made multilaterally. Arriving
pollution. The price of the permits, which are artificially created scarce resources, is
42
at a formula for distributing these property rights may be far more difficult than arriving at
42
43
determined by the market. With the use of tradable permits, companies have an incentive
43
a tax rate and a procedure for its revision, as any such formula may entail substantial
44
to improve the efficiency of their production, thereby reducing their emissions level, as they
44
redistribution.
33
34
I
1.5.5.1
A Double Dividend?
I
1.5.5.3
Tradeable Permit Markets.
2
2
,
Revenues from carbon taxes may allow a reduction in distortionary taxes elsewhere in
3
In order for systems of emission permits to achieve reductions in emissions efficiently,
4
the economy. If the (compensated) elasticity of demand of labor is relatively high, and the
4
there needs to be a market for emissions across International boundaries. There is some
5
revenues from the carbon tax are used to reduce taxes on labor income, then there would be
5
debate about the role of government or international organizations in establishing a market
6
a double dividend from the carbon tax in the reduced dead-weight loss from the labor tax.
6
for such emissions. While some believe that there are private incentives for the
,
which would otherwise be significant.
7
establishment of markets, others contend that government can play a key market facilitation
8
8
role through establishing centralized clearinghouses for information or even provide for
9
At least three objections have been raised to this idea. First, conceptually: rationalizing
9
permit banking (storage) or brokerage (trading) to facilitate trades between private parties.
10
the tax system by reducing the most distortionary taxes is certainly a worthy goal. but
10
These services would prove especially useful in the more complex international context.
"
should not be confused with imposing a carbon tax; neither implies the other. Second,
II
12
empirically: if the (compensated) labor supply elasticity is relatively low. then the dead-
12
1.5.5.4
Combining taxes with tradeable permits
13
weight loss form the labor tax is low, and the commensurate welfare gain is reduced.
13
14
Third, politically: if carbon tax revenues are used to offset the existing deficit, rather than
14
While carbon taxes and tradeable permits are typically presented as alternatives, policy
13
to reduce taxes on labor and capital, then the carbon tax acts more like an ordinary tax
15
makers may prefer to combine them. The major disadvantage is the additional
16
increase, increasing distortions from taxation to pay for budget items with a lower return
16
administrative cost. The advantage is more subtle: the market value of tradeable permits is
17
than the extra burden imposed. The gains to total welfare (reductions in dead-weight loss)
17
reduced as taxes increase. With an optimal carbon tax, and with a tradeable permit supply
IS
depend on the welfare losses associated with these other distortionary taxes, as well as the
18
set equal to the optimally chosen level of emissions, the price of a permit should be zero.
19
cross-elasticities of demand between carbon and other taxed commodities".
19
More generally, the greater the tax, the less the value of a permit (for a fixed supply of
20
20
permits): and thus, the less the distributive consequences of alternative rules for allocating
21
Even though carbon taxes may have 8 positive effect on economic welfare, they can at
21
the initial endowments of permits. Another possible combination is permits for large
22
the same time have a negative effect on measured economic growth. since those measures
22
sources and a tax (set to equal the permit price) on small sources.
23
typically do not include the value of environmental degradation Researchers differ on the
23
24
size of the loss. The wide spread in the numerical results, however, should not obscure
24
1.5.5.5 Intertemporal patterns of taxation
25
agreement among researchers on a number of important points. All models used in the
25
26
major comparison studies to date have projected: first, that intervention would be required
26
If the target is the long-run atmospheric concentrations of greenhouse gases, then climate
27
to achieve the emissions largets; second, that the size of the required tax increases with the
27
change damages will be approximately the same for emissions in any particular year,
28
stringency of the carbon limit; and third, that the size of the appropriate carbon tax varies
28
although the optimal carbon tax must be adjusted for differences in costs, discounting. and
29
over time, even for the same emissions or concentration target".
29
risk". The focus on concentrations also implies that early reductions are more valuable
30
30
than later reductions.
31
1.5.5.2
Energy Taxes
31
32
32
For exhaustible natural resources, such as oil, economic efficiency requires that those
33
Energy taxes as a means of controlling greenhouse emissions must be viewed as "second
33
deposits with the lowest cost of extraction be extracted first. Hotelling (1931) argued that
34
best" taxes in that they do not directly tax the externality, greenhouse emissions. While
34
competitive equilibrium implies that rents (price minus costs of extraction) must rise at the
35
carbon taxes directly penalize the externality-generating activity, less targeted alternatives.
35
rate of interest. The price of the backstop technology (an energy source assumed to be
36
such as energy taxes, may be politically more acceptable. Carbon taxes reduce emissions
36
available in unlimited quantities at a certain price after a certain date, such as electricity
37
first, directly, by moving up the demand curve; second, indirectly. by encouraging
37
from solar photovoltaic cells) determines the set of resources to be ultimately exploited, i.e.
38
consumers to switch to less carbon-intensive energy sources. On the other hand. energy
38
all resources for which the cost of extraction is less than the price implied by the backstop
39
taxes work through the first path, by reducing total energy consumption. But to the extent
39
technology's price. Thus, it is the tax on oil or gas at the date of switching to the backstop
40
that certain kinds of energy, like hydroelectric, have, at least in the short run, a relatively
40
technology which determines the ultimate amount of oil and gas that will be extracted, and
41
inelastic supply, a major impact will be on oil, gas, and coal; and to the extent that oil and
41
thus the total burden of CO2 placed on the atmosphere by oil and gas. If that were the
42
gas supplies are best described by a model of an exhaustible natural resource, with
42
only matter of concern, one could imply announce a commitment to impose such a tax
43
relatively low extraction costs, most of the supply reduction will occur in coal. Thus,
43
sometime in the future, when relevant backstop technologies become available and
44
indirectly, there will be a considerable amount of switching, through the indirect effects.
44
competitive. That announcement would, if believed, have an immediate effect on current
35
36
I
captured by special interest groups".
I
prices.
2
2
3
Figure 3 shows the dynamics of exhaustible resources. S denotes the total available
3
stock of gas and oil. S. represents the amount of stock left in the ground. a function of the
4
1.6
SUSTAINABLE DEVELOPMENT
4
,
"backstop" price. Imposing a uniform emission tax lowers the cut-off price, and flattens the
,
price curve. The effect on current price depends on how close the current stock is to
6
The concept of sustainable development was formulated about 1980 as a response to the
6
,
economic exhaustion (S,). Given the currently high levels of rent, it is probable that the net
7
apparent conflict between environmental concerns and the need for economic growth,
8
current effect on producers is to raise the producer price, not lower it. Thus, short-run
8
especially in developing countries. At the time, preserving biodiversity and maintenance of
9
9
leakage effects will be the opposite of that predicted in many models.
environmental quality seemed incompatible with a 5- or 10-fold increase in world output, as
10
would be necessary if per capita incomes of the South were eventually to approach those of
10
"
the North now. The sustainable development debate rekindled interest in the question of
II
1.5.6 Regulatory Approaches
12
resource scarcity, originally addressed in the economics literature by Malthus (1798), and
12
13
Regulation of greenhouse emissions may take many forms, including fuel restrictions,
13
revived in the policy arena with the publication of Limits to Growth (1972). A variety of
technology standards, and various economic incentives; Chapter II discusses these options
14
definitions of sustainable development have been proposed. The Bruntland Commission
14
in detail. Economists have long argued for the use of economic incentives for
15
offered this interpretation:
15
environmental management, although governments have so far relied on traditional
16
Malthos Principle Penulation Lorder McMollan
16
regulations almost exclusively. as traditional approaches have been more acceptable to the
17
Sustainable development is development that meets the needs of the present without
17
IS
public and industry. For example, Corporate Average Fuel Economy (CAFE) standards in
IS
compromising the ability of future generations to meet their own needs".
19
19
the U.S.A. require automobiles to meet certain mileage standards.
20
Although the Commission clearly had in mind environmental considerations, its report did
20
Proponents of the traditional approaches often claim that these approaches "force"
21
not spell out exactly what sustainable development included. The Interamerican
21
technology. with less redistribution than forcing technology through taxes Thus, if
22
Development Bank explicitly included environmental concerns in its formulation:
22
23
23
automobile makers are required to attain a certain mileage standard. they will meet the
24
standard; on the other hand. gasoline taxes might have to rise significantly to reduce fuel
24
[Sustainable] development distributes the benefits of economic progress more equitably,
consumption by the same amount. Evidence for the claim of technology forcing. however,
25
protects both local and global environments for future generations, and truly improves
25
26
is equivocal. In several cases in which industry failed to meet the applicable standards.
26
the quality of life.
27
regulators withdrew the standard in the face of unacceptably high economic costs The
27
apparent advantage of technology forcing one large club instead of the subtle and
28
1.6.1 The Economic Concept of Sustainable Development
28
continuous incentives provided by market forces - is often in fact a disadvantage.
29
29
30
Although sustainable development began as an ethical principle, it is at the same time
30
31
Other disadvantages to the traditional approach are: First, traditional regulations do not
31
an economic concept, focusing on two issues: 1) intertemporal equity and 2) capital
in general result in economic efficiency, since those in one sector face implicit or explicit
32
accumulation and substitutability.
32
33
incentives at the margin that differ from those in other sectors. Second. traditional
33
regulations fail to account for offsetting private responses that may neutralize the
34
Intertemporal equity. Robert Solow's definition (Solow 1992), which focuses on
34
regulation's intended effects and even cause environmental harm. Third, traditional
35
intertemporal equity, has enjoyed wide currency among economists: sustainable
35
regulations provide no incentives for exceeding the given target, even when doing so might
36
development requires that future generations be able to be at least as well off as current
36
37
result in little additional cost".
37
generations. The central implication is that any environmental degradation should be offset
38
by increases in capital stock sufficient to ensure future generations at least the same
38
Traditional regulations that focus on inputs and technology rather than outputs have the
39
standard of living. Sustainable development does not preclude the use of exhaustible
39
40
further disadvantage of not directing research toward meeting performance objectives at
40
natural resources, but requires that any use be appropriately offset.
least cost. For instance, when stack gas scrubbers are required, research will be directed at
41
41
42
producing scrubbers at least cost, rather than reducing emissions at least cost. Hence. a
42
In practice, sustainability as defined by Solow provides few constraints on growth paths
dynamic inefficiency is added to the obvious static inefficiencies. Finally. because of the
43
for the developed countries, so long as steady increases in productivity continue. Technical
43
nature of the regulatory process, traditional regulatory designs are more likely to be
44
change alone, without further capital accumulation, may well sustain future living standards
44
38
37
I
and offset any effects of environmental degradation. To see this with a numerical example,
I
following statement on sustainable development and the environment introduced during the
2
note that even if estimates of adaptation costs are taken to be I to 3 percent of GDP should
2
INC negotiations in 1991:
3
significant warming occur, and if even moderate rates of technical progress of I to I 5%
3
4
per annum continue to occur, then future generations 45 to 70 years from now will have
4
Protection of the global climate against human-induced change should proceed in an
5
twice the income of the current generation. Even with no discounting. it would be hard on
5
integrated manner with economic development in light of the specific conditions of each
6
this account alone to justify the sacrifice of further consumption by this generation in order
6
country, without prejudice to the socio-economic development of developing countries.
7
to enhance the standard of living of the future generation.
,
Measures to guard against climate change should be integrated into national
8
8
development programmes taking into account that environmental standards valid for
9
Capital accumulation and substitutability. To what extent can technology. skills, and
9
developed countries may have inappropriate and unwarranted social and economic costs
10
capital equipment substitute for a decline in exhaustible resource stocks or a decline in per
10
in developing
11
capita environmental amenities? Solow's definition, and much of economic theory to date,
II
12
implicitly assumes that substitutes exist or could be found for all resources. As formulated
12
13
by Pearce (1988), if substitution possibilities are high, as most evidence from economic
13
1.7
CONCLUSIONS
seep. book
14
history indicates, then no single resource is indispensable, and intertemporal equity stands
14
15
as the only crucial issue. If on the other hand, human and natural capital are complements
15
Climate change presents the analyst with a set of formidable complications: large
16
or only partial substitutes (e.g., if because of the irreversibility of extinction' capital
16
uncertainties, the potential for irreversible damages or costs, a very long planning horizon,
17
accumulation is only a partial substitute for biodiversity) then different classes of assets
17
long time lags between emissions and effects, an irreducibly global problem, wide regional
18
must be treated differently, and some assets are to be preserved at all costs.
18
variation, and multiple greenhouse gases of concern. The risks of climate change are
"
19
highly asymmetrical, with a large probability of a small loss, and a small probability of a
20
Pearce (1992) distinguished between strong and weak sustainability. Weak
20
large loss. Even in the presence of significant uncertainty, this asymmetry, plus the
21
sustainability-requires that any depletion of natural capital be offset by increases in human-
21
principles of risk aversion and portfolio balancing provide the rationale for going beyond
22
produced capital the Solow criterion - or by the substitution of other forms of natural
22
no-regreis policies to those that incur net costs.
23
capital, such as renewable assets in place of nonrenewable assets. Strong sustainability
23
24
requires that some natural capital, being irreplaceable, must be preserved". It has been
24
The atmosphere is an international public good, in that all countries benefit from each
25
argued that there are no close substitutes for the atmosphere and the climate it produces,
25
country's reduction in greenhouse emissions; greenhouse gases are an international
26
implying no substitution possibilities and hence the need to preserve the atmosphere
26
externality, in that emissions by one country affect all other countries to the same extent.
27
27
28
1.6.2 Implications of Sustainable Development for Developing Countries
28
Both public goods and externalities require a legal framework within which the
29
29
problems they pose can be addressed. No such legal framework now exists for global
30
In many developing countries, Solow's definition would not be viewed as
30
climate change. Mechanisms for control of international public goods may include the
31
acceptable, since it seems to place no weight on their aspirations for growth and
31
definition of property rights, the definition of limits to emissions and a consensus for
32
development. Developing countries have also implicitly criticized the debate over
32
distributing the same in a fair and equitable manner. If, on the other hand, each agent acts
33
substitutability for the same reason: if some natural assets must be preserved at any cost,
33
in its individual interest, the result will be too little of the public good and too much of the
34
then there may be no trade off with development. A leading spokesman for the G-77 has
34
externality.
35
emphasized the importance of economic growth in achieving sustainable development:
35
36
36
Climate change demands a decision process that is sequential, can respond to new
37
None of these linked [development] issues can be resolved unless and until there is
37
information with mid-course corrections, and can include insurance, hedging, and the option
38
broad-based development in the South. Only such broad-based development can
38
value of alternative courses of action. The challenge today is to identify short-term
39
provide the foundation of international security. The Northern approach is to attack the
39
strategies in the face of long-term uncertainty. The question is not, what is the best course
40
symptoms, with a residual emphasis on poverty eradication. But the international
40
over the next 100 years, but rather, what is the best course for the next few years, knowing
41
community must insist on addressing the underlying causes for concern. Development,
41
that a prudent hedging strategy sill allow time to learn and change course.
42
environmental protection, peace, and security are indivisible
42
43
43
Policy measures to reduce risks to future generations include 1) immediate reductions
44
Similarly, the G-77 and China emphasized the need for economic growth in the
44
in emissions; 2) R&D on new supply and conservation technologies; 3) continued research
39
40
I
on how much change is likely and what its effects will be: and 4) investments to assist in
I
2
adaptation if significant climate change occurs. A well-chosen portfolio of policies will
2
1.8
ENDNOTES
3
yield greater benefits for a given cost than any one option undertaken by itself. Striking the
1
4
appropriate balance requires taking into account costs, benefits, and risks.
I
I.A warmer climate would directly affect the temperature-sensitive sectors of the economy: agriculture, forests,
6
and fisheries; and construction. Because of the risk of drought, arid and semi-arid regions are likely to be most
5
7
vulnerable to warming. A warmer climate may also encourage insect populations; this, in turn, is likely to
6
In an interrelated global economic system, an attempt to reduce greenhouse gas
$
decrease agricultural yields in some regions. On the other hand, increased CO2 concentrations increase
7
emissions in one region or one sector of the economy may be offset by increases in other
9
photosynthesis and the effect of water consumption in controlled settings, and may also do so in farmers' fields,
$
regions or sectors. This may occur through a) the loss of comparative advantage in the
10
increasing crop yields.
9
carbon-intensive sectors of the regions that limit emissions; b) the relocation of industries;
11
or c) changes in world energy prices and the resulting shift in consumption. Any control
12
A warmer climate would hurt agriculture in some regions and help in others. Recent global studies find that
10
13
Northern temperate regions could benefit, particularly for small increases in temperature, but low latitude areas
II
strategy must account for these global effects.
14
could lose. The global studies show significant losses from climate change alone that are significantly offset by
12
IS
the direct beneficial effects of higher ambient levels of carbon dioxide on plant growth. Reilly, et (1993)
13
Issues of efficiency and equity can largely be separated. The Framework Convention on
16
calculate that global agricultural production will increase for global mean warming of up to 2°C.
14
Climate Change calls on all parties to implement cost-effective measures for abatement,
17
Most analysts expect that agriculture can adapt without significant added costs for the world as a whole (although
enhancement of sinks, and adaptation. The Framework Convention also explicitly requires
IS
some regions will suffer greater-than-sverage losses), in part because of crop shifting and other adaptation by
15
an equitable sharing of the burdens of response, recognizing the common but differentiated
19
farmers. Trade in international commodity markets (Ausubel, 1990; Rosenzweig and Parry, 1994; Reilly,
16
20
Hohmann, and Kane, forthcoming: Kane, Reilly, and Tobey, 1992) can ensure that agricultural prices change only
17
responsibilities of the parties. Different countries will be affected differently by climate
21
moderately, though particular regions may gain or lose significantly.
IS
change and by policy responses to it. The South is more likely to be adversely affected
19
than the North: moreover, developing countries often lack the financial and technical
22
2.A warmer climate implies both a poleward movement of forests and changes in forest composition. The result
20
resources to respond. The Framework Convention does not, however, include a formula for
23
would be to increase boreal forests and, by a smaller percentage, to decrease tropical forests. On balance, doubling
24
CO2 concentrations would modestly reduce both standing biomass and forest area (Sedjo and Solomon, 1989).
21
sharing the costs of addressing climate change.
25
Some analysts also calculate an additional temporary decline in forest cover during the transition to a warmer
22
Efficiency requires that emission reductions occur where their cost is lowest,
26
climate EPA, 1989). this decline could last for several hundred years, as more forest death occurs on southern
23
27
edges than additional growth occurs on northern edges. Higher CO2 concentrations might also increase tree
24
irrespective of who bears the financial responsibility Efficiency calls for removing energy
1
28
growth, partially offsetting direct losses from warming.
25
subsidies, reforming and clarifying property rights that affect energy use and carbon
26
storage, and reducing non-greenhouse externalities that have the side benefit of reducing
29
3.Some analysts conclude that a warmer climate would significantly reduce water supply (Gleick, 1987). Most
27
greenhouse emissions. Efficiency may also be promoted, and greenhouse emissions
30
climate models now predict hotter and drier weather for the mid continents (IPCC, 1990a). If this occurs, the
28
reduced, by better information dissemination and by addressing capital market imperfections
31
downward pressures on water supply will be intensified by upward pressures on water demand.
that inhibit the adoption of energy-efficient technology. Dynamic analysis indicates large
32
29
33
Because of the key role of local conditions, further generalizations are difficult. River basin runoff is the
30
potential gains from flexibility in timing of greenhouse reductions to allow for the
34
difference between precipitation and evaporation plus absorption (IPCC, 1990a). Hence small variations in these
31
economical turnover of capital stock, and to allow time for the development of low-cost
35
three terms-each of which depends on the accuracy of regional climate forecasts-can make a large difference
32
substitutes. Policies that promote efficiency by requiring nations to face the full costs of
36
in the runoff.
33
their actions will also address equity concerns.
37
34
Efficiency also calls for international mechanisms such as joint implementation and
38
4. A warmer climate implies reduced heating costs and greater cooling costs. AI least for countries in
35
36
coordinated economic instruments. Coordinated carbon taxes and tradable carbon emission
39
temperate and cool climates, this would move the population-weighted average temperature closer to the ideal
40
indoor temperature. Nordhaus (1991) finds that warming of approximately 3.0°C in a 1981-sized U.S.
37
permits can correct the market failure associated with greenhouse emissions.
41
economy will increase electricity costs by $1.65 billion and decrease heating costs by $1.16 billion (1981
42
dollars). resulting in net costs of $0.49 billion. Rosenthal, Gruenspecht, and Moran (1994) find that global
43
warming lowers the cost of heating and cooling in the U.S.. Cline (1992) finds net costs of $9.9 billion.
44
45
5.Global mean sea level appears to have risen 10-20 cm over the past century Warrick, Oerlemans of al., 1990)
46
The IPCC has projected a warming-induced rise in mean sea level of 21 to 71 cm by 2070 (IPCC, 1990a). sharply
47
reduced from estimates of 3 to 8 m. made in the early to (Schneider and Chen, 1980> Hoffman et al.,
48
1983; Hoffman et al., 1986). which assumed a disintegration of the West Antarctic Ice Sheet. The current "best
41
42
All 8/15 draft
I
estimate" for mean global sea level rise by 2030 under "business as usual" is 18 cm (Warrick, Oerlmans at of
I
12. Scientists have proposed several methods to measure the warming potential of different greenhouse gases
2
1990; Reper-er al., 1990)
2
(IPCC Special Report 1994). Although global warming potential (GWP), endorsed by the IPCC, is a useful
3.
Wisky
3
concept in formulating comprehensive approaches to greenhouse mitigation policies (Stewart and Wiener,
4
A rising sea level could inundate some small island nations, flood some low-lying coastal cities (eg in
4
1990), some analysts have recently criticized the concept on the grounds that GWP implicitly sets the
5
Bangladesh), damage coastal farmland, and contaminate water supplies (IPCC, 1990a) More severe storms
,
opportunity costs of an increment in radiative forcing equal for all periods in the future (Schmalemse, 1993).
6
would accelerate coastal erosion and affect aquifers, intensifying problems existing today. Warmer oceans might
6
If all greenhouse gases had the same rate of decay, then this problem would not arise.
