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
415892669
label
Morality and Economics, [1988]
core
doc
dtoType
document
citationUrl
pageCount
1
Source metadata
id
415892669
contentType
document
title
Morality and Economics, [1988]
citationUrl
identifierLocal
13896-005
collections
Records of the White House Office of Speechwriting (George H. W. Bush Administration)
Tony Snow Subject Files
imageCount
1
hasImages
yes
source
import
hasTranscription
no
Source extras
naId
415892669
levelOfDescription
fileUnit
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
6dd5181bf93dc496
ocrText
Originally Processed With FOIA(s):
FOIA Number:
S
FOIA
MARKER
This is not a textual record. This is used as an
administrative marker by the George Bush Presidential
Library Staff.
Record Group/Collection:
George H.W. Bush Presidential Records
Collection/Office of Origin:
Speechwriting, White House Office of
Series:
Snow, Tony, Files
Subseries:
Subject File, 1988-1993
OA/ID Number:
13896
Folder ID Number:
13896-005
Folder Title:
Morality and Economics, [1988]
Stack:
Row:
Section:
Shelf:
Position:
G
18
29
2
4
SP
SOCIAL
PHILOSOPHY
PC
&
POLICY CENTER
MORALITY AND ECONOMICS
DE MORIBUS EST DISPUTANDUM
Viktor Vanberg
6
Transaction Books
New Brunswick (USA) and London (UK)
Editor of Original Papers: Ellen Frankel Paul
Managing Editor: John Ahrens
The Social Philosophy and Policy Center, founded in 1981, is an inter-
disciplinary research institution whose principal mission is the ex-
amination of public policy issues from a philosophical perspective. In
pursuit of this objective the Center supports the work of scholars in the
fields of political science, philosophy, law, and economics. In addition
to the Original Papers Series, the Center hosts scholarly conferences,
and publishes both a professional journal, SOCIAL PHILOSOPHY &
POLICY, and a book series. For further information on the Center,
write to the:
Social Philosophy and Policy Center
Bowling Green State University
Bowling Green, OH 43403 USA
About the Author:
Viktor Vanberg is an Associate Professor of Economics at George
Mason University in Fairfax, Virginia, and a Research Associate at the
Center for Study of Public Choice. Before coming to the United States in
1983, he taught at universities in West Germany. His current research in-
terests are constitutional economics and the economic analysis of rules
and institutions. He is the author of numerous publications (in both Ger-
man and English) on the individualistic tradition in social theory, and
particularly the application of this theory to the study of social institu-
tions. Among his more recent publications are articles in Economics and
Philosophy and the Journal of Law, Economics, and Organization.
SP
SOCIAL
PHILOSOPHY
PC
& POLICY CENTER
MORALITY AND ECONOMICS
DE MORIBUS EST DISPUTANDUM
Viktor Vanberg
Original
Transaction Books
Papers No. 7
Published by the Social Philosophy and Policy Center
and by Transaction, Inc. 1988
Copyright © 1988 by the Social Philosophy and Policy Center
All rights reserved. No part of this publication may be used or
reproduced without written permission from the Social Philosophy and
Policy Center, except for brief quotations in critical articles or reviews.
Library of Congress Catalog Number: 87-19213
ISBN: 0-912051-19-1
Printed in the United States of America
Library of Congress Cataloging-in-Publication Data
Vanberg, Viktor.
Morality and economics.
(Original papers; no. 7)
At head of title: Social Philosophy and Policy Center.
1. Economics-Moral and ethical aspects. I. Bowling Green State
University. Social Philosophy & Policy Center. II. Title. III. Series:
Original papers (Bowling Green State University. Social Philosophy &
Policy Center); no. 7.
HB72.V268
1987
174'.26
87-19213
ISBN 0-912051-19-1 (pbk.)
MORALITY AND ECONOMICS
DE MORIBUS EST DISPUTANDUM
tion and those who misperceive reality can "rationally" be moral is not
an attractive alternative either. What one would certainly prefer is to
live in a society which is structured in such a way that morality is a
rational choice, at least for most of the individuals involved. And to the
extent that appropriate structures do not emerge spontaneously, it will
be a matter of deliberate political action to make sure that conditions
prevail which allow for a rational morality.
Table of Contents
I. Introduction
1
II. Morality and Moral Behavior
3
III. Morality: Preference or Constraint?
7
IV. Morality and Choice: Theories as Subjective Constraints
11
V. De Moribus Est Disputandum
17
VI. The Calculus of Morality
19
VII. Alternative Approaches: Supergames and
Gauthier's Rational Morality
25
VIII. Morality: Individual Adaptation and Collective Choice
29
IX. Conclusion
33
34
IX. Conclusion
It has often been suggested (e.g., by Edmund Burke in the epigraph to
this essay) that there exists some kind of substitutability between indi-
vidual morality and social enforcement.⁷⁰ In some sense this may be
true. As I have argued in this essay, however, individual morality is
essentially conditioned by social enforcement. In order to be viable,
morality must be perceived as an efficient strategy, and this will only be
the case if rewards for moral conduct and punishments for rule viola-
tions are appropriately allocated in a social community. The current
level of morality in a community always reflects past enforcement of
moral rules, and it will only be maintained by ongoing enforcement.