7
damage coral reefs, a natural defence for some coastal areas. To account for the costs of protective measures,
8
the costs of defending the world's developed constlines against 8 I - rise over the next 100 years have been
7
,
13. Although forests cannot be expanded Indefinitely, and thus increased carbon sequestration is not a permanent
estimated to range from 0.01% for the former USSR to 0.74% for Indian Ocean small islands, with a world
8
solution to increasing greenhouse gas emissions.
10
average of 0.028% of GNP per annum (IPCC, 1992).
9
II
6. Some analysts believe that the justification for costly and more restrictive actions rests on Intangible costs
10
14.Advances is cost-benefit analysis have allowed the introduction of risk and equity issues in a systematic way.
12
(Nordhaus, 1993). Intangible refers to the difficultly to measuring, and include migration, comfort, health,
13
leisure activities, urban infrastructure, and air pollution (Fankhauser 1992; Cline 1992). A warmer climate
II
15. Frank Knight's often-quoted distinction separates risk, for which the probabilities of different outcomes-are
14
would Improve human comfi " in cold areas, and in the winter generally, while decreasing comfort in warm
12
15
areas. Means of al (1984) calculate a threefold increase in best waves for a 1.7 degree C. rise in U.S mean
13
known, from uncertainty, in which either they be unknown, Boston: or some potentjal putcomes are not specified. Ref.
Knight,
Frask,
1924
Risk.
Uncertainty.
and
Profit
16
temperature It is not yes clear whether net comfort will rise or fall for a given rise in temperature. Chapter
17
6 covers these issues in more detail.
14
16. This division of responses is from Nordhaus (1994)
18
7 Both the natural rate and human contribution to the process of species less are difficult to estimate (EPA, 1989)
13
17. While most insurance markets do not reduce overall risk, there are a few exceptions: fire insurance companies,
19
Predicting the effect of clamate change on species distribution - more difficult will The magnitude and even the
16
by requiring commercial buildings to install sprinklers, do reduce the overall level of losses.
20
sign of these intangibles remains - dispute, for uncertainty about the duration and type of environmental changes
21
that would be caused by climate changes make the long-term projection of species change highly complex
17
IS This assumes decreasing absolute risk aversion (about which there is a general consensus). Even if northern
22
Moreover, while some species may have difficulty adapting to chmetic changes, opportunities for other species
18
countries were adversely affected. so long as the adverse effects were proportionately smaller, then, assuming
23
that otherwise may have become extenct may open up Population pressures have added to pressure on
19
decreasing relative risk aversion (about which there is less consensus), the north would be in 8 position to insure
24
ecosystems, particularly in the third world Climate change may exacorbase these damages, particularly IN Africa
20
the south
25
where environmental degradation has been particularly pronounced during the last 11 years (UN. 1989)
21
22
Insurance contracts may also be created because of differences in judgments concerning the probability of the
26
8. A warmer climate might also increase climate variability. though clamatologists cannot say with assurance
23
insured event occurring. This suggests the potential of an important principle to be invoked in future international
27
whether climate will become more or less variable dealy and seconnally The normal variability of existing
24
negotiations/agreements: countries that believe that the risks of climate change are low, and are therefore
28
climate also makes it difficult to detect any warming that might be occurring The "signal to none" problem
25
seemingly unwilling to take strong actions to mitigate these risks, ought to be willing to provide insurance against
29
makes it possible for observers to mistakenly consider a normal extreme event as evidence of . trend, or to
26
climate change at low cost (since it has, from their perspective, an actuarially low value). (See below for a
30
fail to see B trend in a noisy date series. Because of the signal-to-more problem. the scientific community n
27
discussion of problems of enforce.) (Heal and Chichilnisky **date, reference**; carlier, Stiglitz date ** on
31
unable to indicate confidently whether extremely warm years of the 1980's are evidence of climate change or
28
social risk bearing.)
32
not (IPCC 1992, A. Solow 1992).
29
30
Insurance markets, if appropriately designed, would have one further advantage: some of the losses associated
33
9:Recent ice core data from Greenland points to earlier temperature rises of several degrees occurring within a
31
with climate change are easily avoidable, e.g. construction of durable ocean-front bouses. Insurance firms work
34
few decades (IPCC 1994): reasons for these sudden changes are still not understood, but might have come from
32
hard to mitigate the losses that are incurred. Either the insurance should be based on exogenous events (e.g. not
35
changes in deep ocean currents. The other side of this debate holds that on the whole the biosphere is
33
on the dollar losses incurred, but on the rise in the sea water), or the insurance companies should be given broad
36
homeostatic or self-correcting: this "Gais hypothesis" compares the biosphere to 0 living being once moved away
34
discretion to require the insured to undertake actions to mitigate losses.
37
from equilibrium, self-correcting forces naturally move in back to equilibrium (Lovelock, 1979). The two
38
hypotheses are not necessarily Inconsistent. Within a range of variation, homeostatic properties could dominate,
35
19.Chapter 6 provides estimates of expected regional damage.
39
even if stability was not guaranteed outside that range.
36
20.1n the absence of such rec-irements, moral hazard problems arise. It may be desirable to focus government
40
0.Countries make century-long choices implicitly. as when they choose population policies, policies affecting
37
intervention not on the primary insurance market but on the reinsurance market. See recent U.S. government
41
long-term capital formation and productivity growth, or protection of environmental assets
38
enalysis of the market failures and the design of appropriate responses to those market failures in insurance
39
markets for natural disasters.
42
11. Although bargaining theory has contributed some basic principles, such as the importance of threat points.
40
21. Computer modelers participating in the Energy Modeling Forum examined an "accelerated R&D"
41
scenario in which the cost of nonelectric backstop falls from $100 to 50$ per barrel of oil equivalent, and the
42
cost of the electric backstop falls from 75 to 50 mills per Kwh. The four models used were remarkably
43
consistent in their estimates of economy-wide costs, reporting GDP losses falling by 65 per cent for the 20
43
44
I
percent emissions reduction scenario (EMF 1993)
-
34.Although for counter examples to the received wisdom see Coase (1990).
2
22. For example, low-lying states may suffer permanent damage.
2
5.Formally. if A measures the quality of the atmosphere, then each individual's or country's welfare, U, is 8
3
function of its own consumption, C, and the shared public good, A: U (C, A). This does not mean that all
3
23. Weitzman et al. (1981). cited in Lind (1993). make these points in formulating . sequential decision strategy
4
individual's (country's) value changes in A the same; that is may differ in magnitude, and
4
for developing synthetic fuels.
5
even in sign.
5
24.Manne and Richels, Nordhaus, and Peck and Teisberg all report a high value of information on and
6
36.Although there may also be trade-offs: reducing those gases that contribute most to local pollution may
6
in their computer simulations. in addition, also reports a high value to information on
,
sometimes be at the expense of increased emissions of greenhouse gases.
F.Y.
1881
Londonic Kegan Paul +6.
7
25.Richels and Edmonds (1993) provide a demonstration of this proposition; they calculate relatively low costs
8
37.This kind of optimization problem was first studied by Edgeworth Mathematical Psychics (
The
8
for stabilizing co, concentrations if flexibility in timing is allowed, compared to capping and stabilizing
9
importance of incentive effects for the analysis of distributional issues was first emphasized by Millices (1971)
,
emissions to achieve the same atmospheric concentration.
10
and Fair (1970). There are, obviously, important Incentive effects: if the LDCs were able to classify any
11
expenditure that had some effect on mitigation as a mitigation expenditure, with the cost bome by the developed
10
26. Alternatively, the Kaya identity may be written
12
countries, they would have an incentive to undertake excess expenditures of this type. The GEF (Global
II
13
Environmental Facility) directly addresses this issue, by only providing funds for incremental costs, that is those
12
Growth rate
growth rate
decline in energy
emissions
14
costs that go beyond what would have been the efficient level of expenditures ignoring the public good benefits
13
of CO2 emissions
of output
per unit output
per und energy use
15
of greenhouse gas mitigation.
14
15
That is. CO2 emissions will not rise as long as output grows no faster than the combined decline in energy
16
38 The importance of the assumptions of the absence of transactions costs, including the presence of perfect
16
intensity per unit of production and CO2 emissions per und of energy use This formulation applies most usefully
"
information, has only gradually come to be recognized See Farrell I Stiglitz (1988) for an elementary
17
to the developed countries
IS
textbook treatment.
18
27. Chapter 8. "Evaluating the Costs of Mitigation," treats the important issue of inertia and technology
19
39 That is, for instance, it makes no difference whether smokers or non-smokers are given the property rights to
20
air. Whether smokers value smoking more or less than non-smokers value clean air will determine whether
19
28 Note that energy efficient development packs for developing countries have been proposed (Goldemberg and
21
smoking occurs. How property rights are assigned does make an important difference for the distribution of
20
Reddy 1988).
22
welfare. Coase's result that outcomes are unrelated to the initial assignment of property rights obviously ignore
21
2)
potentially Important income effects.
24
22
29. That is, for conventional exhaustible resources, there is 0 stock. $ Welfare depends on flows out of the stock
25
A slight extension of this perspective says that social scientists should simply describe the outcome of the
23
each year:
26
bargaining process by which property rights are assigned. Beginning with the important work of Nash
a
24
UKS, S. S,- S,......). where S, is the stock at the end of period I
27
variety of bargaining theories have been developed, most of which emphasize the Importance of threet
25
28
points"-the outcomes which arise in the absence of a bargaining agreement- to the determination of the eventual
26
Here, welfare depends directly only on the stocks, though indirectly, through the effects on consumption, on
29
outcome. In this case, the fact that the net losses of many developed countries may be limited relative to those
27
emissions, which affect the change in stock:
30
of many of the less developed countries suggests a bargaining solution in which much more of the costs of
28
31
mitigation are bome by the less developed countries than under the "social welfare function" allocations described
29
U(S_C,(S, - S.),S,,C,(S,- S,) S, C,(S,,
32
earlier.
30
30. Even when the flow exceeds the long-run sustainable level, it will not be optimal to reduce the flow
33
40.For small taxes, these are "compensated" taxes, and have no welfare effect, though they have a substitution
31
instantaneously, unless - or even when - there are zero costs of adjustment.
34
effect, and therefore do reduce pollution.
32
33
For the atmosphere, a sustainable stock of greenhouse gases means stable concentrations Current emissions are
35
41.The loss in welfare (ignoring the benefits from reduced greenhouse gas warming) are the Harberger triangles,
34
estimated to be about twice emissions consistent with stable concentrations.
36
and can thus be shown to be proportional to the product of the elasticity of demand for energy and the share of
37
energy in national output. Since poorer countries are likely to have less access to alternatives which increase the
35
31. Postponing action may lead to some irreversible damages, for example the flooding of low-lying states.
38
elasticity of demand, and since the share of energy is larger in richer countries, the burden of the tax is
39
progressive.
36
32. This is not quite correct, since price affects Incentives for exploration, and some marginal wells would not
37
be drilled if oil prices fall too low.
40
42. The polluter pays principle endorsed by the OECD is exclusively prospective.
38
33. For analyses of market and optimal responses to uncertainty about the arrival of backstop technologies, see
39
Dasgupta, Gilbert, and Stiglitz
and Desgupts and Stiglitz
Nash, J.F. The Bangaming Problem. "Exametrica 155-62. 18,
45
46
I
43.Using either an egalitarian social welfare function approach, or a Rawlsian analysis "behind the veil of
I
2
ignorance" (Rawls 1971) leads to the rejection of the "polluter pays principle Since al the time the relevant
52. Only a few models take into account international capital flows. Thus, most models do not address issues
2
)
actions are taken, the polluter is not cognizant of the effects, such fees have no incentive effects, but rather appear
of industry relocation (McKibben and Wilcoxen, 1992). Chapter II provides a more complete discussion on
3
4
as random taxes, lowering each person's expected utility, and in particular the expected utility of the worst-off
leakages.
5
individual.
4
6
53. In the case of production of highly substitutable commodities, carbon leakage will, of course, be much greater.
,
7
There is a further ethical issue: ascertaining who the true beneficiary of escaping paying for the pollution is
8
generally difficult. It need not be the individual, firm, or country actually engaging in the externality generating
6
9
activity; in competitive markets, when firms are not charged the full social costs of production, product prices
54. Whether taxes in fact have this effect depends in part on the shape of the demand curves. With intertemporal
,
10
will fall, giving consumers 8 substantial fraction of the benefits
separability in demand curves and constant elasticity, with no backstop technology, a constant ad valorem tax has
I
no effect on the pattern of consumption.
11
9
44. Manne and Richels (1992) show that, under the IPCC emissions scenarios, even the most drastic controls on
12
10
emissions from developed countries would be insufficient 10 stabilize greenhouse concentrations without some
Coal presents markedly different issues, not so much because of its greater emissions per unit energy, but because
13
II
means of controlling emissions from developing countries.
of in higher cost of extraction to price ratio. Lowering producer prices may result in less coal being consumed,
12
provided alternative energy sources become available. Thus, taxes on coal are likely to have significant general
14
13
Public goods exist when property rights are not or cannot be clearly assigned. The atmosphere is an
equilibrium as well as partial equilibrium effects; the increase in the price of coal will lead to a substitution of
15
14
international public good because assigning property rights to the atmosphere is difficult for one nation acting
gas and oil. If alternative energy sources are not available, such policies will only affect the intertemporal timing
16
15
alone, and particularly difficult when many sovereign states must agree among themselves. Tradable greenhouse
of coal consumption (given the much more limited resources of gas and oil). But even that might be of some
17
16
emission permits, discussed below, attempt to resolve the problem by explicaly assigning property rights to
value in reducing long run greenhouse gas emissions, as the ability to extract energy from coal may increase
18
17
greenhouse emissions
significantly over time. Analyzing the optimal intertemporal structure of taxes, to minimize long run ambient
19
IS
levels of greenhouse gases, taking into account both intertemporal substitution and substitution across energy
19
sources, is a complicated technical issue, that to due has not been adequately analyzed.
20
46 See Abrew
20
55. United Nations Conference on Environment and Development, Framework Convention on Climate Change,
21
21
47. In some cases, equity considerations may prevent Couse (1960). in hrs discussion of externalities, emphasized
May 9, 1992.
22
the separability between efficiency and equity issues Though there have been several important qualifications
23
22
to Coase's conjecture, emphasizing the importance of public goods, imperfect information, and transaction costs
56 The most difficult problem is posed by those investors who invested in these resources, on the basis of one
24
(see Farrell
23
Stiglitz (1988)). still, the basic ensight remains applicable here
regime (where these resources were not taxed). Do they have any special claim to compensation for a "change
24
in regime." Changes in demands and supplies occur for virtually all resources, and are an inevitable part of the
25
48. Chapter 7 discusses this issue at greater length
25
risks in investing. While most economists would argue that arbitrary and capricious changes in policies contribute
26
to business uncertainty, and therefore have an adverse effect on economic growth, reasoned changes in policies
26
49. These concerns are not just theoretical possibilities, as the following two examples illustrate Assume that
27
in response to changes in information are an inevitable part of the business risk.
27
the North imposes high energy taxes, but the South fails to do so Energy intensive industries, such as aluminum,
28
migrate from North to South. But energy efficiency in the south is much less than in the north, so that the total
28
57. For example, some U.S.A. electric utilities are already making decisions in anticipation of some future policies
29
29
energy used to produce a ton of aluminum could increase substantially. While economic efficiency would call
to limit greenhouse emissions.
30
for locating energy intensive industries where energy efficiency is greatest, a system of partial controls would
31
30
results in energy intensive industries being located where energy efficiency is lowest. Similarly, the reduced
These issues also arise among countries: countries with large coal deposits will find the value of their natural
32
31
energy consumption by the North will result in lower producer prices of oil and gas. leading to increased
wealth eroded, and quite naturally will be less enthusiastic about international agreements having that
33
consumption of energy in the South, partially offsetting any energy conservation induced in the North
32
consequence.
34
50.The precise manner in which this should be done is a technical matter, treated in the literature on Global
33
59. The studies show a marked variation in GDP losses across models. For example, stabilizing emissions at their
35
Warming Potential (see, e.g., IPCC 1990). To the extent that there are large differences in atmospheric lifetimes,
34
1990 levels is estimated to reduce U.S. GDP by 2% to 8% in the year 2010 roughly a $20 billion to $80 billion
36
then the relative weighting of different greenhouse gases should change over time, since the "shadow price"
35
loss for that year. Estimate of the costs of reducing emissions by 20% below levels in the year 2010 range
37
associated with effects on relative concentrations at different dates will differ.
36
from 9% to 1.7% of GDP. Aggregated models (top-down) have generally reported higher costs, while
37
disaggregated models (bottom-up) have shows lower costs. Chapter 9 contains a more complete discussion.
38
51. Similarly, if the developed countries restrict forest cutting, the price of lumber may rise, inducing the
38
39
developing countries to cut down more of their own trees. While total carbon sequestration may not increase,
39
These GDP losses occur when the carbon taxes lead to investments that are more expensive than those that would
40
environmental and economic efficiency will decrease if, as some researchers have concluded, hardwood forests
40
take place in the absence of the taxes. The higher the carbon taxes, the greater the investment in price-induced
41
in the less developed countries may be the least desirable ones to cut down from an ecological or economic
41
conservation and the more fuel switching toward the less carbon-intensive substitutes.
42
perspective (Edmonds and Reilly, 1987).
42
43
The overall impact of a carbon tax will depend not only on the size of the tax but also on the uses to which the
44
revenues are put. In the standard EMF scenarios, is was assumed that tax revenues will be redistributed in 8
45
neutral manner (i.e., without affecting the marginal tax rates). There are, of course, numerous ways in which tax
CK
ypons
47
48
I
67 This issue has been addressed in several papers. See, in particular, Stiglitz [1975, 1984] and Grossman [1981].
I
revenues can be used. These include: reducing budget deficits; reducing marginal rates of income, payroll,
2
corporate or other taxes: granting tax Incentives to preferred activities, or increasing the level of government
2
68 This is both because those who would, under "true" disclosure, be of 0 competitive disadvantage have an
3
expenditures. The costs of the tax will vary widely depending on how the revenues are recycled
)
incentive to add "noise" and because there are strong market forces for product differentiation; in markets with
4
homogeneous commodities profits, even with a limited number of suppliers, will be driven to zero (in Bertrand
4
60.Top-down models estimate that for developing countries there. exist low-cost options to reduce emissions in
5
competition.) For a discussion of these and related issues, see Salop [1977] Salop and Stiglitz [1977, 1982],
5
the near term. but eventually costs would exceed I to 2% of GDP (EMF 1993) For economies in transition,
6
Stiglitz [1977, 1988].
A58
6
because of historical inefficiencies and energy subsidies, there exist large opportunities to reduce emissions at
7
little or no cost. For developing countries, problems of informal economies make hard estimates difficult, but
,
The standard reference in the organizational literature is March and Simon Economic theories emphasizing
I
the cost of stabilizing emissions would likely be large enough to cut into economic growth
8
the non-value maximizing behavior of managers include the works Baumol
Marris
and
Leibenstein
9
1. The principal agent literature [Ross, 1973, Stiglitz, 1974] provided the foundations for
9
61. Recent comparisons indicate that the most important differences between top-down and bottom-up models arise
10
understanding the divergence: of interests. See Stiglitz [1988]. A more recent overview is provided by Stiglitz
10
from differences in input parameters, rather than from differences in model structure.
II
and the symposium in the Journal of Economic Perspectives,
11
Government institutions and regulations often hinder the efficient use of energy. The developing countries
This -MEJEP key point, while noted in March and Simon's (1958) original work, was elaborated upon by Rednet
Main,
12
12
are least able to absorb the costs of these inefficiencies. Thus, while some developing countries argue that they
13
Hannaway
cannot afford to reduce greenhouse emissions the same countries have the most to gain from reforming
Capvalism.
13
14
government-caused inefficiencies. At least in the short nm, International agreements committing countries to
eliminate at least the most egregious of these practices might go a long way to addressing the problem of
14
71. The facts that time is a scarce commodity and that decision making in large organizations decentralization
15
15
do not in themselves constitute a market failure; they do not prove that resources are not efficiently allocated
16
emission reductions,
16
given the real constraints facing society, which include time. However, Greenwald and Stiglitz [1986, 1988] have
17
17
established a very general theorem showing that when information is imperfect and costly, market equilibrium
18
is, in general, not (constrained) Pareto efficient. Thus, there is no presumption concerning the efficiency of the
IS
63 This may also be a problem with electricity generated by the private sector, as regulation has historically any
19
market economy, even in the absence of the kinds of externality and public goods problems that are associated
19
price equal to average cost, rather then allowing in to meach the competitive price. In many countries, the increase
20
with greenhouse gases. For a more extended discussion, see Stiglitz (1994) not in publication list.
20
of competitive pressures has moved electricity prices closer to the merginal cost of production
21
22
One of the main insights of recent advances in the economics of Information is to provide a sound
21
64.Full unilization of non-fossil ford energy sources. taking account of other environmental impacts For example.
when hydroelectric power generation, which does not greenhouse CURISSIONS, can be cost-effectively
23
micro-foundations for these theories of the firm. And indeed, the importance of the limitations on the availability
22
24
of information, and the consequent importance of attention directing efforts applies to individuals as well as to
23
expanded without other environmental effects, # should be done
25
organizations. Some studies have suggested that the limited success of the special tax provisions in the United
24
25
Eliminating Regulations Impeding Efficient Energy Unitization Many. perhaps most, countries have a host of
26
States designed to encourage savings (IRA accounts) was primarily due to the competitive efforts of banks to
27
recruit these accounts, and the attention which savings got as a result.
26
regulations which increase energy use as they impode aconomic efficiency For asstance, the United States has
28
27
had 8 policy of restricting oil exports to Alaska Whatever the merits of that policy. N has forced Japan to import
28
oil from Indonesia and Saudi Arabia. World oil transportation costs have thus been greatly increased. at the
29
72 Network externalities are manifested in other ways: builders fail to install energy efficient light bulbs, because
29
expense of the American economy. Another example of government reform, included IN the US' Action Plan
(1993). encourages efforts to expand and improve natural gas markets through continued regulatory reform These
30
customers dislike them, because stores do not carry replacements; and stores do not carry them because the
30
31
demand for them is too low.
31
reform efforts include guidelines to allow greater natural gas use in the summer in coal- and oil. fired power
32
32
plants.
33
When there are important network externalities, market equilibria are frequently inefficient. The economy
33
34
might, for instance, get "stuck" in the wrong equilibrium. Government action can, in these instances, "force" the
34
Other regulation. Unintended effects of many tax. expenditure and other policies have contributing further to
inefficiencies in land use. Among the unfortunate effects of the U.S. Superfund program for the management of
35
economy to move from one equilibrium to another.
35
36
36
hazardous wastes has been the creation of large unoccupied holes in the center of major cities.
37
73.This is not the only explanation of differences between bottom-up and top-down models. There are several
37
65.Consider the following thought experiment: compare an optimally designed road system which only carried
cars; and contrast that with an optimally design road system which also carries trucks. The incremental cost of
38
other features of market behavior that bottom-up models often ignore.
38
carrying trucks is, in most countries, much larger than the proportionate share of the cost they bear in gasoline
39
39
40
(a) Hidden Costs: Consumers value a range of attributes difficult to include in an engineering model. For
40
taxes and other fees.
41
example, auto buyers value not only initial costs and fuel economy (which computer models can easily calculate).
66.For example, in many countries, governments have taken an active role in the dissemination of Information
42
but also performance, safety, and durability, which they typically do not.
41
to the agriculture sector. These programs are in some measure responsible for the large increase in agricultural
43
42
44
(b) Divergence between laboratory and in-use performance: Especially for new technologies, actual energy use
43
productivity in countries with agricultural extension services.
45
often differs significantly from energy use calculated in the laboratory. It is the latter upon which purchasers
44
50
"
I
I
focus.
83. Analysts now use two methods to estimate stocks. The first assumes a fixed stock of a natural resources such
2
2
as oil. Consumption of oil then depletes the stock by the amount of consumption. The second begins by treating
3
3
(c) Variation across individual CORSUMERS Engineering models generally assume an average consumer, actual
discovered reserves as the asset. Thus, additions to reserves increase the asset, while consumption reduces it.
4
4
consumers may display a wide range of characteristics and usage patterns. Except when demand functions are
If in any given year, new discoveries match resource utilization, then according to this method no net depletion
5
has occurred.
5
linear in the relevant variables, the consumption of the "average" individual is not equal to the average
6
consumption; and what is optimal for the average person may not be optimal for a significant fraction for the
7
population.
0
For
a
survey,
see
JES 1981 Credit Ratiming Mkh of lufo Impartate 8
84. Many researchers have noted deficiencies in standard national income accounts. First, national Income
accounts do not, in general, provide an adequate measure of welfare; second, they do not provide the correct
Juffer
and
Stiglitz
The basic theory of credit rationing was developed inSuglitz
information relevant for making policies relevant to sustainable development. Sustainable development is
Weigs [1981] and the theory of credit retioning is developed in Greenwald, Stightz and Weiss (1984) and Myers AER 71(3)10
9
9
concerned with society's resources; an economy is growing when its resource base (capital stock combined with
10
and Maliuf 11284
10
natural resources) is growing GDP does not, and is not intended to be, a measure of resource availability. Firms
tere 395-1710.
have two sets of accounts-cash flow (income) statements and balance sheet statements. GDP is . statement of
"
75.This generally accepted methodology is, for instance, reflected in the guidelines issued by the Office of
the former type.
12
Management and Budget in the United States for the evaluation of projects and regulations. The applied Interature
13
13
does not address the question of whether this procedure is appropriate in the presence of certain types of time and
14
Standard accounting procedures require that firms, in an attempt to present an accurate account of "true income,"
14
risk non-separabilities.
15
take account of depreciation. GDP measures gross output; it does not take into account depreciation, either of
16
natural or physical capital stocks. The reason is simply that it is hard to get accurate measures of depreciation.
15
76. Though if the variance of the net benefits is increasing over time in a particular manner, the differences in the
17
16
18
two methodologies may not be large.
Net national product does take into account depreciation, the change in capital stock. And it is this account which
19
should be most subject to criticism, since it accounts for changes in the physical capital stock, but not in other
17
77 Again, under certain restrictive conditions, where the shadow value of a capital constraint is changing
20
capital assets, in particular, environmental assets and natural resources.
18
systematically over time, the differences - the two methodologies may not be great.
21
22
A number of difficult conceptual problems face the analyst defining levels and changes in levels of these assets.
19
78
See
Vilson
Though the original discussion of winners curse focused around bidding in auctions, it has
23
20
subsequently come to be applied to a range of other market phenomena See. " Stightz I I
24
25
First, how should the "stock" of natural resources be defined? Coal poses perhaps the easiest situation. The
21
79.For a discussion of the role of the state in capital markets. are Stiging 119941
26
location of coal reserves is known. Costs of extraction are high, so the rents (the value of coal in site) is low.
27
The depletion can be measured not by the coal used times the market price, but the coal used times the in site
22
80.1n some industrialized countries, energy efficient home mortgage lending may help correct the problem
28
value. But for oil and other minerals, information about where reserves are located is vital. Two models have
23
Lenders generally set criteria for the maximum loan amount based on the borrowers' ability to repay, which in
29
been proposed. One sees the world as having a fixed stock of natural resources (say oil). When one uses oil,
24
turn depends on income and wealth. The fact that a particular expenditure which would enhance efficiency and
30
one is depleting this stock. Thus, to calculate the value of depletion, one does not need to know the entire stock;
25
reduce utility bills is not given special attention. Energy efficient mortgages provide funds to households to
31
the flow (the amount of oil consumed) provides an accurate measure of the change in stock.
26
make energy efficiency enhancing investments intended to pay for themselves, ie. result in reduced utility bills
32
27
equal to or greater than the interest payments. With capital construints, builders may have an incentive to trade
33
The alternative model looks at the size of discovered reserves. Reserves are treated as the asset. Additions to
28
off initial capital costs for higher maintenance costs (lower energy efficiency). Building codes specifying minimal
34
reserves thus are viewed as increasing the resource base. If in any given year, new discoveries match resource
29
levels of energy efficiency and full disclosure of expected life cycle energy costs may help address these market
35
utilization, then there is DO net depletion. This is the approach being taken by the U.S. Department of Commerce.
30
distortions.
36
This accounting framework would be correct if there were an infinite supply of the resource (reflected in zero
31
37
rents); the essential "capital" good is information about where the resource is located.
32
38
39
Environmental assets-such as air quality-present another set of problems, because there are not market prices
33
81. A country that rapidly depletes its natural resources may show a high rate of growth under conventional
40
to value the asset. Dynamic optimization problems of the kind described earlier can be used to calculate shadow
34
income accounting. but a lower rate of growth when resource depletion is taken into account. Repetto (1991,
41
prices. How sensitive these shadow prices are to specific assumptions remains to be investigated.
35
1992) calculated resource-adjusted GDP for several countries rapidly harvesting their stocks of hardwoods and
42
36
other resources, arguing that conventional measures sharply overstated GDP.
43
Accounting systems do not, however, have to aggregate all information together. Just is Information about
44
longevity and other Indicators of well being (see below) serve to complement information from national income
37
82.Cobb and Daley (1989) have even claimed that U.S. per capita GDP, when adjusted for environmental damage,
45
accounts concerning standards of living, so 100 can information about physical environmental measures be used
38
was stughent 1950 and 1986. This assertion is hard to reconcile with the steady improvement in most
46
to complement information from the extended national income accounts.
39
measures of environmental quality since 1970, when measurements standards were established.
47
85. There is some concern that excessively broad and long patents may actually impede innovation. When
48
technological progress occurs by building on previous innovations, later Innovator require permission of earlier
49
innovators to realize the returns on their innovation. While advocates of broad patent coverage argue that the
51
52
1
parties always arrive at efficient bargaining solutions, critics point out that the outcomes of bargaining models
I
simpler emissions permit system (EPS) issues permits on the basis of source emissions and ignores what effects
with incomplete information often entail large inefficiencies
2
2
those emissions have on the receptor points. Within a given region or zone, the polluter would have only one
3
market to deal with and one price. Finally, there is the pollution offset (PO) system wherein the permits are
3
86. Matters are more complicated, since the patent does not reward the innovator with his marginal contribution-
4
defined in terms of emissions, trade takes place within a defined zone. However, the standard has to be met at
4
the increase in the present discounted value of benefits as a result of the innovation occurring earlier than it
5
all receptor points. The exchange value of the permits is then determined by the effects of the pollutants at the
5
otherwise would have occurred. For a fuller discussion, see Stigling as 1. Desgupts and Stightz 1980a. 1980bl.
6
receptor points. The PO system thus combines characteristics of the EPS and the APS. (Pearce and Turner
6
Barzel I 1
7
1990) These distinctions are of limited relevance for greenhouse gases, where what is of concern is global
B
emission levels. The specific location of the emissions is of no concern.
,
87.
If less developed countries fail to implement fully a set of corrective taxes or tradeable permits, or if less
,
8
developed countries fail to adopt and enforce effectively intellectual property rights, there will be insufficient
,
incentives to produce energy and emission savings innovations, particularly those appropriate for the level of
10
90. The choice between taxes and tradeable permits depends on the objectives of the policy maker and nature of
10
technological knowledge, human capital, and factor prices in those countries. If less developed
II
the uncertainty about the marginal cost and marginal benefit curves for carbon emission reductions (Weitzman,
"
countries do take these actions, there is concern that they will result in higher prices of innovations, and thus the
12
1974). Theory tells us that if the nature of the curves is known with very little certainty, but the marginal cost
12
pace of adoption will be retarded. An effective form of aid, targeted to reducing greenhouse gas emissions. may
13
curve is known to be relatively steeper (i.e. a change in the level of pollution allowed brings about a greater
13
be subsidies directed at producing appropriate energy saving and emission reducing technologies for LDCs.
14
change in the marginal costs of mitigation compared to the marginal benefits) then taxes should be the policy of
14
15
choice. This is because, in this case, an erroneous estimation of the optimal tax rate will lead to a relatively small
16
deviation from the optimal pollution level. On the other hand, an erroneous estimation of the optimal level of
13
$8 Edmonds of al (1994) has studied the importance of available advanced energy technologies such as those
17
total emissions in a permit scheme will lead to a relatively large deviation from the optimal cost of the permits.
16
proposed by Johannson (1993) Edmonds " of use the Edmonds-Resily-Rarnes model for energy related
II
17
greenhouse gas emissions. the MAGICC model for atmospheric composition, climate response. and see level me,
19
If the marginal benefit curve is known to be relatively steeper than the marginal cost curve, however, tradable
IS
the IPCC scenario 1592a (IPCC, 1992) as the reference base case and five alternate energy scenarios that are for
20
permits are the better option. Here, an erroneous estimation of the optimal tax rate will lead to a relatively large
19
more advanced over today's energy supply and transformation technologies The five energy scenarios are
21
deviation from the optimal level of emissions while an erroneous estimation of the optimal level of emissions in
20
22
a permit strategy will lead to 8 relatively small deviation from the optimal cost of the permits.
21
a advanced fossil fuel technologies
23
22
b. advanced liquified hydrogen fuel cells
24
la the case of greenhouse gas emissions, the time horizon for adjustment is sufficiently long that many of these
23
c. advanced hydrogen fuel cells without liquified hydrogen
25
uncertainties become less important. If the tax rate initially chosen yields too high I level of emissions in one
24
d. low cost biomass
26
year, it can be increased, and the net impact of the erroneous initial estimate on global warming (or the total cost
25
c. accelerated rate of exogenous end-use energy intensity improvement
27
of achieving 8 given level of atmospheric concentration) will be negligible. In any case, as the earlier discussion
26
28
on sequential decision making has emphasized, there is likely to have to be continued revisions in either tax rates
27
Combined, the energy technologies reduce annual conssions from fossil fuel use to levels that stabilize
29
or permit levels.
28
atmospheric concentrations below 550 ppmv (ie double the concentration prior to the Industrial Revolution)
30
29
The tax rate, assumed to apply globally, used was the marginal cost of stabilizing fossil fuel carbon emissions
II
30
in the reference case. With values reflected for only carbon dioxide emission reductions, the cost of global
32
Still, there is some argument that the required adjustments under a permit scheme may be less burdensome.
31
emission reductions grow from approximately $35 (US) Billion in 2005 to $230 (US) billion per year in the year
"
(Tistenberg 1992) For instance, if the authority feels that the old standard needs some tightening they may enter
32
2095. With advanced fossil fuels, low cost solar electric power, low cost fuel cell vehicles, the present discounted
34
the market themselves and buy some of the permits, holding them out of the market.
33
value of adding low cost biomass fuel to the energy technology bundle is almost half a trillion dollars (US) The
35
34
present discounted value of the advanced energy technologies taken together 19 SI $ trillion (US)
36
There must be effective, competitive markets in tradeable permits, if such schemes are to achieve efficient
35
37
outcomes. There are transactions costs of running such schemes, just as there are transactions costs associated
36
The introduction of advanced biomass energy production technology was found to play a key role in reducing
38
with collecting tax revenues. Whether transactions costs gives one system a decided advantage over the other
37
emissions. Biomass energy at $2.00/GJ growing to become the core energy supply technology by 2050 could
39
is not clear. There seems to be no compelling reason to believe that good markets in tradeable permits would
38
significantly reduce emissions highlighting the potential role of technology development and deployment relative
40
not develop.
39
to that of fiscal and regulatory intervention.
41
40
42
41
These results should be viewed as illustrative rather than predictive. in this analysis, the gains from introduction
42
and deployment of advanced energy technologies is dependent on the order of technologies evaluated in the study.
43
91.11 is possible to design allocations of trading permits which (i) on average, impose no net burden on developing
43
44
countries (thus conforming to the ability to pay principle); (ii) provide those economies which are growing faster
44
45
per capita with commensurately greater permits, thus imposing no net drag on economic growth, provided the
46
economy exhibits an increase in fuel efficiency at least equal to the average of fast growing LDCs; and (iii)
45
89. The literature has identified three types of permit systems. The ambient permit system (APS) works on the
47
rewards those economies which are able to reduce greenhouse gas emissions faster than benchmark, either through
46
basis of permits defined according to exposure at the receptor points. Each polluter, then, may face quite
48
greater control of population growth, through larger increases in energy efficiency, or through switching from
47
complex markets different permit markets according to different receptor points, and hence different prices. The
49
higher to lower carbon fuels.
50
53
54
I
The extent to which individual circumstances of countries should be taken into account in setting benchmarks
I
For a 45% reduction in baseline emissions by 2020, the required tax would be in the range of $150-$325 per ton
2
remains a question for international negotiations; to the extent that high emissions is due to natural endowments
2
of carbon and the cost might be in the range of 1.5%-2.9% of world GDP. A 70% reduction in baseline
3
(e.g. the availability of cosl rather than natural gas as a source of energy). 0 persuasive case can be made for
3
emissions by 2050 could require a tax between $230 and $880 and a loss in world GDP of 2.4%-3.8% (Dean
4
benchmarks to reflect initial emission levels To the extent that high emissions are due to inappropriate energy
4
1994).
5
pricing policies, the case that benchmarks should reflect initial emission levels is for more tenuous
5
6
The required carbon taxes and associated costs vary significantly across regions in all of the models. This
6
92. Similarly, some developing countries have asked, should the North be given higher levels of permits. simply
7
indicates that the same proportional reductions in emissions across all regions would give rise to very different
1
because is has, in the past, been the chief source of greenhouse gases?
$
costs in different regions and would thus be globally inefficient with great potential for savings in the global
9
cost of reducing emissions through the use of emission trading between countries or regions (see section 1.5.2.1)
8
93. For a discussion of the polluter pays principle, see above.
10
or a global carbon tax.
11
,
94. Standard tradeable permit schemes essentially take the revenue from a carbon tax. and distributes it to current
12
Three insights emerging from the OECD study are (Dean, 1994):
10
user emitters, rather than using the revenue to reduce other taxes. An alternative to these standard schemes is
13
a. Small amounts of emissions reduction can probably be achieved with low taxes;
11
for the government to auction off the tradeable permits.
14
b. Large reductions can only be achieved at high tax rates (i.e. marginal reduction costs rise with emissions
12
15
reductions).
13
If taxing carbon leads to reduced labor supply or reduced savings, then government revenues from wage or capital
16
c. Carbon-free backstop technologies are likely to slow the rise of the carbon tax, or halt k altogether, if they
14
taxes may be reduced, more than offsetting the direct revenue gain from the carbon tax. Cross elasticities of this
17
are available at constant marginal cost.
15
magnitude are unlikely, though any such cross elasticity will reduce the net gain from the carbon tax. The
IS
16
magnitude of the double dividend has been the subject of some dispute, with Goulder 11 and taking opposite
19
Energy Modeling Forum Project 12 (Impact of Carbon Emission Control Strategies)
17
views.
20
A recent study at Stanford, the Energy Modelling Forum Project 12, examines the cost of reducing CO2
18
21
emissions (Energy Modeling Forem, 1993). A diverse group of economic models, employing common
22
assumptions for selected numerical inputs, were used to analyze a standardized set of emission reduction
19
95. Two main studies provide insights into the root of the variance in estimates of the economic effects of carbon
23
acenarios. In all, 14 top-down models participated in the study.
20
taxes: The Energy Modeling Forum Study (12) and the OECD comparison project. in each case, sophisticated
24
21
sensitivity analyses are run by standardizing key economic assumptions along with use of common reference case
25
The EMF model comparison provides the most comprehensive application of top-down methodologies to date.
22
scenarios of reductions. The magnitude of the effect on economic growth will depend both on assumptions
26
The study addresses 8 wide range of policy questions. How large are emissions likely to grow in the absence
23
concerning the effect of carbon taxes on savings and labor supply. and the induced investment to offset the higher
27
of controls? How much market intervention will be required to meet alternative targets? What will be the price
24
energy prices. If higher energy prices do not lead to much capital substitution and of the cross elasticity with
28
lag? In exploring economic costs, the modelers were asked to examine the Impacts of timing, research and
25
savings and labor are low then the likely effect on economic growth will be small
29
development, and revenue recycling.
26
30
27
31
The EMF exercise provides a wealth of useful information for policy making. Although the focus was primarily
28
OECD Project on Economy-Wide Cost Estimates of Carbon Taxes
32
on the U.S., many of the insights are applicable to developed countries in general.
29
An OECD model comparisons project was conducted to compare economy wide estimates of the effects of carbon
33
30
taxes. Time horizons as well as the key economic assumptions on growth, population and resource prices, and
34
In selecting parameters for standardization, the EMF study focused on what were felt to be the most influential
31
the reduction scenarios for six global models were standardized. The global models compared were the GREEN
35
determinants of mitigation costs. These included: GDP, population, the fossil-fuel resource base, and the cost
32
model, the IEA model and four North American models [Edmonds-Reilly Model (ERM). Global 2100 of Alan
36
and availability of long-term supply options. la addition, although the EMF models differed considerably in their
33
Manne and Rich Richels (MR). the Carbon Rights Trade Model (CRTM) of Tom Rutherford. and the Whalley-
37
technology representation, the study attempted to impose uniformity with regard to world oil prices, the oil and
34
Wigle model] (Deas 1994).
38
gas resource base, and the cost of backstop technologies. For its reference case, EMF adopted the average of the
35
39
IPCC high and low economic growth cases (IPCC, 1990c). Also for consistency with the IPCC, the study
36
There is significant variation in tax rates and costs for the same amount of emissions reduction among models
40
adopted the population growth projections based on Zachariah and Va.
37
due to differing assumptions regarding several key considerations. Several factors explain the differences between
41
38
model results. The most important factors are:
42
The modelers generally used taxes based on the carbon content of the fossil fuels in order to achieve a prescribed
39
43
emissions reduction. The magnitude of the tax provides a rough estimate of the degree of market intervention
40
a. The degree of substitution between fucts the ease with which producers and consumers can switch from
44
that would be required to achieve the carbon emissions target. Estimates range from $20 to $140 per ton for the
41
high-carbon content fuels to luw-carbon content fuels;
45
carbon taxes required to hold emissions at 1990 levels in 2010. Estimates of the carbon taxes required to reduce
42
b. Expectations about future energy prices and taxes;
46
emissions by 20% below 1990 levels in 2010 range from $50 to $330 per ton.
43
c. The speed of emissions reduction;
47
44
d. The way in which revenue is recycled;
48
Two parameters are particularly important in explaining the differences in tax projections: the price elasticity of
45
c. The treatment of the removal of energy subsidies; and
49
energy demand and the speed with which the capital stock adjusts to higher energy prices. Neither were
46
f. Assumptions regarding backstop technology and a host of other technical and economic factors.
50
controlled in the EMF experiments. Those models using lower price elasticities required higher taxes to achieve
47
51
the same emissions goal. Those models which assumed greater malleability of capital required lower taxes.
48
Because of the varying approach to these questions, the range of estimated tax rates and costs are quite wide.
55
56
I
1.9
REFERENCES
I
2
2
96. Rapid capital stock retirement may add to the cost of immediate CO2 reductions Also, reductions may he
1
Abreu (**date**)
3
cheaper later because of technical changes in the intervening years
4
,
97. Indeed, the "ratchet effect," 8 commonly observed phenomens in the command and control economics
Arrow 1962. Learningly Doing," reprented in Anow KJ
4
6
5
(Stiglitz, 1975, Weitzman, 19 ) leads firms to just satisfy the target.
1984 Pagers 1 KM. Camb Mass: Bellevag Presed
7
Ausubel, J.H. 1996, Conventional Wisdom About Impacts of Global Change. Rockefeller
6
98. Two examples from the United States Clean Air Act. 1) the requirement that coal-burning power plants
8
University, New York. ?pp.
7
install stack-gas scrubbers even if they burn low-sulfur Western coal; 2) the mandates for ethenol If reformulated
Haward a.
9
$
gasoline.
10
Baumol (**date**)
9
99. World Commission on Environment and Development, Our Common Future. Oxford Oxford University Press,
II
12
Barzel (**date**)
10
1987.
11
"
100 It has been argued that new species and varieties produced by genetic engineering may be able to more than
14
Bongaarts, J. 1994. "Population Policy Options in the Developing World," Science (263),
12
offset the loss in genetic variability from climate impacts, particularly because of the increasing capacity to direct
15
11 February 1994, pp. 771-776.
13
those mutations toward socially desirable objectives
16
"
World Commission on Environment and Development 1987. Our Common Future.
14
101 Other views represented in the debate on sustainable development but not generally accepted in the
economics profession include those of Dely (1977. 1980) and Dely and Cobb (1991). and those of the new field
18
London: Oxford University Press.
15
16
of ecological economics (Costanza 1991).
19
20
Burnisux, J.M., Martin, J.P. and Oliviers Martins, J. 1992. 'The effects of existing
17
102 Tanq Ownen Hyder, "Clunate Negotiations the North/South Perspective," in Mintaer, ed Hyder has headed
21
distortions in energy markets on the costs of policies to reduce CO2 emissions: evidence
18
the Pakistani delegation at meetings of the INC. and GET (Global Environment Facility)
22
from GREEN.' OECD Economic Studies No. 19
103 "Principles of the G.77 and China OR the Chemic Convention." red - Tanq Owner Hyder. "Climate
23
19
Negotiations, the North/South Perspective," - I od. Confronties Closses Change Roke implications.
24
Chso, Hung-po, 1992. "Managing the Risk of Global Climate Catastrophe," Electric Power
20
21
and Responses Cambridge: Cambridge University Pres. 1992
25
Research Insitute.
26
27
Cline, W.R. 1992. The Economics of Global Warming. Institute for International
28
Economics, Washington, D.C.
29
30
Casse (1990)
31
32
Coase (1960)
33
34
Dasgupta, Gilbert, andStiglitz (**date**)
35
Athinson AB + stighty JE lectures Reldin Economics. M: McGH
36
Dasgupts, and Stiglitz (1980a, 1980b)
37
38
Dasgupts, P. and K.G. Maler 1990. The Environment and Emerging Developing Issues,"
39
Proceedings of the World Bank Annual Conference on Development Economics, pp.
40
101-131.
41
42
Dean, A. 1994. 'The Effectiveness of Carbon Taxes at the International Level.' In:
43
Climate Change: Policy Instruments and their Implications: Proceedings of the Tsukuba
44
Workshop of IPCC Working Group III, 17-20 January, 1994. pp.46-59.
57
58
1982
Dasyupta P., The Central 1 Beamer Lambudge Mt: HUP.
7
Dowlatabedi H., and MGM 1993
I
Dixem et al. (1994)
-
Hannaway (**date**)
2
2
3
Edgeworth, Mathematical Physics,
)
Heal and Chichinisky (**date**) which?
4
4
,
Edmends, J., Wise, M. and MacCracken, C. 1994. 'Advanced Energy Technologies and
,
Hoffman, J.S., Keyes, D. and Titus, J.G. 1983. Projecting Future Sea Level Rise:
6
Climate Change: An Analysis Using the Global Change Assessment Model (OCAM).'
6
Methodology. Estimates to the Year 2100, and Research Needs. US GPO 055-000-0236-3.
1
Global Environmental Change Program, Pacific Northwest Laboratory. Wahington D.C.
1
GPO, Washington, D.C.
$
Draft
8
,
9
Hoffman, J.S., Wells, J.B. and Titus J.O. 1986. 'Future global warming and sea-level rise.'
10
Energy Modeling Forum 1993. EMF 12 Global Climate Change: Impacts of Greenhouse
10
In: Iceland Coastal and River Symposium. G. Sigbjarnason (ed) National Energy Authority,
II
Gas Control Strategies. Palo Alto, Ca
II
Reykjavik.
12
12
13
Environmental Protection Agency (EPA) 1989. 'Sea level rise.' In Report to Congress
11
Hotelling. H. 1931. "The Economics of Exhaustible Resources," Journal of Political
14
on the Potential Effects of Global Climate Change on the United States.
14
Economy.
15
15
16
Fair (1970)
16
Interamerican Development Bank (**date**)
17
17
IS
Fankhauser (**date**)
18 got Intergovernmental Panel on Climate Change (1994)
19
19
20
Farrell (**date**)
20
Intergovernmental Panel on Climate Change 1992. The Supplemental Report to the
21
21
IPCC Scientific Assessment. Cambridge University Press, Cambridge. 200 PP.
22
Gleick, P.H. 1987. Regional Hydrologic Consequences of Increases in Atmospheric CO2
22
23
and Other Trace Gases'. Climatic Change 10. PP 137-61
21
Intergovernmental Panel on Climate Change 1990a, Climate Change: The IPCC
24
24
Scientific Assessment. Report Prepared for IPCC by Working Group I, World
25
Goulder (**date**)
25
Meteorological Organization and United Nations Environment Programme, New York.
26
26
27
Greenwald and Stiglitz (1986, 1988)
27
Intergovernmental Panel on Climate Change 1990b, Climate Change: The IPCC
28
28
Impacts Assessment. Report Prepared for IPCC by Working Group I, World Meteorological
29
Greenwald, Stiglitz, and Weiss (1984)
29
Organization and United Nations Environment Programme, New York.
30
30
31
Grossman (1981)
31
Jaffe and Stiglitz (**date**)
32
32
33
Grubb, M. et al. 1993 The Costs of Limiting Fossil-Fuel co, Emissions: A Survey and
33
34
Analysis.' Annual Review of Energy and the Environment 18 pp.397-478.
34
Johannsen (1993) revewable Energy Sources for Fuels + Elee,
35
35
Wishington: Island Puen.
36
Grubb, M., Minh Ha Duong and Thierry Chapuis 1993. "Optimising Climate Change
36
Kalkstein, L.S. and Davis, R.E. 1989, 'Weather and Human Mortality: an Evaluation of
37
Abatement Responses: on inertia and induced technology development," IIASA Workshop
37
demographic and interregional responses in the United States'. Annals of the Association of
38
on Integrative Assessment of Mitigation, Impacts, and Adaptation to Climate Change,
38
American Geographers 79, pp. 44-64.
39
IIASA, October 1993.
39
40
40
Kane, S., Reilly, J. and Tobey, J. 1992, 'An Empirical Study of the Economic Effects of
41
Grubb, M., Thierry Chapuis and Minh Ha Duong 1995. "The Economics of Climate
41
Climate Change on World Agriculture'. Climatic Change 21, pp. 17-35.
42
Change: Implications of adaptability and inertia for optimal climate policy," forthcoming.
42
43
43
Kaya (1989)
44
Hall et al. (1993)
44
59
60
I
Konomev (1993)
I
Mendelsobn, R. 1993. Global Climate Change Impacts on Coastal Resources. Summary
2
2
Workshop Report to EPRI and NOAA.
3
Kepp. R.J. and Smith, V.K. 1993. Valuing Natural Assets. The Economics of Natural
3
4
Resource Damage Assessment. Resources for the Future, Washington D.C.
4
Meyers and Sathaye, J. 1988. Electricity in Developing Countries: Trends in Supply and
,
,
Use Since 1970. Lawrence Berkeley Laboratory, LBL-2166.
6
Larson, E.D., Ross, M.H., and Williams R.H., "Beyond the Era of Materials." Scientific
6
7
American. vol. 254, no. 6, pp.34-41, June 1986
1
Mirriees (1971)
8
I
9
Leemans, Rik and Solomon, A. 'Modeling the Potential Change in Yield and Distribution
9
Myers and Maljuf (1984)
10
of the Earth's Crops Under a Warmed Climate'. Climate Research 3. PP. 79.96.
10
11
11
Nash (**date**)
12
Leemons, Rik 1992, 'Modelling Ecological and Agricultural Impacts of Global Change on
12
13
8 Global Scale'. Journal of Scientific & Industrial Research 51, pp. 709-724.
13
Nordhaus, W.D. 1991. 'To Slow or Not to Slow: The Economics of the Greenhouse
14
14
Effect.' The Economic Journal 101, pp. 920-937.
IS
Leibenstein (**date**)
IS
16
16
Nordhaus, W.D. 1993, Managing the Global Commons: The Economics of Climate
17
Lind, R.C et al 1982 Discounting for Time and Risk M Energy Policy Resources for
"
Change. MIT Press, place? ??pp.
18
the Future, Washington DC.
IS
19
19
Orians, O., G. Brown, W. Kunin, and J. Swierzbinski. The Preservation and Valuation of
20
Lind, Robert C. 1994. "Intergenerational Equity, Discounting. and the Role of Cost-Benefit
20
Biological Resources. Seattle: University of Washington Press, 1985.
21
Analysis in Evaluating Global Climate Policy," in N Nakicenovic, WD. Nordhaus, R.
21
22
Richels, and F.L. Toth, eds., Interactive Assessment of Miliestion. Impacts and Adaptation
22
Pachauri and Khanna (1993)
23
to Climate Change: Proceedings of 8 Workshop held on 11.15 October 1993 of the Institute
"
24
for Applied Systems Analysis, Laxenburg, Austria
24
Pearce, D.W. and R.K. Turner 1990. Economics of Natural Resources and the
25
23
Environment. Baltimore: Johns Hopkins University Press.
26
Lovelock, J. 1979. Gaia: A New Look di Life on Earth London Oxford University Press
26
27
27
Pearce, D.W., E. Barbier, and A. Markandya 1989. Economics and Environment in the
28
Maler, M.G. 1991. "National Accounts and Environmental Resources," Environmental and
28
Third World, London: Edward Elgar.
29
Resource Economics 1. pp. 1-15.
29
30
30
Manne, Alan S., Richels, Richard G. 1992. Buving Greenhouse Insurance: The Economic
"
Information: An Analysis using CETA," Electric Power Research Insitute. Assessment,In
Peck, S., and Teisberg, T., 1991. "Global Warming Uncertainties
1 CETA: A Modelfa Carbon and the Value Em of Trajectory
31
32
Costs of CO, Emission Limits Massachusetts Institute of Technonlogy
32
The Peck and Engrey bynal, (1993) val. 13, no. 1, 55-27.
33
"
34
March, James, and Simon, Herbert. 1958, Organizations. New York: Wiley.
34
Polasky, 5. S. and Solow, A.R. 1994. 'The value collection of species. of an analysis
Teilburg T. L Global wanng uncertantos Journal + the value of up
35
35
36
Marris (**date**)
36
Environmental Economics and Management, forthcoming.
37
37
38
McKibben and Wilcoxen (1992)
38
Polasky, S., Solow, A.R. and Broadus, J.M. 1993. 'Searching for uncertain benefits and
SugCETA Resure
39
39
the conservation of biological diversity.' Environmental and Resource Economics 3, PP.
40
Mearns et al. (1984)
40
171-181.
41
41
42
Mendelsoba, R. and Nordhaus, W. 1994, The Impacts of Climate on Agriculture: A
42
Radner (**date**)
Energy
43
Ricardian Approach'. American Economic Review, (forthcoming).
43
44
44
Ramsey, F.P. 1928, 'A Mathematical Theory of Saving'. The Economic Journal Vol??, Economic
61
62
V.15ne.1
87-107
I
wisky 543-559. T. Michael Holine and Smah Rapar 1413. AGICE Madel
I
Schueider and Chen (1980)
2
for the Assessment I GH induced cc,
2
,
Raper H al. (1990)
1
4
,
Rawis,
John.
1971,
Theory Uux, of Justice, Consactions Belknap Press for Afimer. Rich, Boulder lolo
Sedjo, R.A. and Solomon, A.M. 1989, 'Climate and Forests' 0?
4
,
6
Sen, A.K. 1993, 'The Economics of Life and Death'. Scientific American Vol?, pp. 40-47
6
7
Reddy and Goldemberg J., "Energy for the Developing World," Scinetific American,
,
Sea, A.K. 1982, 'Approaches to the Choice of Discount Rates for Social Benefit Cost
$
September 1990, pp. 111-118.
8
,
Analysis'. In: Discounting for Time and Risk in Energy Policy, R. Lind et al.(eds).
7
,
Reilly (1993)
Resources for the Future, Washington, DC, pp. 325-53.
10
1993.
10
II
"
in
12
Reilly. J., Hohman, N. and Kane, S., Climate Change and Agricultural Trade: Who
Shah, A. and Larsen, B. 1991. "Carbon Taxes, the greenhouse effect and developing
12
13
Benefits, Who Loses?' Global Environmental Change, forthcoming.
countries.' World Bank, Washington.
sathy
13
14
14
15
Richels, R, and J. Edmonds 1993. "The Economics of Stabilizing
Solow, A.R., Polasky, S. and Broadus, J.M. 1993. 'On the measurement of biological
15
16
Atmospheric CO2 Contentrations," draft, 11/2/93
diversity.' Journal of Environmental Economics and Management 24, pp. 60-68.
16
17
17
18
Robey et al. (1993)
Solow, A.R. 1993. 'The response of sea level to global warming.' In: The World at Risk:
IS
19
Natural Hazards and Climate Change, R. Bras (ed). AIP Conference Proceedings 277, pp.
19
Romer (**date**)
38-42. American Institute of Physics, New York.
20
20
21
21
Rosentbal. DH., Gruenspecht, HK. and Moran, E.A 1994. 'Effects of Global Warming
Solow, Andrew (1992)
sdan, R. "An Almost.. View 1 Svstanable Dec
22
22
23
on Energy Use for Space Heating and Cooling in the United States Paper presented at the
RFF 1992
23
24
Annual Meeting of the Western Economics Association, July 1994, Vancouver, British
Solow, R.M. 1974. 'Intergenerational Equity and Exhaustible Resources' Review of
24
25
Columbia, Canada.
Economic Studies 1970. pp. 29-45.
25
26
26
27
Rosearweig. Cynthia and Martin Parry 1994, 'Potential Impact of Climate Change on
Steer, A. and E. Lutz 1993, "Measuring Environmentally Sustainable Development",
27
28
World Food Supply'. Nature 367, pp. 133-138.
Finance and Development, pp. 20-23.
28
29
29
30
Ross (1973)
Stewart, R. and Wiener, J.B. 1990, 'A Comprehensive Approach to Climate Change: Using
30
31
the Market to Protect the Environment'. In: The American Enterprise, Vol.1, 777
31
32
Rethenberg. J. 1993. Time Comparisons in Public Policy Analysis of Climate Change:
Schneider, S.H. and Chen, R.S. 1980. 'Carbon dioxide warming and coastline flooding:
32
33
An Economic Analysis. MIT.
physical factors and climate impact.' Amer. Rev. Energy 5, PP. 107-140.
33
34
34
35
Salop (1971)
Stiglitz (1975, 1977, 1979, 1984, 1988, 1994)
35
36
37
Salep and Stiglitz (1977, 1982)
36
Stiglitz, J.E. 1982. "The Efficiency of the Stock Market Equilibrium," Review of Economic
37
38
Studies 49 pp. 241-61.
39
Sandler, T. 1981 "Tropical Deforestation: Markets and Market Failures," Land
38
Economics, August, 225-33.
39
$
Stiglitz, J.E. 1974. "Wage Determinism and Unemployment in LDC's. I: The Labor
40
41
Turnover Model," Quarterly Journal of Economics 88, pp. 194-227.
Schmalensee, R. 1993, 'Comparing greenhouse gases for policy purposes'. The Energy
41
42
42
43
Journal 14, no. 1. pp.??
Stiglitz, J.E. and P. Dasgupta 1971. "Differential Taxation, Public Goods, and Economic
44
Samuelsan, ?.4. 1964 the Pure Theory of Public Exenditme[?] )
Efficiency," The Review of Economic Studies 37, pp. 151-74.
reprented in Collected Pager of PAS.
64
I
Stiglitz and Weiss (1981)
I
World Bank 1992, World Development Report 1992: Development and the Environment.
2
2
Washington, D.C.
3
Tietenberg (1992)
,
4
4
5
Tites, J.G. 1987. "Causes and effects of sea level rise.' In: Preparing for Climate
6
Change. Proceedings of the First North American Conference on Preparing for Climate
7
Change. Government Institutes, Inc., Rockville, MD.
$
9
Tobia, J., and W. Nordhaus 1972, "Is Growth Obsolete?" ...
10
II
Hartwick 1990,
12
13
United Nations Environment and Development Branch, U.N. Center for Transnational
14
Corporations, "Options to Increase Transfer of Environmentally Sound Technologies to
13
Developing countries on Favorable Terms," New York, 1991, 89pp.
16
17
United Nations 1989, 'African Alternative Framework to Structural Adjustment
18
Programmes for Socio-Economic Recovery and Transformation'. UNECA/OAU.
19
E/ECA/cm. 15/6/Rev 3.
20
21
Unterwurzacher and Wird (1991)
22
23
U.S. Action Plan (1993)
24
25
U.S. Congress, Office of Technology Assessment Technologies to Maintain Biological
26
Diversity. Government Printing Office: Washington, D.C., 1987.
27
28
Warrick, R., Oerlemans, J. et al. 1990. 'Sea level rise.' In: Climate Change The IPCC
29
Scientific Assessment. J.T. Houghton, O.J. Jenkins and JJ. Ephrams (eds). Cambridge
30
University Press, Cambridge.
31
32
Weitzman, M.L. 1993. 'What to preserve? An application of diversity theory to crane
33
conservation.' Quarterly Journal of Economics 108, pp. 157-184.
34
35
Weitzman, M.L. 1991, 'On Diversity'. Discussion Paper 1533, Harvard University Institute
36
of Economic Research, Cambridge, MA. 54 pp.
37
38
Weitzman M.L., Newey, W., and Rabin, M. 1981. "Sequential R&D Strategy for
39
Synfuels," The Bell Journal of Economics 12(2), 574-590.
40
41
Weitzman (1974)
42
43
Wilson, E.O. 1988. 'The Current State of Biological Diversity,' in E.O. Wilson (ed.)
44
Biodiversity, National Academy Press, Washington, PP. 3-18.
65
66
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
1
SUMMARY FOR POLICYMAKERS
I
The Convention thus incorporates, and the literature assessed in this report (Chapters
SECOND ASSESSMENT REPORT, WORKING GROUP III
2
2
1. 2. 8 and 9) supports a strategy of "act - then learn then act" where the actions to
3
3
address climate change are periodically adjusted in light of new knowledge and
4
4
The Context
developments. The ultimate objective of stabilizing atmospheric concentrations of
5
5
greenhouse gases, although not yet operationally defined, provides guidance for betw 450-750
This assessment of the socio-economic literature related to climate change has been
6
6
policymakers when considering initial actions to address climate change.
modestreductions
7
undertaken in the context of sustainable development as reflected in the Framework
7
8
Convention on Climate Change. This Convention is now a key element in the
8
The Convention establishes a collective decision making process where actions,
in near.term
decision making framework on climate change (Chapter 2).
9
9
reflecting their common but differentiated responsibilities, are agreed upon by the
10
10
Parties. The report identifies a number of general principles and findings that are
11
The ultimate objective of the Convention is "stabilization of greenhouse gas
11
relevant regardless of the decisionmaking process, such as those concerning the
concentrations in the atmosphere at a level that would prevent dangerous
12
12
design of cost-effective policies. should be recognized that a collective decision
anthropogenic interference with the climate system. Such a level should be achieved
13
13
4
making process may not reach the same conclusions as analyses that adopt a global
within a time-frame sufficient to allow ecosystems to adapt naturally to climate
14
14
optimization perspective.
IS
change, to ensure that food production is not threatened and to enable economic
15
16
development to proceed in a sustainable manner."
16
A Portfolio of Possible Actions to Limit Climate Change
17
17
The Convention specifies that to achieve this objective the "Parties should take
18
18
The precautionary principle (Article 3.3) and the estimated damages due to current
precautionary measures to anticipate, prevent or minimize the causes of climate
19
19
emissions of greenhouse gases (Chapter 6) argue for action beyond "no regrets"
20
change and mitigate its adverse effects. Where there are threats of serious or
20
measures.4 The decisions to be taken in the near future will necessarily have to be
bigarsmall
21
irreversible damage, lack of full scientific certainty should not be used as a reason
21
taken under great uncertainty. However, these decisions are not very sensitive to the
22
for postponing such measures, taking into account that policies and measures should
22
level at which atmospheric concentrations are ultimately stabilized and can be
as
23
be cost-effective
23
adjusted based on experience, new knowledge and other developments.
shartan
24
24
25
Each of the Annex I Parties is committed to "adopt national policies and take
25
Actions analysed in the literature which countries can take, individually or jointly, to ofr solin in
26
corresponding measures on the mitigation of climate change, by limiting its
26
limit climate change and its impacts include:
27
anthropogenic emissions of greenhouse gases and protecting and enhancing its
27
nobustbur
28
greenhouse gas sinks and reservoirs."
28
implementing low cost measures to reduce emissions of greenhouse
29
29 30 are these
gases. There is broad agreement that in most countries there are some
Whatnit?
measures, such as energy efficiency, that can reduce emissions of
31
being
I United Nations Framework Convention on Climate Change, Article 2. The
33
recommended
greenhouse gases at negative to slightly positive costs (Chapter 9).
32
Achieving such reductions may require policy intervention;
30
31
Intergovernmental Negotiation Committee (INC) requested that the IPCC assess the
34
phasing out existing distortionary policies, such as some subsidies to
32
scientific evidence that might help policy makers interpret Article 2 of the
35
fossil fuels, that reduce welfare and, directly or indirectly, increase
33
Convention. This topic will be addressed in the IPCC's Second Assessment Report
34
to be published in 1995.
35
2 United Nations Framework Convention on Climate Change, Article 3.3.
36
4 No regrets measures are those whose benefits, such as reduced energy costs and
37
reduced emissions of conventional pollutants equal or exceed their cost, excluding
36
3 United Nations Framework Convention on Climate Change, Article 4.2(a).
38
the benefits of climate change mitigation.
I
2
Cour some $ some &
n
Cut: "Don't have to do anything for duades
careful
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
greenhouse gas emissions. Global emissions reductions of 5% to 20%
where does this
I
An illustrative simulation of global options with marginal costs of up to $50 per $50
2
have been estimated for removal of energy subsidies worldwide (Chapter
2
tonne of carbon suggests about 50% of the total reduction/sequestration would be come
3
11);
3
forestry measures, 25% would be through energy conservation and efficiency
Latercites tran $5-
4
4
improvements, and the remainder would come from a mix of renewable energy use,
5
implementing fuel switching measures, such as coal-fired generation to
5
fuel switching, CO2 removal and disposal, and nuclear energy. The mix of options
20?
6
natural gas, to reduce emissions of greenhouse gases (Chapters 9 and 11);
6
and the magnitude of the reduction/sequestration relative to current emissions varies
7
7
significantly by region (Chapter 7).
8
implementing measures to enhance sinks or reservoirs of greenhouse
8
9
gases;'
9
Opportunities for Action Vary by Country and Over Time
10
10
II
instituting forms of international cooperation to limit greenhouse gas
11
Each country's approach to climate change can be represented as a portfolio of
12
emissions such as joint implementation and technology transfer;
12
actions that reflects its objectives, economic conditions and constraints. The actions
13
13
taken by a particular country may reflect considerations such as their cost, secondary
14
undertaking research aimed at better understanding the causes and
14
benefits, the impacts on different socio-economic groups, international equity,
15
impacts of, and adaptation to, climate change. Economic studies suggest
15
intergenerational equity, the expected damage due to climate change or sea level rise
16
that such research can yield high returns by reducing the uncertainty as
16
within the country, and economic strategies based on development of new
17
to the most cost-effective actions to address climate change (Chapter 2);
17
technologies. The size of the investment will reflect the country's obligations under
18
18
international agreements, adjusted for resources transferred to or received from other
19
conducting technological research aimed at enhancing energy efficiency,
19
countries, as well as domestic goals and priorities. A country's climate change
20
minimizing emissions of greenhouse gases from continued use of fossil
20
actions will also reflect other priorities aimed at sustainable development, including
21
fuels, and developing commercial non-fossil energy sources. The timing
21
eradication of poverty, increased levels of education, protection of environmental
22
and availability of cost-effective of non-fossil energy technologies is one
22
resources, and improved human health.
23
of the major determinants of addressing climate change in the long run
23
24
(Chapters 9 and 10);
24
Governments can and should adjust their portfolio of climate change actions in
25
25
response to new information and developments. Indeed, both the investment in
26
planning, and implementing as necessary, measures to adapt to the
26
climate change limitation and the mix of policies will probably change over time.
27
consequences of climate change; and
27
That does not imply complete flexibility. Many of the measures involve capital and
28
28
technology commitments with long life times. Reversing such commitments could be
29
developing institutional mechanisms, such as insurance, to share the risks
29
relatively costly, however that risk would be incorporated into the analysis of the
30
of damages due to climate change (Chapters I and 2).
30
options before the commitments are made. This is particularly important for
31
31
developing countries and countries with economies in transition where significant
32
infrastructure investments will be made over the next few decades.
33
32
34
Stabilization of Greenhouse Gas Concentrations
$ Energy prices that do not reflect the externalities associated with energy use can be
33
considered a distortionary policy. Under some circumstances, kerosene subsidies to
35
34
reduce fuelwood use, can increase welfare or decrease emissions (Chapters 2 and
36
Climate change impacts are related to the stock of greenhouse gases (concentration)
35
11).
37
in the atmosphere, rather than the annual emissions, hence the ultimate objective of
36
6 Some actions to increase sinks may have indirect effects, such as a change in
37
albedo. In such cases the net effect of the measure should be considered.
3
4
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
1
the Convention is to stabilize atmospheric concentrations of greenhouse gases.'
I
Stabilization of the atmospheric concentrations of other greenhouse gases at today's
2
Currently, the atmospheric concentration of CO2 is about 360 ppmv; the pre-
2
levels requires lesser reductions, but nevertheless reductions, of the emissions of
3
industrial concentration was about 280 ppmv." The danger of significant harmful
3
those gases." Thus, stabilization of atmospheric concentrations of greenhouse gases
4
effects from climate change increase as the concentration of greenhouse gases in the
4
will require very low population and economic growth rates (as assumed for IS92c),
5
atmosphere rises. The discussion below focuses on the effects associated with
5
a major "natural" change in energy technologies, or intervention to reduce emissions
6
stabilization of atmospheric concentrations of CO2 within the range of 350 to 750
6
and enhance sinks. The effect of the initial commitments of Annex I Parties,
7
ppmv "could be attained only with global anthropogenic emissions that eventually
7
assuming that they are achieved and continued to 2050, is shown in Figure 2. The
8
drop to substantially below 1990 levels."
8
current commitments are not sufficient to stabilize atmospheric concentrations of
9
9
CO2 except under scenario IS92c which has emissions below current commitments
10
The IPCC's six emissions scenarios assume virtually no new policies specifically
(Chapter 12).
11
designed to reduce greenhouse gas emissions (Chapter 12). Three scenarios that
Applicability of Cost-Benefit Analysis
Annex 1"
12
span the full range (IS92a, IS92c and IS92e) are compared with the illustrative
13
profiles of anthropogenic emissions consistent with stabilization of atmospheric
14
concentrations at levels between 350 and 750 ppmv in Figure 1. These emissions
Edmind But wanto 12
A stabilization objective could be based upon an absolute standard, an affordable
Key is LDCs
15
profiles indicate that at some stage major reductions of CO2 emissions (of the order
16
of 50% of current emissions) will be required to eventually stabilize concentrations.
Averynt
15
safe minimum standard, or a cost-benefit analysis, the approach favoured by most
Weyans
16
economists. These approaches are illustrated in Figure 3. The absolute standard
17
The pattern and timing of reductions of emissions can have a major effect on the
17
approach would define a maximum atmospheric concentration of greenhouse gases
18
overall costs of mitigation. "Emissions for all stabilization levels studied are lower
18
that is considered to constitute "dangerous anthropogenic interference with the
19
than those for IS92a and e, even in the first few decades of the 21" century.
19
climate system" on the basis of predicted changes to the climate system or of the
20
Emissions for the IS92c Scenario lie between the emissions [profiles] which
20
predicted impacts of climate change regardless of the cost of achieving this
21
achieve stabilization at 450 and 550 ppmv.""
21
concentration. Environmental standards designed to safeguard human health are
22
22
often established using this approach.
But fig. 1 decs not based m min C.
23
24
The safe minimum standard approach would specify a maximum atmospheric
25
concentration of greenhouse gases based on an assessment of the risks associated
23
, It is not clear from the Convention whether the ultimate objective is to stabilize the
26
with different atmospheric concentrations and the costs of achieving those
24
atmospheric concentration of each gas at a particular level or to stabilize greenhouse
27
concentrations. The standard would seek to reduce the risks of climate change to an
25
gas concentrations, expressed in terms of CO₂ equivalents, at a particular level.
28
acceptable level at an acceptable cost. Cost-benefit analysis attempts to value the
29
impacts of climate change and seeks the atmospheric concentration where the
26
. ppmv is parts per million by volume, a measure of concentration.
30
marginal value of the net impact is equal to the marginal cost of abatement.
31
27
, Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
28
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p. 4.
32
Working Group III was specifically requested to assess whether cost-benefit analysis
29
Analyses of emissions profiles consistent with stabilization of atmospheric
33
is applicable to climate change. Cost-benefit analysis was broadly interpreted to
30
concentrations of CO2 at levels lower than 350 ppmv or higher than 750 ppmv are
34
include traditional project level cost-benefit analysis, cost-effectiveness analysis,
31
presently not available in the literature.
35
multi-criteria analysis and decision analysis. The traditional cost-benefit approach is
32
10 The 1994 Special Report of IPCC, Summary for Policy Makers, IPCC, 1994, p. ?.
33
" Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
36
12 Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
34
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p.
37
Assessment Working Group of IPCC. Summary for Policy Makers, IPCC, 1994, p.
35
15.
38
18 and 20.
5
6
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
1
depicted in Figure 3(c). Cost effectiveness analysis seeks to find the lowest cost
I
The literature contains a few aggregate (across all sectors and regions) estimates of
option to achieve a specified objective; the marginal cost of abatement curves in
2
2
the damages associated with a climate characterized by a globally-averaged 2.5°C
3
Figures 3(b) and 3(c) represent cost-effective options. Multi-criteria analysis is
3
warming. The estimates suggest that if such warming occurred now it would impose
designed to deal with problems where benefits and/or costs can not be valued; it
4
4
net damages equivalent to 1.5% to 2.5% of world GDP on the present world
could be used for the approaches described in Figures 3(a) or 3(b), where non-
5
5
economy (Chapter 6). The regional variation in climate change damage would be
6
economic effects are brought into the analysis, but without a formal way of valuing
6
substantial. Damage in developing countries is estimated to be at least 2% with a
7
those effects relative to the economic impacts. Decision analysis focuses expressly
7
maximum central estimate of 9% of GDP for some regions. Small islands and low
on making decisions under uncertainty and conceptually could be used in any of the
8
8
lying coastal areas are particularly vulnerable. Climate changes of this magnitude are
9
approaches described in Figure 3.
9
not expected to be realized for some decades and damages in the interim would be
10
10
smaller.
11
Cost-benefit analysis (broadly defined) could contribute to resolution of three
11
12
questions facing decisionmakers. These are:
12
Most estimates of the additional damage done by CO₂ emitted now range from $5 to
13
13
$25 (1990 $ U.S.) per tonne of carbon, although a few estimates range up to $125
14
By how much should emissions of greenhouse gases be reduced or sinks
14
per tonne of carbon." The models used to develop these estimates are simplistic and
15
of greenhouse gases be enhanced?
15
are poor representations of dynamic processes, but they reflect the current state of
16
16
the art. The wide range of estimates reflects variations in models, scenarios, discount
17
When should emissions be reduced or sinks be enhanced?
17
rates and other assumptions.
18
18
19
What methods should be employed to reduce emissions and enhance
19
Climate change benefits of emissions reduction measures can be very complex to
20
sinks?
20
estimate. Some energy efficiency and fuel switching measures simultaneously reduce
21
21
emissions of greenhouse gases and sulphate aerosols. The negative radiative forcing
22
Because of the large uncertainties surrounding estimates of damages due to climate
22
from sulphate aerosols is significant. "Tropospheric aerosols have a lifetime of only
23
change (Chapter 6) and the costs of mitigation (Chapter 9) make it difficult in
23
a few days So the atmospheric concentration of aerosols responds rapidly to
24
practice to rely solely on cost-benefit analysis for the selection of an appropriate
24
changes in emissions. Control of sulphate emissions would immediately reduce
25
stabilization objective at this time. Still, the information provided by cost-benefit
25
the amount of aerosols in the atmosphere." A recent model analysis finds that
26
analysis is critical in making judgements not only about that decision, but also about
26
replacement of fossil fuels with carbon-free energy sources leads to lower
27
the timing of emission reductions and the design of cost-effective strategies.
27
temperature increases in the long run, but higher temperatures in the interim due to
28
Integrated assessment models best suited to such analyses are quite new (Chapter
28
the reduced sulphur emissions (Chapter 10).
29
10).
29
30
30
Costs of Response Options
31
However, cost-effectiveness analysis can and should be used to evaluate adaptation
31
32
and mitigation options.
32
What matters from a policy perspective is the net cost (total cost adjusted for
33
33
secondary benefits and costs) of a mitigation or adaptation option apart from its
34
The Benefits of Limiting Greenhouse Gas Emissions and Enhancing Sinks
35
The benefits of limiting greenhouse gas emissions and enhancing sinks are the
34
36
13 A tonne is a metric ton or 1,000 kg. $5 per tonne of carbon corresponds to
climate change damages avoided. Since some increase in atmospheric concentrations
35
approximately 0.35 per litre of motor gasoline.
37
38
from the current level is virtually inevitable, some impacts and damages due to
36
14 Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
39
climate change seem unavoidable.
37
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p.
40
38
24.
7
8
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
climate change benefits (Chapter 8). A wide range of estimates of the costs of
I
same set of input assumptions, such as the cost and availability of different energy
2
response options is found in the literature. The wide range is due in large measure to
2
sources.
3
differences in assumptions. Alternative views of the costs of emissions reduction are
3
4
shown schematically in Figure 4. Curve A assumes that there are no reducible
4
Top-Down and Bottom-Up Models
5
market or government induced imperfections; thus there are no "no regrets" options.
5
6
Curves B, C and D all assume market failures or government induced imperfections
7
8
9
benefits yield net cost savings. Curves C and D incorporate secondary economic and
perfect 1mg Pating Bor Pni BED
In the jargon of modelers "top-down" models are aggregate models, often
exist in one or more sectors that can be corrected by policies that reduce costs and
also lower greenhouse gas emissions. Curve B includes only measures whose direct
give 9 8 7
macroeconomic models, that analyse how changes in one sector of the economy
affect other sectors and regions. Early top-down models tended to have little detail
on energy consumption, especially at the technology-specific level, but explicit
10
environmental benefits into the calculation of the net cost of the option.
0
treatment of behaviour and economic interactions. In contrast, "bottom-up" models
11
II
tended to describe energy consumption in detail, however consumer behaviour and
12
The potential for "no regrets" policies is represented by the line OD2. The area
12
interactions with other sectors of the economy tended to be down played.
13
OD,D₂ is the potential net benefit to society of implementing all "no regrets"
13
14
measures. Some analysts have argued that measures with a positive cost should be
14
This simple characterization of top-down and bottom-up models is increasingly
15
implemented to the point where the total net cost to society is zero, i.e. to a point
15
misleading as more recent versions of each approach have tended to provide greater
16
D₃D₂ such that the area under curve D between D₂ and D, is equal to that area
16
detail in the areas that were less developed in the past. This convergence means that
17
OD,D2.
17
differences in model results are increasingly driven by differences in input
18
18
assumptions rather than model structure. Most results available in the literature do
19
Despite significant differences in views, there is agreement that some energy
19
not reflect the effects of this convergence of model structure since the evolution in
20
efficiency improvements (perhaps to 10% to 30% of current consumption) can be
20
model structure is a relatively recent occurrence.
21
realized at negative to slightly positive costs in many countries, depending on the
21
22
baseline assumptions and the time frame (Chapter 9). Measures to reduce greenhouse
22
Nevertheless, differences in model structure remain important as different models are
23
gas emissions often yield additional economic benefits (such as reduced traffic
23
best suited to answer different types of questions.
24
congestion) and/or environmental benefits (such as reduced emissions of acid rain
24
25
and urban smog precursors). The magnitude of these secondary benefits depends on
25
Equity and Social Considerations
26
local circumstances. Studies for Norway, the U.K. and some other countries indicate
26
27
that secondary benefits could offset 30% to 100% of abatement costs (Chapter 6).
27
Equity considerations are an important aspect of climate change policy and feature
28
28
prominently in the Convention. Inaction is likely to impose costs on future
29
The long run costs of addressing climate change will ultimately be determined by
29
generations and on regions where damages occur, often regions with low greenhouse
30
the rate of capital replacement, the discount rate and the effect of research and
30
gas emissions. Action to address climate change is likely to impose some costs on
31
development. Implementation of "no regrets" measures will increase the time
31
the present generation. Since actions to address climate change are deliberate,
32
available to learn about climate change and to bring new technologies to the market.
32
explicit decisions can be made as to how to share these burdens.
33
Infrastructure decisions are crucial because they can enhance or restrict future
33
34
options and different choices can lead to very different cost outcomes.
34
The Convention recognizes the differentiated responsibilities of the Parties, but it
35
35
does not include an explicit formula for sharing the costs of addressing climate
36
Estimates of the costs of stabilizing CO2 emissions vary widely as a result of
36
change. It also lists factors to be considered in determining how the burden should
37
differences in the baseline scenario, the estimated long run costs of mitigation
37
be shared and provides for development of mechanisms for sharing the burdens.
38
measures, and the date for achieving the stabilization objective. Differences among
38
Issues of equity and efficiency can for the most part be separated. Cost-effective
39
models are less significant than differences in the underlying assumptions. Indeed
39
strategies can be developed embracing a variety of ways for sharing the burden.
40
widely differing methods can produce quite similar results when calibrated to the
40
9
10
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
Equity issues involved in addressing climate change may be divided into four
I
contributions to the problem, while emissions from others may well start to dominate
2
categories:
2
it.
3
3
4
distributing the costs of adaptation;
4
Countries (such as low lying coastal regions and islands or drylands) and special
5
5
groups within society (such as the poor, and sometimes women or children, or
6
distributing future emissions rights;
6
specific occupations or regions) that are especially vulnerable to climate change -
7
7
those on whom the costs of abatement and coping would be especially burdensome
8
distributing the costs of abatement; and
8
- merit special attention.
9
9
10
ensuring institutional and procedural fairness.
10
All Parties have equal status under the Convention. Procedural equity also requires
11
II
that all Parties be able to participate effectively in international negotiations related
12
The Convention offers some guidance on all of these issues, but doesn't resolve any
12
to climate change. Assistance to enable developing country Parties to participate
13
of them.
13
effectively in negotiations would increase the prospects for achieving effective,
14
14
lasting and equitable agreements on how best to address the threat of climate
15
A variety of ethical principles, including the importance of meeting people's basic
15
change.
16
needs, are relevant to sharing the burden of addressing climate change. No single
16
17
rule will command universal agreement. Furthermore, any arrangement may need to
17
Intergenerational Equity
18
change over time to continue to be perceived as equitable and to take new
18
19
developments into account. Analysts can not prescribe how the burden of addressing
19
Climate policy raises particular questions of equity among generations, because
20
climate change should be shared, they can only clarify the implications of different
20
future generations are not able to influence directly the policies being chosen today
21
principles (Chapter 3).
21
that will affect their well-being and because it might conceivably not be possible to
22
22
compensate future generations for reductions in well-being that might be caused by
23
Countries differ in terms of current greenhouse gas emissions (see Figures 5 and 6),
23
current policies.
24
vulnerability to climate change (Chapter 6), wealth, resource endowments, and
24
25
institutional capacity to respond effectively. The Convention provides that these
25
Sustainable development and the concept of an environmental trust have been
26
circumstances be reflected in differentiated responsibilities. Developed countries
26
proposed to address intergenerational equity. Sustainable development would meet
27
have undertaken commitments through the Convention for technology transfer and
27
"the needs of the present without compromising the ability of future generations to
28
contributions to the financial mechanism.
28
meet their own needs." An environmental trust, as proposed by Brown-Weiss, would
29
29
require each generation to ensure that the next inherits "a planet and cultural
30
A rigid "north-south" delineation of equity issues is inappropriate and may be highly
30
resource base at least as good as that of previous generations."
31
damaging in the long run, although it permeates much of the debate at present. The
31
32
implications of climate change for developing countries are different than those for
32
Discount Rate
33
developed countries because the former are generally poorer, have contributed much
33
good.
34
less to past emissions and still emit much less per capita, have shorter policy time-
34
The discount rate is the analytical tool economists use to compare economic effects
35
horizons, often weaker institutions, other urgent priorities, and are generally more
35
that occur at different points in time. Identifying the proper discount rate is probably
36
vulnerable to climate change. However, there are substantial variations within the
36
the single most important analytical step in economic analysis of global warming.
37
"north" and within the "south" in terms of absolute and per capita emissions, the
37
The choice of discount rate is important in climate change analyses because the time
38
likely impacts of climate change, institutional strengths and preferences, and
38
horizon is extremely long and the abatement costs tend to come earlier than the
39
endowment of natural resources that may be affected by mitigation. Similarly, over
39
benefits of avoided damages. It is also the most profound ethical question, since it
40
the next century some developing countries will continue to make marginal
11
12
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
inherently confronts present pain of abatement cost against damages suffered by
I
climate change, however, society is not likely to be able to set aside investments
future generations if no action is taken.
2
2
over the next three centuries and earmark the proceeds the eventual compensation of
3
3
those adversely affected. Determining the discount rate on the basis of descriptive or
The literature on the appropriate discount rate for climate change analysis can be
4
4
market considerations yields a global rate in the range of 3% to 6% real, with rates
divided into two approaches. The prescriptive approach asks whether a particular
5
5
declining within this range as income rises.
action would make society better off, given a plausible social welfare function. The
6
6
7
discounting rules specify how consumption at different dates should be valued to
7
Increasing the discount rate because of risk is generally inappropriate.
8
provide the correct answer to that question. The descriptive approach focuses on
8
9
intertemporal efficiency and opportunity costs.
9
Policy Instruments
10
10
11
The prescriptive approach discounts consumption accruing to different generations
II
A clear distinction needs to be drawn between actions to reduce the impacts of
12
using the "social rate of time preference" (SRTP). The SRTP is composed of two
12
climate change and actions to reduce emissions. The former also involve
13
components, "pure time preference" and the incremental welfare derived from future
13
consideration of adaptation policies.
14
14
-------------------------
consumption:
15
15
Policy instruments designed to mitigate climate change need to be assessed at the
16
Pure time preference is a reflection of impatience. Although there is
16
international and domestic levels. Such instruments can help participating countries
17
debate in the literature, values tend to range from zero to 1%
17
comply with multilateral commitments on climate change abatement, including
18
18
international transfers of resources and technology. The international and domestic
19
The incremental welfare of future consumption accounts for rising per
19
policy instruments need not be the same, but international agreements may constrain
20
capita income over time and the resulting decline in the incremental
20
the choice of domestic policy instruments.
21
welfare (marginal utility) derived from the higher future consumption.
21
22
22
The policy instruments available internationally and for adoption by groups of
23
Depending upon the values assumed for the different parameters, values for the
23
countries include: non-tradeable quotas, tradeable quotas, joint implementation,
24
SRTP on a global basis tend to fall between 0.5% and 3.0% per year. Developing
24
harmonized domestic carbon taxes, international carbon taxes, various international
25
countries may have much higher (10% to 20%) rates of SRTP due to high rates of
25
standards and technology transfer. Domestic policy instruments include carbon taxes,
26
per capita income growth or high elasticities of marginal utility for small increases
26
tradeable permits, deposit-refund systems, technology standards, performance
27
in per capita income from subsistence levels.
27
standards, product bans, voluntary agreements, energy taxes, energy efficiency
28
28
subsidies, non-fossil fuel subsidies, and removing existing market distortions.
29
It is important remember that in making choices concerning alternative policies, that
29
30
while differences in consumption at different dates will be valued using the SRTP,
30
Among the criteria that governments can consider when selecting from among the
31
those differences in consumption will be affected by the opportunity cost of capital,
31
available policy instruments are: distributional (including intergenerational) equity,
32
which typically is in the range of 7% to 15% real. Thus, where mitigation projects
32
flexibility in the face of technological change and new information, efficiency or
33
displace other investments yielding higher returns, the mitigation projects typically
33
cost-effectiveness, compatibility with the institutional structure and existing policies,
34
will not be chosen unless they yield returns at least equal to those on the displaced
34
and understandability to the general public. A mix of instruments may be needed to
35
capital.
35
achieve the best results. Governments may apply different criteria with different
36
36
weights to the selection of international and domestic policy instruments.
37
The descriptive approach to the discount rate looks at returns to investments in the
37
38
real world. Market rates of return usually exceed the values estimated for the SRTP.
38
Cost-effectiveness should always be a criteria for selecting policy instruments, but it
39
Conceptually, funds could then be invested in projects that provide a higher return
39
becomes more important at both the international and domestic levels as the
40
with the proceeds being used to increase future consumption. In the context of
40
abatement effort becomes more stringent. The economic literature suggests that the
13
14
a
7
6
5
4
3
2
-
15
depend on their acceptability to the public and to governments.
design and implementation of those policies. The effectiveness of policies will also
change policy will be determined by the choice of policy instrument as well as the
consistent with equitable international burden sharing. The consequences of climate
Although there are implementation problems associated with each, both can be made
more economically efficient instruments include carbon taxes and tradeable quotas.
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
opt path doesn't
Figure 1
divence
Illustrative Anthropogenic Emissions Profiles of CO, Leading to
Stabilization at Concentrations of 350 to 750 ppmv Using a Mid-range
Carbon Cycle Model and CO2 Emissions for the IS92a, c and e Scenarios
Anthropogenic emissions (GtC/yr)
20
IS92e
IS92a
15
S750
take winds Ryer E mb
10
S650
S550
IS92c
Anthropogenic CO,
emissions in 1990
5
S450
S350
0
2000
2100
2200
2300
Year
930005 - 1/95.20
Source: 1994 Special Report of the IPCC
Figure 3
Global Warming, Cost-Benefit
anthropogenic CO,
emissions in 1990
Analysis and Sustainable Development
Damage
and
Abstement
2050
Costs
Minimum
Risk Zone
Zone of Uncertainly
S750
S650
S550
S450
S350
Comparison of IPCC Emissions Scenarios Adjusted for Current
Commitments and Emissions Profiles Consistent with Stabilization of
Unscceptable Resk
0
2040
AT
Atmospheric Concentrations of 350 to 750 ppmv
Source: Based on Chapter 12/adjusted for current commitments by Annex I Parties extended to 2050
A Trax
IS92c
(a) Absolute Standard Approach
IS92eC
IS92aC
2030
Damage
Marginel Cost
and
of Abstement
Abstement
Unscceptable Flisk
Costs
Figure 2
IS92a
2020
Legend: IS92aC (IS92eC) is scenario IS92a (IS92e) 4
Unacceptable
Cost
0
A Tran
AT
Anthropogenic emissions (GtC/yr)
IS92e
2010
(b) Safe Minimum Standard Approach
2000
Marginal Cost
Marginal
Damage
of Abalement
Damage
and
Abstement
Costs
1990
13
12
11
10
6
8
7
9
5
4
3
2
1
0
0
s Topt
AT
wish 1 I
(c) Cost-Benelit Approach
Source: Chapter 6
938005 Age L/BS -
Figure 4
Alternative Views on Costs of
Emission Reduction
un inch me 1 revenues
Marginal
Cost
A
B
CA not harefits carlyon
D
(no Ace CAST Inackelton analysis
0
(Gorlder papa)
D2
D₃
Emission
Reduction
also chap 8
D,
A - No reducible market imperfections and/or distatcon. Nardhave (woh)
B - - "Negative cost potential" achieved
C - Curve B plus "economic double dividends"
Tan recycling BUT
D . Curve C plus "environmental double dividends"
71 pm recta +
938005 tige 1/95 ap
that SOu aerosols could 80 other way
Source: Chapter 8
sector spanders an piblic
7
Prb finance question?
7
Figure 5
Carbon Emissions per Capita and Population, 1993
8
tC/cap
6
4
2
0
Africa
Former
India
China
Japan
Other
Latin
US
USSR
Asia
America
EU12
Other
Australia
Canada
Europe
Population, Billion
Source: Derived by M. Grubb from BP Statistical Review of World Energy 1994 and World Population Prospects, UN
938005 - 1/95.ap
Figure 6
Carbon Emissions/GNP and Total GNP, 1992
Carbon emissions/GNP (kg/USSm)
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Africa
Former
India
Japan
Australia
Latin
US
Canada
USSR
China
America
EU12
Other
Other
Europe
Asia
World total GNP, Sbn (market exchange rate)
Source: Chapter 3
930000 - Visap
Chapter 5:
Applicability of Techniques of Cost-Benefit
Analysis to Climate Change
Overall: This chapter is very lengthy and repetitive with other
chapters, especially chapter 1. Basic terminology should be used
more carefully, especially couching benefits and costs in present
value terms. The chapter needs careful editing.
SECTION II: SUMMARY OF COMMENTS
This chapter is long and reproduces information found in
other chapters (particularly the characteristics of the
climate change problems found in chapter 1).
Benefit and cost estimates should be couched in present
value terms.
Treatment of the discount rate is brief and not entirely
consistent with that found in chapter 4.
The discussion of the developing countries' public
statements condemning joint implementation is not
appropriate for the assessment. Host countries have
something to lose if emissions reductions they undertake
affect the baseline against which further emissions
reductions are measured.
Table 5.2 detracts from the section's description of
multicriteria analysis. Since 3-4 of the criteria are
considered to be central elements of the approach, the table
seems superfluous and could be deleted.
Some of the figures require more detailed information to
improve the readers' comprehension of the material.
SECTION III: DETAILED COMMENTS
p. iii 1. 16-17 The global scale of the greenhouse gas control
analysis derives from the characteristics of the climate change
problem and is not " a fundamental result of cost-benefit
analysis".
Replace last sentence in the paragraph with the following:
"The global scale for the analysis of greenhouse gas
emissions control comes from the characteristics of the
problem; that is, the composition of the atmosphere is a
global public good, whereby the services the atmosphere
provides are commonly held globally and the primary
underlying forces that influence the earth's climate are
also global (see chapter 1)."
p. 6 line 16 Section 3. Unique Features of climate change
This section overlaps greatly with the discussion in chapter
1. The points should be summarized, referring the reader to
the earlier chapter.
p.
11 Figure 4.1
Terms in the figure are not defined in the
text: CHP, CCGTs, PWRs
p.
12
1.
4
What are "T&D losses"?
p. 14 2nd line below table 4.1
Market impacts should be emphasized where possible in the
summary of damage estimates. Insert ", particularly the
market impacts," after "the economic impact of climate
change".
p. 18 1. 18 Replace "true" with "the actual budget curve".
p. 18 1. 20 Replace "true" with "the actual budget curve".
p. 22 1. 41-42
The second full paragraph seems to imply that
the risk premiums for lotteries with expected
return X will necessarily be equal -- which
is not true. The wording should be changed.
p.
22
1.
47 The two alternatives do not seem to have the same
expected reduction of 100 tons.
p. 23 1. 12-13
The sentence should be replaced with wording
similar to the following: "Even if not
enough is known about the probability
distribution of a variable to express it as a
single expected estimate, we should note
qualitatively the information that is
available; the information should not be left
out of the analysis completely."
p. 23 1. 23-24
Increased knowledge about climate change will
not automatically narrow the range of
uncertainties in the future. In some cases,
additional knowledge results in an
appreciation of additional influencing
factors of which individually little may be
known.
p. 24 1. 30
Interaction between the assumptions is important
as well. Add text such as: "Moreover, there is
considerable uncertainty about the interactions of
these assumptions. To some extent, economists
have little idea about the sensitivity of general
equilibrium results to assumed parameters; there
is no good analogue to confidence intervals used
in empirical analyses."
p. 24 point 6, 1. 6
Information imperfections are a good
example of market imperfections. Insert
"(e.g. information imperfections") after
"market imperfection".
p. 25 1. 7-8 This point made in this sentence is supported by
articles in the literature (Kahnman and Tversky --
among others) i This work suggests that individuals
have a difficult time rationally dealing with low
probability events.
p. 25 1. 47-48
Since the total US budget on defense was
between $200-$300 per year and no decline of
more than $100 billion has taken place since
the end of the cold war, it is highly
unlikely that this statement is true.
"hundreds of billions of dollars" should be
changed to "many billions of dollars".
p.
26 1. 4-5 Expected benefits and costs should be discussed in
present value terms. Insert "present value of"
before "expected" in line 4 and before "expected
costs" in line 5.
p. 28 footnote 59 The footnote is incomplete.
p. 31 1. 44-47
The treatment of CVM needs additional
clarification. As part of the rulemaking
exercise in the US under the Oil Pollution
Act, a blue-ribbon panel of economists was
convened to judge the adequacy of CVM for use
in damage assessments. The findings from the
panel should be included in this box.
p. 33 1. 16-18
How is the 5 percent increase calculated? is
it an increase in the annual rate of growth
in the "price" of the environment?
p.
33
1.
22
Replace "correct" with "optimal"; replace "optimal
social rate" with "social discount rate".
p. 33 1. 23-25
Why is the saturation and decline in economic
growth rates more likely than remaining
constant or increasing?
p.
34
1.
27
The discussion of the resistance of developing
countries to joint implementation is too
simplistic. Host countries have something to lose
if emissions reductions affect the baseline
against which further emission reductions are
measured. Lines 26-35 should be deleted. The
discussion of reasons developing countries
publicly renounce joint implementation does not
further the assessment of CBA.
p. 40 1. 5
The estimate of typical planning horizons of 15-25
years was used in the chapter. Why has the
estimate of 20 years used in the summary?
p. 40 1. 14-15
The text should use costs and benefits in
present value terms.
Minor changes:
p. iii 1. 6 Insert "area of" before "climate change."
p. 5 footnote 9, third line Remove "," after basin and insert
"s" after require
p. 7 1. 10
Insert "for greenhouse gases" after "calculus of
CBA"
p. 8 1. 23
Insert "damages" before "are related to "
p. 8 1. 48
"Special" is misspelled
p. 16 1. 17
Depletion of natural resources is important but a
lesser concern to that of the appropriate measure
of welfare. "Equally" should be changed to
"Also".
p. 18 footnote 43
"Kroner" should follow each currency
estimate, (i.e. 2.6 billion "Kroner", with a
range of 1.0-3.8 billion "Kroner")
p. 20 1. 23 Where are the results "above"?
p. 23 1. 25
Delete the word "value" since it is used again
beginning of line 26.
p. 24 1. 5
Insert "of emissions reductions" after "estimating
the costs" .
p. 24 1. 12
Insert ")" after "is difficult."
p. 26 1. 10
Replace "to buy (and sell) with "make future
purchases (or sales)".
p. 32 1. 34
Insert "affecting climate change" at the end of
sentence.
Chapter 5
p. iii, line 3. The criticism of narrow cost-benefit analysis
attacks a strawman.
p. iii, line 13. Cost-benefit analysis can help answer the first
question, but it would be simplistic to rely upon it unless
losers can be fully compensated by winners. For example, if
people have a "right" to a clean environment, then polluters
should not necessarily be allowed to discharge, just because the
costs may be greater than the benefits.
p. 32, line 35. In chapter 4 most of the discussion concerns the
case where the discount rate reflects time preference, not
opportunity cost. Opportunity cost can be reflected through the
shadow price of capital.
p. 33, line 1. The opportunity costs of different investments
could vary, depending on the share of costs drawn from
consumption and investment, for example.
p. 33, line 22. As a whole, the discussion of the discount rate
here is very brief and not particularly consistent with chapter
4. The treatment here is more straightforward, though, about the
difficulty of knowing the precise discount rate with certainty.
p. 34, line 25. The idea that receiving countries have nothing
to lose is too simplistic, if emissions reductions they undertake
affect the baseline against which further emissions reductions
are measured. This issue is more properly treated on line 39.
p. 35, table 5.2. The table presents what appear to be far too
many overlapping criteria. Basically, a tough problem is made
even worse by this type of presentation. Fortunately this is
recognized in line 42; however, the reader is dragged through an
unnecessarily difficult presentation in the process.
9
Chapter 12: An Evaluation of the IPCC IS92 Emission Scenarios
Section II: SUMMARY OF COMMENTS
This chapter is poorly written, has a defensive tone, and is
in some cases internally inconsistent.
The presentation of the material in mostly bullet form is
not easy to follow.
The summary for policy makers for emissions scenarios should
be used as the chapter core. The "scenarios summary" has
several figures, one of which is included in the overall
"summary for policy makers". If the emissions chapter
carries no figures, what is the source material for the
figures currently carried in the overall "summary for policy
makers"? If a figure is carried in the emissions scenarios
chapter, it should reflect current literature suggesting
that near-term optimal paths are very similar regardless of
the desired concentration level.
The discussion of "purposes" should be dropped completely
because it is confusing. Too much effort is expended in
describing the different stages of policy analysis that
utilize emissions projections when the very scenarios under
discussion can only be used for the first out of four
stages.
SECTION III: DETAILED COMMENTS
This chapter (chapter 12) is quite literally the technical
summary for the chapter on emissions scenarios found in the 1994
Special Report. The tone is defensive -- justifying
unsuccessfully why true analysis of emissions projections
developed in 1992 was not accomplished. The text excessively
discusses to what ends emissions projections should be used, why
only energy and land-use activities are covered, why only a few
geographic regions were studied, and why only a subset of
greenhouse gas emissions were included in the analysis. It is
abundently clear that minimal effort was expended to actually
compare models that generated the emissions paths, to examine the
reasonableness of the assumptions, and to test the sensitivity of
the results to the models and the underlying assumptions.
The information in the chapter is not well presented. Much
of the text is organized into bullets. The writing is imprecise
and confusing, making it difficult to follow the arguments even
if one is technically trained in economics. A set of "purposes"
identified for the use of emissions scenarios adds little to the
understanding of the IPCC 1992 scenarios but underscores their
limited utility beyond that as input data for general circulation
models.
The easiest way to fix this chapter is to use the summary
for policy makers found in the 1994 Special Report as its core.
More detailed information can be added where appropriate. The
writing of the summary for policy makers for emissions scenarios
is much cleaner and technically correct than that of the
technical summary of the chapter. The summary for policy makers
has been through contentious government review and plenary
debate. Its tone is matter-of-fact and presents information
carefully.
No figure is currently found in the chapter. The "policy
makers summary" needs to reference material in the underlying
chapters. If figures are used in the emissions scenario, they
should be vetted extensively.
Detailed information from the underlying chapter in the
special report or the technical summary could be added to the
emissions summary for policy makers to enhance its contribution
to the readers' understanding of climate change.
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
SUMMARY FOR POLICYMAKERS
1
The Convention thus incorporates, and the literature assessed in this report (Chapters
2
SECOND ASSESSMENT REPORT, WORKING GROUP III
2
1, 2, 8 and 9) supports a strategy of "act - then learn - then act" where the actions to
3
3
address climate change are periodically adjusted in light of new knowledge and
4
The Context
4
developments. The ultimate objective of stabilizing atmospheric concentrations of
5
5
greenhouse gases, although not yet operationally defined, provides guidance for
6
This assessment of the socio-economic literature related to climate change has been
6
policymakers when considering initial actions to address climate change.
7
undertaken in the context of sustainable development as reflected in the Framework
7
8
Convention on Climate Change. This Convention is now a key element in the
8
The Convention establishes a collective decision making process where actions,
9
decision making framework on climate change (Chapter 2).
9
reflecting their common but differentiated responsibilities, are agreed upon by the
10
10
Parties. The report identifies a number of general principles and findings that are
11
The ultimate objective of the Convention is "stabilization of greenhouse gas
11
relevant regardless of the decisionmaking process, such as those concerning the
12
concentrations in the atmosphere at a level that would prevent dangerous
12
design of cost-effective policies. It should be recognized that a collective decision
13
anthropogenic interference with the climate system. Such a level should be achieved
13
making process may not reach the same conclusions as analyses that adopt a global
14
within a time-frame sufficient to allow ecosystems to adapt naturally to climate
14
optimization perspective.
15
change, to ensure that food production is not threatened and to enable economic
15
16
development to proceed in a sustainable manner."
16
A Portfolio of Possible Actions to Limit Climate Change
17
17
18
The Convention specifies that to achieve this objective the "Parties should take
18
The precautionary principle (Article 3.3) and the estimated damages due to current
19
precautionary measures to anticipate, prevent or minimize the causes of climate
19
emissions of greenhouse gases (Chapter 6) argue for action beyond "no regrets"
20
change and mitigate its adverse effects. Where there are threats of serious or
20
measures. The decisions to be taken in the near future will necessarily have to be
21
irreversible damage, lack of full scientific certainty should not be used as a reason
21
taken under great uncertainty. However, these decisions are not very sensitive to the
22
for postponing such measures, taking into account that policies and measures should
22
level at which atmospheric concentrations are ultimately stabilized and can be
23
be cost-effective #2
23
adjusted based on experience, new knowledge and other developments.
24
24
25
Each of the Annex I Parties is committed to "adopt national policies and take
25
Actions analysed in the literature which countries can take, individually or jointly, to
26
corresponding measures on the mitigation of climate change, by limiting its
26
limit climate change and its impacts include:
27
anthropogenic emissions of greenhouse gases and protecting and enhancing its
27
28
greenhouse gas sinks and reservoirs."
28
implementing low cost measures to reduce emissions of greenhouse
29
29
gases. There is broad agreement that in most countries there are some
30
measures, such as energy efficiency, that can reduce emissions of
31
greenhouse gases at negative to slightly positive costs (Chapter 9).
32
Achieving such reductions may require policy intervention;
30
I
United Nations Framework Convention on Climate Change. Article 2. The
33
31
Intergovernmental Negotiation Committee (INC) requested that the IPCC assess the
34
phasing out existing distortionary policies, such as some subsidies to
32
scientific evidence that might help policy makers interpret Article 2 of the
35
fossil fuels, that reduce welfare and, directly or indirectly, increase
33
Convention. This topic will be addressed in the IPCC's Second Assessment Report
34
to be published in 1995.
35
2 United Nations Framework Convention on Climate Change, Article 3.3.
36
4 No regrets measures are those whose benefits, such as reduced energy costs and
37
reduced emissions of conventional pollutants equal or exceed their cost, excluding
36
3 United Nations Framework Convention on Climate Change, Article 4.2(a).
38
the benefits of climate change mitigation.
1
2
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
greenhouse gas emissions. Global emissions reductions of 5% to 20%
I
An illustrative simulation of global options with marginal costs of up to $50 per
2
have been estimated for removal of energy subsidies worldwide (Chapter
2
tonne of carbon suggests about 50% of the total reduction/sequestration would be
3
11);
3
forestry measures, 25% would be through energy conservation and efficiency
4
4
improvements, and the remainder would come from a mix of renewable energy use,
5
implementing fuel switching measures, such as coal-fired generation to
5
fuel switching, CO2 removal and disposal, and nuclear energy. The mix of options
6
natural gas, to reduce emissions of greenhouse gases (Chapters 9 and 11);
6
and the magnitude of the reduction/sequestration relative to current emissions varies
7
7
significantly by region (Chapter 7).
8
implementing measures to enhance sinks or reservoirs of greenhouse
8
9
gases;6
9
Opportunities for Action Vary by Country and Over Time
10
10
11
instituting forms of international cooperation to limit greenhouse gas
11
Each country's approach to climate change can be represented as a portfolio of
12
emissions such as joint implementation and technology transfer;
12
actions that reflects its objectives, economic conditions and constraints. The actions
13
13
taken by a particular country may reflect considerations such as their cost, secondary
14
undertaking research aimed at better understanding the causes and
14
benefits, the impacts on different socio-economic groups, international equity,
15
impacts of, and adaptation to, climate change. Economic studies suggest
15
intergenerational equity, the expected damage due to climate change or sea level rise
16
that such research can yield high returns by reducing the uncertainty as
16
within the country, and economic strategies based on development of new
17
to the most cost-effective actions to address climate change (Chapter 2);
17
technologies. The size of the investment will reflect the country's obligations under
18
18
international agreements, adjusted for resources transferred to or received from other
19
conducting technological research aimed at enhancing energy efficiency,
19
countries, as well as domestic goals and priorities. A country's climate change
20
minimizing emissions of greenhouse gases from continued use of fossil
20
actions will also reflect other priorities aimed at sustainable development, including
21
fuels, and developing commercial non-fossil energy sources. The timing
21
eradication of poverty, increased levels of education, protection of environmental
22
and availability of cost-effective of non-fossil energy technologies is one
22
resources, and improved human health.
23
of the major determinants of addressing climate change in the long run
23
24
(Chapters 9 and 10);
24
Governments can and should adjust their portfolio of climate change actions in
25
25
response to new information and developments. Indeed, both the investment in
26
planning, and implementing as necessary, measures to adapt to the
26
climate change limitation and the mix of policies will probably change over time.
27
consequences of climate change; and
27
That does not imply complete flexibility. Many of the measures involve capital and
28
28
technology commitments with long life times. Reversing such commitments could be
29
developing institutional mechanisms, such as insurance, to share the risks
29
relatively costly. however that risk would be incorporated into the analysis of the
30
of damages due to climate change (Chapters I and 2).
30
options before the commitments are made. This is particularly important for
31
31
developing countries and countries with economies in transition where significant
32
infrastructure investments will be made over the next few decades.
33
34
Stabilization of Greenhouse Gas Concentrations
32
5 Energy prices that do not reflect the externalities associated with energy use can be
35
33
considered a distortionary policy. Under some circumstances, kerosene subsidies to
36
Climate change impacts are related to the stock of greenhouse gases (concentration)
34
reduce fuelwood use, can increase welfare or decrease emissions (Chapters 2 and
37
in the atmosphere, rather than the annual emissions, hence the ultimate objective of
35
11).
36
6 Some actions to increase sinks may have indirect effects, such as a change in
37
albedo. In such cases the net effect of the measure should be considered.
3
4
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
1
the Convention is to stabilize atmospheric concentrations of greenhouse gases.'
I
Stabilization of the atmospheric concentrations of other greenhouse gases at today's
2
Currently, the atmospheric concentration of CO₂ is about 360 ppmv; the pre-
2
levels requires lesser reductions, but nevertheless reductions, of the emissions of
3
industrial concentration was about 280 ppmv.' The danger of significant harmful
3
those gases.¹² Thus, stabilization of atmospheric concentrations of greenhouse gases
4
effects from climate change increase as the concentration of greenhouse gases in the
4
will require very low population and economic growth rates (as assumed for IS92c),
5
atmosphere rises. The discussion below focuses on the effects associated with
5
a major "natural" change in energy technologies, or intervention to reduce emissions
6
stabilization of atmospheric concentrations of CO₂ within the range of 350 to 750
6
and enhance sinks. The effect of the initial commitments of Annex I Parties,
7
ppmv "could be attained only with global anthropogenic emissions that eventually
7
assuming that they are achieved and continued to 2050, is shown in Figure 2. The
8
drop to substantially below 1990 levels."
8
current commitments are not sufficient to stabilize atmospheric concentrations of
9
9
CO2, except under scenario IS92c which has emissions below current commitments
10
The IPCC's six emissions scenarios assume virtually no new policies specifically
10
(Chapter 12).
11
designed to reduce greenhouse gas emissions (Chapter 12). 10 Three scenarios that
11
12
span the full range (IS92a, IS92c and IS92e) are compared with the illustrative
12
Applicability of Cost-Benefit Analysis
13
profiles of anthropogenic emissions consistent with stabilization of atmospheric
13
14
concentrations at levels between 350 and 750 ppmv in Figure 1. These emissions
14
A stabilization objective could be based upon an absolute standard, an affordable
15
profiles indicate that at some stage major reductions of CO2 emissions (of the order
15
safe minimum standard, or a cost-benefit analysis, the approach favoured by most
16
of 50% of current emissions) will be required to eventually stabilize concentrations.
16
economists. These approaches are illustrated in Figure 3. The absolute standard
17
The pattern and timing of reductions of emissions can have a major effect on the
17
approach would define a maximum atmospheric concentration of greenhouse gases
18
overall costs of mitigation. "Emissions for all stabilization levels studied are lower
18
that is considered to constitute "dangerous anthropogenic interference with the
19
than those for IS92a and e, even in the first few decades of the 21" century.
19
climate system" on the basis of predicted changes to-the climate system or of the
20
Emissions for the IS92c Scenario lie between the emissions [profiles] which
20
predicted impacts of climate change regardless of the cost of achieving this
21
achieve stabilization at 450 and 550 ppmv.""
21
concentration. Environmental standards designed to safeguard human health are
22
22
often established using this approach.
23
24
The safe minimum standard approach would specify a maximum atmospheric
25
concentration of greenhouse gases based on an assessment of the risks associated
23
7 It is not clear from the Convention whether the ultimate objective is to stabilize the
26
with different atmospheric concentrations and the costs of achieving those
24
atmospheric concentration of each gas at a particular level or to stabilize greenhouse
27
concentrations. The standard would seek to reduce the risks of climate change to an
25
gas concentrations, expressed in terms of CO₂ equivalents, at a particular level.
28
acceptable level at an acceptable cost. Cost-benefit analysis attempts to value the
29
impacts of climate change and seeks the atmospheric concentration where the
26
a ppmv is parts per million by volume, a measure of concentration.
30
marginal value of the net impact is equal to the marginal cost of abatement.
27
9
31
Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
28
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p. 4.
32
Working Group III was specifically requested to assess whether cost-benefit analysis
29
Analyses of emissions profiles consistent with stabilization of atmospheric
33
is applicable to climate change. Cost-benefit analysis was broadly interpreted to
30
concentrations of CO2 at levels lower than 350 ppmv or higher than 750 ppmv are
34
include traditional project level cost-benefit analysis, cost-effectiveness analysis,
31
presently not available in the literature.
35
multi-criteria analysis and decision analysis. The traditional cost-benefit approach is
32
10 The 1994 Special Report of IPCC, Summary for Policy Makers, IPCC, 1994, p. ?.
33
11 Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
36
12 Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
34
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p.
37
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p.
35
15.
38
18 and 20.
5
6
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
1
depicted in Figure 3(c). Cost effectiveness analysis seeks to find the lowest cost
I
The literature contains a few aggregate (across all sectors and regions) estimates of
option to achieve a specified objective; the marginal cost of abatement curves in
2
2
the damages associated with a climate characterized by a globally-averaged 2.5°C
3
Figures 3(b) and 3(c) represent cost-effective options. Multi-criteria analysis is
3
warming. The estimates suggest that if such warming occurred now it would impose
designed to deal with problems where benefits and/or costs can not be valued; it
4
4
net damages equivalent to 1.5% to 2.5% of world GDP on the present world
could be used for the approaches described in Figures 3(a) or 3(b), where non-
5
5
economy (Chapter 6). The regional variation in climate change damage would be
6
economic effects are brought into the analysis, but without a formal way of valuing
6
substantial. Damage in developing countries is estimated to be at least 2% with a
7
those effects relative to the economic impacts. Decision analysis focuses expressly
7
maximum central estimate of 9% of GDP for some regions. Small islands and low
on making decisions under uncertainty and conceptually could be used in any of the
8
8
lying coastal areas are particularly vulnerable. Climate changes of this magnitude are
9
approaches described in Figure 3.
9
not expected to be realized for some decades and damages in the interim would be
10
10
smaller.
Cost-benefit analysis (broadly defined) could contribute to resolution of three
II
11
questions facing decisionmakers. These are:
12
12
Most estimates of the additional damage done by CO2 emitted now range from $5 to
13
13
$25 (1990 $ U.S.) per tonne of carbon, although a few estimates range up to $125
By how much should emissions of greenhouse gases be reduced or sinks
14
14
per tonne of carbon.¹ The models used to develop these estimates are simplistic and
15
of greenhouse gases be enhanced?
15
are poor representations of dynamic processes, but they reflect the current state of
16
16
the art. The wide range of estimates reflects variations in models, scenarios, discount
17
When should emissions be reduced or sinks be enhanced?
17
rates and other assumptions.
18
18
19
What methods should be employed to reduce emissions and enhance
19
Climate change benefits of emissions reduction measures can be very complex to
20
sinks?
20
estimate. Some energy efficiency and fuel switching measures simultaneously reduce
21
21
emissions of greenhouse gases and sulphate aerosols. The negative radiative forcing
22
Because of the large uncertainties surrounding estimates of damages due to climate
22
from sulphate aerosols is significant. "Tropospheric aerosols have a lifetime of only
23
change (Chapter 6) and the costs of mitigation (Chapter 9) make it difficult in
23
a few days So the atmospheric concentration of aerosols responds rapidly to
24
practice to rely solely on cost-benefit analysis for the selection of an appropriate
24
changes in emissions. Control of sulphate emissions would immediately reduce
25
stabilization objective at this time. Still, the information provided by cost-benefit
25
the amount of aerosols in the atmosphere." A recent model analysis finds that
26
analysis is critical in making judgements not only about that decision, but also about
26
replacement of fossil fuels with carbon-free energy sources leads to lower
27
the timing of emission reductions and the design of cost-effective strategies.
27
temperature increases in the long run, but higher temperatures in the interim due to
28
Integrated assessment models best suited to such analyses are quite new (Chapter
28
the reduced sulphur emissions (Chapter 10).
29
10).
29
30
30
Costs of Response Options
31
However, cost-effectiveness analysis can and should be used to evaluate adaptation
31
32
and mitigation options.
32
What matters from a policy perspective is the net cost (total cost adjusted for
33
33
secondary benefits and costs) of a mitigation or adaptation option apart from its
34
The Benefits of Limiting Greenhouse Gas Emissions and Enhancing Sinks
35
The benefits of limiting greenhouse gas emissions and enhancing sinks are the
34
36
13 A tonne is a metric ton or 1,000 kg. $5 per tonne of carbon corresponds to
climate change damages avoided. Since some increase in atmospheric concentrations
35
approximately 0.35c per litre of motor gasoline.
37
38
from the current level is virtually inevitable, some impacts and damages due to
36
14 Radiative Forcing of Climate Change: The 1994 Special Report of the Scientific
39
climate change seem unavoidable.
37
Assessment Working Group of IPCC, Summary for Policy Makers, IPCC, 1994, p.
40
38
24.
7
8
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
climate change benefits (Chapter 8). A wide range of estimates of the costs of
I
same set of input assumptions, such as the cost and availability of different energy
2
response options is found in the literature. The wide range is due in large measure to
2
sources.
3
differences in assumptions. Alternative views of the costs of emissions reduction are
3
4
shown schematically in Figure 4. Curve A assumes that there are no reducible
4
Top-Down and Bottom-Up Models
5
market or government induced imperfections; thus there are no "no regrets" options.
5
6
Curves B, C and D all assume market failures or government induced imperfections
6
In the jargon of modelers "top-down" models are aggregate models, often
7
exist in one or more sectors that can be corrected by policies that reduce costs and
7
macroeconomic models, that analyse how changes in one sector of the economy
8
also lower greenhouse gas emissions. Curve B includes only measures whose direct
8
affect other sectors and regions. Early top-down models tended to have little detail
9
benefits yield net cost savings. Curves C and D incorporate secondary economic and
9
on energy consumption, especially at the technology-specific level, but explicit
10
environmental benefits into the calculation of the net cost of the option.
10
treatment of behaviour and economic interactions. In contrast, "bottom-up" models
11
11
tended to describe energy consumption in detail, however consumer behaviour and
12
The potential for "no regrets" policies is represented by the line OD2. The area
12
interactions with other sectors of the economy tended to be down played.
13
OD,D, is the potential net benefit to society of implementing all "no regrets"
13
14
measures. Some analysts have argued that measures with a positive cost should be
14
This simple characterization of top-down and bottom-up models is increasingly
15
implemented to the point where the total net cost to society is zero, i.e. to a point
15
misleading as more recent versions of each approach have tended to provide greater
16
D,>D₂ such that the area under curve D between D₂ and D₃ is equal to that area
16
detail in the areas that were less developed in the past. This convergence means that
17
OD,D₂.
17
differences in model results are increasingly driven by differences in input
18
18
assumptions rather than model structure. Most results available in the literature do
19
Despite significant differences in views, there is agreement that some energy
19
not reflect the effects of this convergence of model structure since the evolution in
20
efficiency improvements (perhaps to 10% to 30% of current consumption) can be
20
model structure is a relatively recent occurrence.
21
realized at negative to slightly positive costs in many countries, depending on the
21
22
baseline assumptions and the time frame (Chapter 9). Measures to reduce greenhouse
22
Nevertheless, differences in model structure remain important as different models are
23
gas emissions often yield additional economic benefits (such as reduced traffic
23
best suited to answer different types of questions.
24
congestion) and/or environmental benefits (such as reduced emissions of acid rain
24
25
and urban smog precursors). The magnitude of these secondary benefits depends on
25
Equity and Social Considerations
26
local circumstances. Studies for Norway, the U.K. and some other countries indicate
26
27
that secondary benefits could offset 30% to 100% of abatement costs (Chapter 6).
27
Equity considerations are an important aspect of climate change policy and feature
28
28
prominently in the Convention. Inaction is likely to impose costs on future
29
The long run costs of addressing climate change will ultimately be determined by
29
generations and on regions where damages occur, often regions with low greenhouse
30
the rate of capital replacement. the discount rate and the effect of research and
30
gas emissions. Action to address climate change is likely to impose some costs on
31
development. Implementation of "no regrets" measures will increase the time
31
the present generation. Since actions to address clinfate change are deliberate,
32
available to learn about climate change and to bring new technologies to the market.
32
explicit decisions can be made as to how to share these burdens.
33
Infrastructure decisions are crucial because they can enhance or restrict future
33
34
options and different choices can lead to very different cost outcomes.
34
The Convention recognizes the differentiated responsibilities of the Parties, but it
35
35
does not include an explicit formula for sharing the costs of addressing climate
36
Estimates of the costs of stabilizing CO2 emissions vary widely as a result of
36
change. It also lists factors to be considered in determining how the burden should
37
differences in the baseline scenario, the estimated long run costs of mitigation
37
be shared and provides for development of mechanisms for sharing the burdens.
38
measures, and the date for achieving the stabilization objective. Differences among
38
Issues of equity and efficiency can for the most part be separated. Cost-effective
39
models are less significant than differences in the underlying assumptions. Indeed
39
strategies can be developed embracing a variety of ways for sharing the burden.
40
widely differing methods can produce quite similar results when calibrated to the
40
9
10
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
Equity issues involved in addressing climate change may be divided into four
I
contributions to the problem, while emissions from others may well start to dominate
2
categories:
2
it.
3
3
4
distributing the costs of adaptation;
4
Countries (such as low lying coastal regions and islands or drylands) and special
5
5
groups within society (such as the poor, and sometimes women or children, or
6
distributing future emissions rights;
6
specific occupations or regions) that are especially vulnerable to climate change -
7
7
those on whom the costs of abatement and coping would be especially burdensome
8
distributing the costs of abatement; and
8
- merit special attention.
9
9
10
ensuring institutional and procedural fairness.
10
All Parties have equal status under the Convention. Procedural equity also requires
11
11
that all Parties be able to participate effectively in international negotiations related
12
The Convention offers some guidance on all of these issues, but doesn't resolve any
12
to climate change. Assistance to enable developing country Parties to participate
13
of them.
13
effectively in negotiations would increase the prospects for achieving effective,
14
14
lasting and equitable agreements on how best to address the threat of climate
15
A variety of ethical principles, including the importance of meeting people's basic
15
change.
16
needs, are relevant to sharing the burden of addressing climate change. No single
16
17
rule will command universal agreement. Furthermore, any arrangement may need to
17
Intergenerational Equity
18
change over time to continue to be perceived as equitable and to take new
18
19
developments into account. Analysts can not prescribe how the burden of addressing
19
Climate policy raises particular questions of equity among generations, because
20
climate change should be shared, they can only clarify the implications of different
20
future generations are not able to influence directly the policies being chosen today
21
principles (Chapter 3).
21
that will affect their well-being and because it might conceivably not be possible to
22
22
compensate future generations for reductions in wêll-being that might be caused by
23
Countries differ in terms of current greenhouse gas emissions (see Figures 5 and 6),
23
current policies.
24
vulnerability to climate change (Chapter 6). wealth, resource endowments, and
24
25
institutional capacity to respond effectively. The Convention provides that these
25
Sustainable development and the concept of an environmental trust have been
26
circumstances be reflected in differentiated responsibilities. Developed countries
26
proposed to address intergenerational equity. Sustainable development would meet
27
have undertaken commitments through the Convention for technology transfer and
27
"the needs of the present without compromising the ability of future generations to
28
contributions to the financial mechanism.
28
meet their own needs." An environmental trust, as proposed by Brown-Weiss, would
29
29
require each generation to ensure that the next inherits "a planet and cultural
30
A rigid "north-south" delineation of equity issues is inappropriate and may be highly
30
resource base at least as good as that of previous generations."
31
damaging in the long run, although it permeates much of the debate at present. The
31
32
implications of climate change for developing countries are different than those for
32
Discount Rate
33
developed countries because the former are generally poorer, have contributed much
33
34
less to past emissions and still emit much less per capita, have shorter policy time-
34
The discount rate is the analytical tool economists use to compare economic effects
35
horizons, often weaker institutions, other urgent priorities, and are generally more
35
that occur at different points in time. Identifying the proper discount rate is probably
36
vulnerable to climate change. However, there are substantial variations within the
36
the single most important analytical step in economic analysis of global warming.
37
"north" and within the "south" in terms of absolute and per capita emissions, the
37
The choice of discount rate is important in climate change analyses because the time
38
likely impacts of climate change, institutional strengths and preferences, and
38
horizon is extremely long and the abatement costs tend to come earlier than the
39
endowment of natural resources that may be affected by mitigation. Similarly, over
39
benefits of avoided damages. It is also the most profound ethical question, since it
40
the next century some developing countries will continue to make marginal
11
12
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
I
inherently confronts present pain of abatement cost against damages suffered by
I
climate change, however, society is not likely to be able to set aside investments
2
future generations if no action is taken.
2
over the next three centuries and earmark the proceeds the eventual compensation of
3
3
those adversely affected. Determining the discount rate on the basis of descriptive or
4
The literature on the appropriate discount rate for climate change analysis can be
4
market considerations yields a global rate in the range of 3% to 6% real, with rates
5
divided into two approaches. The prescriptive approach asks whether a particular
5
declining within this range as income rises.
6
action would make society better off, given a plausible social welfare function. The
6
7
discounting rules specify how consumption at different dates should be valued to
7
Increasing the discount rate because of risk is generally inappropriate.
8
provide the correct answer to that question. The descriptive approach focuses on
8
9
intertemporal efficiency and opportunity costs.
9
Policy Instruments
10
10
11
The prescriptive approach discounts consumption accruing to different generations
II
A clear distinction needs to be drawn between actions to reduce the impacts of
12
using the "social rate of time preference" (SRTP). The SRTP is composed of two
12
climate change and actions to reduce emissions. The former also involve
13
components, "pure time preference" and the incremental welfare derived from future
13
consideration of adaptation policies.
14
consumption:
14
15
15
Policy instruments designed to mitigate climate change need to be assessed at the
16
Pure time preference is a reflection of impatience. Although there is
16
international and domestic levels. Such instruments can help participating countries
17
debate in the literature, values tend to range from zero to 1%
17
comply with multilateral commitments on climate change abatement, including
18
18
international transfers of resources and technology. The international and domestic
19
The incremental welfare of future consumption accounts for rising per
19
policy instruments need not be the same, but international agreements may constrain
20
capita income over time and the resulting decline in the incremental
20
the choice of domestic policy instruments.
21
welfare (marginal utility) derived from the higher future consumption.
21
22
22
The policy instruments available internationally and for adoption by groups of
23
Depending upon the values assumed for the different parameters, values for the
23
countries include: non-tradeable quotas, tradeable quotas, joint implementation,
24
SRTP on a global basis tend to fall between 0.5% and 3.0% per year. Developing
24
harmonized domestic carbon taxes, international carbon taxes, various international
25
countries may have much higher (10% to 20%) rates of SRTP due to high rates of
25
standards and technology transfer. Domestic policy instruments include carbon taxes,
26
per capita income growth or high elasticities of marginal utility for small increases
26
tradeable permits, deposit-refund systems, technology standards, performance
27
in per capita income from subsistence levels.
27
standards, product bans, voluntary agreements, energy taxes, energy efficiency
28
28
subsidies, non-fossil fuel subsidies, and removing éxisting market distortions.
29
It is important remember that in making choices concerning alternative policies, that
29
30
while differences in consumption at different dates will be valued using the SRTP.
30
Among the criteria that governments can consider when selecting from among the
31
those differences in consumption will be affected by the opportunity cost of capital,
31
available policy instruments are: distributional (including intergenerational) equity,
32
which typically is in the range of 7% to 15% real. Thus, where mitigation projects
32
flexibility in the face of technological change and new information, efficiency or
33
displace other investments yielding higher returns, the mitigation projects typically
33
cost-effectiveness, compatibility with the institutional structure and existing policies,
34
will not be chosen unless they yield returns at least equal to those on the displaced
34
and understandability to the general public. A mix of instruments may be needed to
35
capital.
35
achieve the best results. Governments may apply different criteria with different
36
36
weights to the selection of international and domestic policy instruments.
37
The descriptive approach to the discount rate looks at returns to investments in the
37
38
real world. Market rates of return usually exceed the values estimated for the SRTP.
38
Cost-effectiveness should always be a criteria for selecting policy instruments, but it
39
Conceptually, funds could then be invested in projects that provide a higher return
39
becomes more important at both the international and domestic levels as the
40
with the proceeds being used to increase future consumption. In the context of
40
abatement effort becomes more stringent. The economic literature suggests that the
13
14
JANUARY 20 DRAFT. DO NOT QUOTE. FOR REVIEW ONLY
1
more economically efficient instruments include carbon taxes and tradeable quotas.
2
Although there are implementation problems associated with each, both can be made
3
consistent with equitable international burden sharing. The consequences of climate
4
change policy will be determined by the choice of policy instrument as well as the
5
design and implementation of those policies. The effectiveness of policies will also
Anthropogenic CO,
emissions in 1990
6
depend on their acceptability to the public and to governments.
7
2300
Figure 1
Illustrative Anthropogenic Emissions Profiles of CO, Leading to
Stabilization at Concentrations of 350 to 750 ppmv Using a Mid-range
Carbon Cycle Model and CO2 Emissions for the IS92a, c and e Scenarios
Source: 1994 Special Report of the IPCC
2200
S750
S650
S550
S450
Year
IS92a
IS92c
2100
Anthropogenic emissions (GtC/yr)
S350
IS92e
2000
20
15
10
5
0
194 130003 , I
15
Figure 3
Global Warming, Cost-Benefit
anthropogenic CO2
emissions in 1990
Analysis and Sustainable Development
Damage
and
Abstement
2050
Costs
Minimum
Risk Zone
Zone of Uncertainty
S750
S650
S550
S450
S350
Comparison of IPCC Emissions Scenarios Adjusted for Current
Commitments and Emissions Profiles Consistent with Stabilization of
Unacceptable Flisk
0
2040
AT
Atmospheric Concentrations of 350 to 750 ppmv
Source: Based on Chapter 12/adjusted for current commitments by Annex I Parties extended to 2050
^ Tmax
IS92c
(a) Absolute Standard Approach
IS92eC
IS92aC
2030
Damage
Marginal Cost
and
Abstement
Abelement
Unacceptable Risk
Costs
Figure 2
IS92a
2020
Legend: IS92aC (IS92eC) is scenario IS92a (IS92e) 4
Unacceptable
Cost
Acceptable Cost
0
A Tmax
AT
Anthropogenic emissions (GtC/yr)
IS92e
2010
(b) Safe Minimum Standard Approach
2000
Marginel Cost
Marginal
Damage
of Abstement
Damage
and
Abstement
Costs
1990
13
12
11
10
9
8
7
6
5
4
3
2
1
0
0
^ Topi
AT
$ 038003 , .
(c) Cost-Benefit Approach
Source: Chapter 6
938005 figs 1/95 -
Figure 4
Alternative Views on Costs of
Emission Reduction
Marginal
Cost
A
B
C
D
0
D₂
Emission
D₃
Reduction
D,
A No reducible market imperfections and/or distantion
B - - "Negative cost potential" achieved
C - Curve B plus "economic double dividends"
D - Curve C plus "environmental double dividends"
Source: Chapter 8
938005 Age 1A5 up
Figure 5
Carbon Emissions per Capita and Population, 1993
8 tC/cap
6
4
2
0
Africa
Former
India
China
Japan
Other
Latin
US
USSR
Asia
America
EU12
Other
Australia
Canada
Europe
Population, Billion
Source: Derived by M. Grubb from BP Statistical Review of World Energy 1994 and World Population Prospects, UN
938005 fign 1/95 mg
Figure 6
Carbon Emissions/GNP and Total GNP, 1992
Carbon emissions/GNP (kg/US$m)
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Africa
Former
India
Japan
Australia
Latin
US
Canada
USSR
China
America
EU12
Other
Other
Europe
Asia
World total GNP, $bn (market exchange rate)
Source: Chapter 3
936005 - 1/95.mp
Chapter 4
This chapter has improved over earlier versions, but still could
use some work. One general suggestion would be to have the
chapter consider efficiency on a level more comparable to the
current emphasis on equity. For example, shouldn't efficiency be
in the chapter title and receive more attention in the summary?
Also, there could be some additional explicit discussion of what
efficiency means in the intergenerational context; the standard
view that efficiency is a less controversial criterion than
equity has some merit for this application.
p. ii, line 20. How accepted in Brown-Weiss's argument? My
sense is that it is not supported by any treaty. Rather it is
more an ethical norm. Does terming it an international legal
principle give it more weight than other philosophical criteria
(e.g., Rawls's difference principle or Nozick's theory of
rights) ?
p. 11, line 32. The chapter does not explicitly recommend
sensitivity analysis, despite the wide range of rates that can be
justified. Shouldn't sensitivity analysis be recommended? If
so, what would the recommendation be -- perhaps use of social
rates of time preference of 2 to 4 percent with commensurate
shadow prices of capital? If so, the shadow price of capital
(SPC) or a range for this measure should be suggested.
Higher discount rates should be considered in the case where no
adjustment is made for the SPC (e.g., 2 to 8 percent).
p. 1, line 14. Why is the alternative chapter title posited? If
efficiency is an important consideration in discounting -- and it
surely is -- then it should be mentioned in the actual title and
should get adequate treatment in the chapter.
p. 2, line 11. The flip side of the collapse of future values
due to discounting is that very small displaced investments can
lead to very large future shortfalls in output.
p. 2, line 38. The implication that markets do not effectively
allocate goods over time (and particularly over generations) is
controversial. It affects more than global warming policy, and
raises significant questions about national savings policy and
use of exhaustible resources. I do not think that there is, in
fact, a consensus among economists that market interest rates are
generally incorrect guides to behaviors in those settings.
p. 2, line 43. "the apply" should be "and then to apply".
p. 3, line 21. The compensation problem deserves some
discussion; it is relevant for any discount rate.
p. 3, line 33. See comment for p. ii, line 20.
5
p. 3, last line. Why "must" discussions of intergenerational
equity go beyond sustainability? Brown-Weiss, for example,
implies that sustainability is a "legal" (though perhaps more
accurately, "a plausible ethical") obligation that generations
face. What evidence is there of a consensus that governments
should be allowed to make poorer generations suffer so long as
richer generations gain sufficiently? Arguably this view is
controversial, yet it is implicit in the utilitarian framework
that drives the equity analysis in this chapter. While this
framework is important ànd merits exploring, it should be made
clear that this is not the only possible framework. It should
also be explained to policymakers that despite the technical
content of this discussion, the material covered is quite
normative, and that while economists have expertise in analyzing
these norms, they are not in a position to determine them for
society.
p. 4, box. The discussion of Schelling's view is too abstract to
be useful to policymakers. Regarding the conclusion in the box,
it is not obvious that the analysis presented in the chapter --
which, in effect, suggests consumption transfers across
generations to help equate marginal utility -- is appropriate in
a world where richer nations make only modest transfers to poorer
ones.
p. 5, line 12. The sentence on Thurow's paper should probably be
moved later in the paragraph.
p. 5, line 30. The discussion in the prior section is also
prescriptive.
p. 6, line 38. Should add the following sentence to this
paragraph: "However, given appropriate estimates of relative
prices, there is no reason to explicitly modify the discount
rate."
p. 7, line 3. The social welfare function and related
relationships used to derive this equation should be stated
explicitly. This is the most important equation in the chapter,
but it is not clearly derived for readers. A reference to the
annex and a full presentation there would be useful.
p. 9, line 19. The finding that the optimization problem is not
well defined should be explained somewhat more.
p. 11, line 22. Insert "absolute value of" before "elasticity".
p. 11, line 26. Are the low rates of productivity a U.S.
phenomenon or a world phenomenon?
p. 11, line 38. See comment for p. 11, line 22.
6
p. 12, line 11. What is meant by "The more weight the society
gives to intergenerational equity"? There are all kinds of
equity criteria, each with different values of Θ. One view of
equity would be very protective of worse-off generations; another
might be less SO.
p. 12, line 27. While there are a number of equity reasons to
intervene to help the very poor, the idea that government should
generally fine-tune the income distribution is less clear. At
some point government intervention can cause inefficiencies and
can run into other equity problems, such as those associated with
confiscation.
p. 12, line 40. Should "even" be "particularly"?
p. 14, line 13. Shouldn't one state that based on "certain"
ethical grounds, it is tough to support the higher time
preference rate?
p. 14, lines 34-5. The World Bank review and other study should
be properly cited. Were they comprehensive or selective reviews
of completed projects? Were poor projects dropped by virtue of
being uncompleted?
p. 15, line 30. Replace "most plausible discount rates" with
"discount rates below 10 percent".
p. 16, line 9. The estimate of a 15.1 percent real return to
nonresidential capital over the period 1975-90 seems very high.
The estimates should be properly attributed to their respective
authors.
p. 16, line 33. Wouldn't general economic growth address a large
share of the damages from warming? Moreover, compensation of
future parties that lose from particular policy choices is
difficult regardless of the discount rate chosen.
p. 17, line 1. The existence of multiple market failures is a
classic problem of the second best. It is not obvious whether
piecemeal fixes would make things better or worse.
p. 17, line 19. Footnote 3 is unclear. Risk aversion (declining
marginal utility) is supported by data, for example.
p. 18, line 17. If there is no intertemporal welfare
maximization by the government, then how can one assert that
market rates are necessarily wrong? Couldn't the lack of
intervention be viewed as an endorsement of market rates?
p. 18, line 29. The work presented by Lyon at the IPCC meetings
shows that a shadow price of capital of 1.5 to 2 may be quite
low, given the rates of time preference assumed by Cline and
7
others. For example, with no reinvestment of capital income
(except for coverage of depreciation), a rate of time preference
of 2 percent, and a return to capital of 8 percent, the SPC will
be 4.0. If 10 percent of net returns to capital are reinvested,
the SPC rises to 6; and if 20 percent of net returns are
reinvested, the SPC rises to 16. Moreover, if the rate of time
preference is 1.5 percent -- as is argued at times in chapter 4 -
- then the SPC under the last set of assumptions would be
infinite. (Note that the total weight on costs will be below
these values, as only the portion of costs coming from investment
gets weighted by the SPC.)
p. 19, line 1. The relationship of the discussion in this
paragraph to the shadow price of capital approach is unclear.
p. 23, line 24. This derivation of the discount rate could be
explained a bit more. Equation 4A.1 should be maximized, and the
relationship of the interest rate to this equation should be
explained.
p. 24, line 4. Again, intergenerational equity can be far more
complex than determination of Θ. See comment for p. 12, line 11.
p. 25, line 20. Again, the discount rate would not be negative
if prices are appropriately adjusted. See comment for p. 6, line
38.
p. 25, line 23. The relationship of the Sandmo-Dreze approach to
the SPC approach is unclear.
p. 27, line 17. Should the statement refer to G=U or G not equal
to U?
8
a) If new technologies will become available in the near to
medium term, it may be much less expensive to wait to reduce
emission levels beyond modest amounts. This is one of the basic
findings of studies cited elsewhere in the IPCC report. If the
authors of this chapter disagree with those studies, they should
state the reason, not just assert an alternative conclusion.
b) The argument given earlier suggests that the shadow price
of abatement today may be considerably lower than in 100 years.
c) Though the argument on p. 9 that there may be large costs
for sudden reductions is correct, the analysis does not take into
account the important role of fixed costs, which suggest that while
new investments should take into account the desirability of
abatement, levels of abatement will increase only gradually.
p. 1 The questions posed on page 1 are not well formulated. Key
terms are not well defined. For instance, what does the word
"tolerated" mean? Question 3 confuses issues of efficiency (where
abatement occurs) and finance (who should pay the costs of
abatement).
p. 2 "In this connection, we note that accordign to the earlier
IPCC report, even present information is adequate to go for the
goal of reducing gloabl emissions by 60% for stabilisation." The
sentence is difficult to follow. "Go for": Does this mean,
justified by a cost benefit analysis? The earlier IPCC report did
not have a cost benefit analysis. "stabilisation": stabilisation
at what level?
p. 7." risks are measured in CO₂ years.' I think I know what
the authors are trying to get at, but it appears to be based on the
wrong model (or at least a questionable model.) The sentence which
follows is even more questionable. (Assume we were fairly
confident that concentration levels below some threshold
substantially higher than current levels were safe. Then increases
in concentration levels in intervening years would pose no risk.
And if the half life was short enough, most of the CO₂ emitted
today would be dissipated by the time threshhold levels were
attained: current emissions would pose little (integrated) risk to
global warming.) The final sentence, however, is correct
qualitatively, that is, increased concentration levels today have
adverse consequences for future generations, by making whatever
task they have in limiting concentrations more difficult.
p. 8 Discussion of absorption capacity confuses analytic and
scientific issues with normative issues. Thus, the fact that
absorption capacity may be endogenous and changing does not affect
the conclusion that absorption reduces net emissions. The question
of the extent to which absorption levels are affected by
concentration levels is an important one, and deserves more than
the passing reference given here.
Allocating the per capita absorption capacity of the world
uniformly is one "normative" basis, but not the only one, as the
authors indicate with the question, "Is this the right way to
allocate absorption?" No hint is given of the other factors that
might be considered. For instance, should countries that have
brought their population growth rates under control be given
"credit" for doing so? Proposals which allocate pollution rights
simply on a per capita basis have severe problems of intellectual
coherency.
p. 8 The risk analysis at the bottom of p. 8 is faulty, or at
least based on special assumptions. The authors should first state
the general optimization problem to be solved, and then state the
asserted qualitative properties, related to certain empirical
parameters. Among the objections to the conclusions stated are the
following:
02/24/95
12:48
9 202 6221294
TREASURY E
002
SECTION I: BIOGRAPHICAL INFORMATION
Name:
Joseph E. Stiglitz
Organization:
Council of Economic Advisors, Washington
Area of Expertise:
Microeconomics
Phone:
202-395-5036
Fax:
202-395-6958
E-mail:
none
The following comments are on
Chapter 2: Decision Making Framework
SECTION II: SUMMARY OF COMMENTS
Chapter is not yet in a form that can be effectively reviewed by experts
or governments.
^
Chapter presents normative conclusions unsupported by
scientific argument. This is not appropriate to a
scientific assessment
Chapter fails to focus attention on its major conclusions;
the conclusions it does highlight are often at variance with
economists' theory and empirical results.
Logical foundation for many claims is shaky.
Chapter does not make a balanced presentation of the issues.
Writing is not up to the standard of a major IPCC
publication.
SECTION III: DETAILED COMMENTS
Notes:
1)
The rough state of the present draft makes it difficult to
give helpful comments yet;
2)
These comments are only a small fraction of the many that
could be offered. Addressing these comments alone without
other major work would probably still not result in a
chapter that would gain acceptance in the economics
profession.
p. <Summary, page unnumbered> 1. 23-24. Causal relationship
not established.
p
2, section 2.1.2. Uncertainties are summarized better in
02/24/95
12:48
9 202 6221294
TREASURY E
003
Chapter 1; this list should be improved or deleted.
p. 3, 1. 2-5. "Some expect " This is what economists
generally view. believe, and should be presented as the prevailing
p. 3, point 3. Text leaves the impression that Cline
damage estimates, which have not received wide acceptance among
economists are the accepted values. Nordhaus, Tol and
Fankhauser have also published comprehensive damage estimates.
p. 4, section 2.1.2.1. This paragraph assumes the
conclusion. NOXT to last rentence, Normative conclusion no epet denved
p. 5, 1. 6-7.
sea level rise in Indian subcontinent
from
could be higher than in, say, North America. If How would the
Indian Ocean rise more than the Atlantic and Pacific Oceans?
analysis
p. 5, 1. 8. Insert words in bold: " could have more or
less impact
"
p. 5, figure 2.1. Delete temperature estimates. It is not
the accepted view that for 2°C warming, damage would be
"substantial. Also, graph appears to present probability
density functions (but then what is the curve that starts highest
on the probability axis?) If that is the case, the areas under
the curves should be the same, the tails of the curves should
extend to the left of the origin (implying some probability of
benefits from warming), and other logical problems should be
corrected.
R.5, figure 2.2. Brobability of welfare impacts is
irrelevant to the discussion I suggest a graph of delta
Temperature (x-axis) against delta Welfare (y-axis).
p. 6, 1. 32. Text missing.
p. 7, 1. 14-16. Assumes the conclusion, which may be wrong.
12 Need to explain 1A what some NICKIS measured -years
p. 8, 1. 10 ff. "
if humanity lived a lifestyle
corresponding to about 0.66 tonne of carbon equivalent emissions
per person, per year " With no warning, sentence suddenly
appears to endorse one side of a delicate political issue; out of
place in a scientific assessment.
p. 8, section 2.2.3. trend risk and not a sudden
risk.
" What does this mean? L. 28: " higher abatement
costs are justified in case of trend risk. " Why is this?
Because CO2 in atmosphere has an infinite lifetime?
p. 9, 1. 1. "Therefore, small but regular
...
if not
impossible.' This may be correct, but it is not self-evident and
does not fit here.
Coreept of "acceptable GT" not dofined
has no generally accepted defin TM in economics)
02/24/95
12:49
9 202 6221294
TREASURY E
004
p. 9, section 2.2.4. Why insert this political argument
here? Political arguments belong in the INC.
p. 11, 1. 1. What is the "apex forum?"
p. 11. 1. 11. Change sentence: "To a few low-lying island
states, climate change may threaten their existence. To oil
exporters, climate change mitigation measures threaten their
revenues."
p. 11, 1. 12. "Again, it concerns differently economists and
environmentalists." What does this mean?
p. 12, 1. 7. Parry and Rosenzweig (cited) do not find large
vulnerability in food supplies, contrary to impression left by
text.
p. 13, section 2.3.2.2. Economists will not take seriously
this presentation of "Ecocentrism, " e.g., "To put it sharply,
humankind's abrogation [?] of decision making power's regarding
nature is questioned. The ecocentric view proposes to curtail
such decision making power Equity is not only among human
beings but also between human beings and other living species."
In economics, value is generally defined with reference to
willingness to pay. Resources have value because of someone's
to
willingness to pay for their use (or even just for their
existence). Section 2.3.2.2 appears rest its argument on the idea
that a resource may have value to other species; do other species
have a willingness to pay, apart from our willingness to pay to
protect them or make them happy?
p. 14, 1. 20. "While the rich can argue " This political
argument does not belong in a scientific assessment.
p. 15, 1. 6. " the practice of discounting the future is
severely questioned, etc. Sentence should be deleted, as
address
Chapter 4 settles this issue. Also, Cline's estimate cited in
first paragraph is not widely accepted by economists. This
reference should carry a note to that effect, or should be
deleted.
"This substitutability assumption is highly questionable. This
statement assumes the conclusion. The section should either make
the case for its assertion, or be silent.
p. 16, 1. 4. " then not doing anything is morally
unacceptable." This is a moral judgement, not a statement of
scientific agreement.
p. 16, section 2.3.3.1. First paragraph should also note
that cutting back on growth in the North -- for example as a
result of climate change programs -- is likely to hurt residents
of the South as well, through a drop in southern exports.
02/24/95
12:50
9 202 6221294
TREASURY E
005
p. 18, 1. 5. "That is, due to delay in abatement, global
risks are externalized." Alternatively, spending a lot now on
the wrong programs could easily impoverish our descendants.
p. 18, 1. 25. "Thus risk tolerance of, say South should
matter more than that of North." This conclusion is unsupported.
"
homeless dying from heat stress
" At least in the U.S.
many more homeless die during winter than during summer.
p. 18, 1. 27. "Most models developed in the North do not
effectively build in the perspectives of the South in adapting to
climate change. Economic models, like physical models, should
be value-free. Do we speak of Northern physics and Southern
physics -- or just good physics and bad physics? An economic
model may be flawed for several reasons: it may be poorly
formulated, or based on flawed estimation results; it may be
internally inconsistent, or may predict poorly. We may say that
it fails to reflect available resources, or misstates likely
behavioral responses, but not that it does or does not "build in
perspectives of the South."
p. 19, section 2.3.3.4. This section presents political
conclusions as if they were scientific conclusions, e.g.,
"Currently there are many countries whose use of environmental
space far exceeds their justifiable share of that space. J.
Parikh has indicated that these privileges are worth US $70
billion per year.'
The section also presents apparently scientific conclusions that
are probably wrong, e.g., "The cost of delay in emission
reduction (by the North) in terms of South's foregone
opportunities to development is substantial." (see Richels and
Edmonds 1993, Schlesinger and Zhang 1992, etc.)
p. 20, section 2.4.1. This section -- and indeed the entire
chapter -- inappropriately presses a certain political point of
view. For example, 1. 28: "Remember that if 550 ppm is chosen,
it is difficult to go back to 440 ppm, but vice versa is
possible."
p. 22, above table. "This is referred as purchasing power
parity." Wrong use of the term.
p. 27, section 2.5. Why begin with the quote from
Dowlatabadi and Morgan? Placed at the top of the section
"Economic Formulation of Individual and Collective Decision
Making", it suggests that the most important consideration is the
prejudices of the decision makers. These may indeed be
important, but the contribution of scientific analysis lies
elsewhere.
p. 28, 1. 17-20. II Bayesian approach, which represents
information in the form of conventional, necessarily precise,
02/24/95
12:51
9 202 6221294
TREASURY E
006
probability distributions " The writer appears to be unaware
of the voluminous and widely disseminated work on decision
analysis that has appeared since the late 1950s.
p. 29, 1. 26 ff. "From such a standpoint, a subjective
probability can always be formulated as a 'bet on a certain
event'. " What does this mean?
p. 29, 1. 12. "In sum, climate change seems to imply a
certain degree of irreducible unpredictability beyond the common
categories of risk and uncertainty." Climate change is a
difficult analytical problem, but not for this reason. Decision
analysis has been applied to many questions involving great
uncertainty.
p. 31, 1. 3. "The most catastrophical fluctuation yet
analyzed " Why describe this event as catastrophic? I don't
believe we have any evidence of any large die off of plants or
animals, or of an unusual number of species extinctions.
p. 31, 1. 7. "The discrepancies between the paleoclimatic
observations of abrupt change and the smooth gradual warming
prognosticated by today's models could not be more pronounced."
This is false and irrelevant. GCMs can in general show rapid
changes.
p. 32, 1. 13. If low probabilities attached to large
potential damages." This is not a fatal problem, as the text
claims.
p. 35, 1. 8. " there might be as many rationalities as
there are cultures." This may or may not be true, but it adds
nothing to an assessment of the state of scientific knowledge.
p. 38, 1. 11 ff. The writer does not seem to be familiar
with the terms that economists use in describing risk. E.g.: "A
risk averter can be a diversifier and spread his portfolio over
different assets--or a "plunger" and invest wholly in bonds and
money. " "Cautiousness: Defined as decreasing risk aversion
...
(Schaifer, 1961; after Pratt, 1995) " It would be useful to
consult a basic text such as Raiffa (1964).
p. 38, 1. 26 ff. This section strings together terms from
decision theory like cranberries on a Christmas tree, with no
explanation of their meaning, and no attempt to relate them.
Throwing around jargon and padding it with scholarly references
will not help most readers.
p.
40, 1. 10. " trade securities insurance contracts."
Of what relevance is this here?
p. 40, 1. 15. "According to Shlyakhter et al. (1995), those
risks that fall below a particular threshold of probability--and
thereby are ignored by a particular group or society--are called
02/24/95 12:52 89 202 6221294
TREASURY E
007
"de minimis" risks. [emphasis in original]. This is not a
research result but merely a definition.
p. 40, 1. 29. " 5 percent chance of a climate-related
catastrophe within their lifetime. It is difficult to imagine
what climate-related catastrophe could occur by, say 2050.
p. 44, 1. 6. implies that the industrialized nations
should take strong measures " "[emphasis in original]. This is a
political conclusion. Why is it included here?
p. 54, 1. 14. "
would it not be cautious to insure
?" This statement assumes the conclusion, and inappropriately
makes a policy recommendation.
p. 55, 1. 5 ff. Paragraph is highly speculative and does not
belong here.
p. 55, footnote 19. "The more risk averse you are, the more
you will insure (Chichilnisky, 1994) "[emphasis in original]
This is not a research result.
p. 56, 1. 8. "Economics is about differences in preferences
leading to trade (Chichilnisky, 1994) " This is not a research
result.
p. 56, 1. 9. "Betting on climate states " What is this
sentence saying?
p. 56, 1. 23. "To illustrate, the Alliance of Small Island
States (AOSIS) has proposed that an "International Insurance
Pool" be established " Here, and in many other places, a
tangential discussion deflects the reader's attention from the
major arguments. This should go in a footnote or be deleted.
p. 60, 1. 4. "An operational model to analyze the optimal
portfolio of climate change policies for a country is not
feasible given the state of our knowledge." No economic model
will produce an error-free, bullet-proof answer to anything; on
the other hand, several models already exist to examine precisely
the question stated. The writer may want to consult recent
papers of the Energy Modeling Forum at Stanford (for projects
EMF-12 and EMF-14) for more details.
.p. 64, figure 2.8. What is the message of this figure?