The level of morality can in this sense, be viewed as social capital that,
in the absence of sufficient reinvestment, will depreciate over time 71 It is
the crucial function of social enforcement-whether formal or infor-
mal-to make sure that morality can, in fact, be perceived as a suc-
cessful strategy. Where informal enforcement ceases to be effective, for-
mal enforcement may serve as a substitute mechanism, though it may
well be that in various respects it is only a poor substitute.
As the foregoing analysis implies, the proper purpose of formal en-
forcement can only be to affect people's constitutional, strategic choices,
not their situational choices. Its purpose cannot be to make sure that to
act morally is always the rational choice. It is simply impossible and
certainly undesirable to organize large-scale social interaction in such a
way. Neither is it necessary. All that is required for a beneficial social
order is that the individual participants have good reasons to assume
that morality is a promising strategy overall. To make sure that always
being moral is a rational strategy is, however, a much easier task than to
make sure that being moral is always rational. Life in a society where the
latter is achieved would probably be extremely unpleasant. And to live
in a society where only those who believe in transcendental remunera-
70 See James M. Buchanan. The Limits of Liberty-Between Anarchy and Leviathan (Chi-
cago: The University of Chicago Press, 1975), pp. 117f., and "The Moral Dimension," pp. 1,
5.
71 It should be kept in mind that throughout this essay, all issues are ignored which arise with
dissent on what "moral rules" ought to be enforced in a social community. The discussion
herein proceeds on the assumption that there exists a set of rules to which obedience is
generally considered desirable, and morality is viewed as the disposition to follow those
rules.
33
forcement mechanisms have changed the relevant payoffs, one should
expect individuals to adjust their "moral strategies" and to redefine their
own demarcation line between the kind of social setting where they
consider morality a reasonable strategy, and those social settings where
this is not the case
If these adaptive changes in people's moral dispositions are considered
I. Introduction¹
a "social problem," moral appeals will hardly be a promising remedy. As
David Hume argued two centuries ago, it is "impossible to change or
Men are qualified for civil liberty in exact proportion to their
correct any thing material in our nature, the utmost we can do is to
disposition to put moral chains upon their own appetites.
So-
change our circumstances and situation.' What would seem to be
ciety cannot exist unless a controlling power upon will and appetite
required are structural changes which-if they are a matter of choice at
be placed somewhere, and the less of it there is within, the more
all-are a matter of social, constitutional choice. Such structural
there must be without."
changes may include institutional reforms, such as various forms of
Edmund Burke
"decentralization," which aim at reactivating to some extent informal
enforcement mechanisms. To a large extent, however, such constitu-
Ever since Adam Smith reasoned about men's Moral Sentiments, as
tional measures may simply be confined to using the means of formal
well as the nature and causes of the Wealth of Nations, the relation
enforcement to compensate for the loss of informal enforcements.
between ethics and economics has been an issue of scholarly interest;
and-judging from publication rates-it seems to be an issue which
recently has attracted considerable renewed attention.²
Current and received contributions on the issue seem to fall roughly
into two categories. The first category includes contributions which, in
one way or another, discuss the potential role of ethical judgments-
judgments about what ought or ought not to be-in economic theory
itself. Such contributions are essentially concerned with the limits and
scope of normative economics and the rational justification of the ulti-
mate criteria of evaluation upon which normative economic statements
are based.³ The second category comprises contributions which, in one
I
I am indebted to James M. Buchanan. Bunshiro Ando, Hartmut Kliemt, and an anony-
mous referee for helpful suggestions and comments on an earlier draft, which I presented
at Professor Karl Brunner's Interlaken Seminar on Analysis and Ideology, May 1986.
2
For a selection of recent publications on the issue, see Allan Buchanan, Ethics, Efficiency,
and the Market (Oxford: Clarendon Press. 1985): James M. Buchanan, "Ethical Rules,
Expected Values. and Large Numbers," Freedom in Constitutional Contract- Perspectives
of a Political Economist (College Station and London: Texas A&M University Press. 1977)
pp. 151-168, and "The Moral Dimension of Debt Financing Economic Inquiry: vol. 23
(1985). pp. 1-6; Georges Enderle. ed., Ethik und Wirtschaftswissenschaft (Berlin: Duncker
& Humblot. 1985): Jack High. "Is Economics Independent of Ethics?" Reason Papers, vol.
10 (1985). pp. 3-16: David Levy, "David Hume's Invisible Hand." The Wealth of Nations-
The Public Choice of Moral Information. Hume Studies, 10th Anniversary Issue (1985), pp.
110-149; Michael S. McPherson. "Limits on Self-Seeking-The Role of Morality in Eco-
nomic Life," D.C. Colander. ed., Neoclassical Political Economy (Cambridge. MA: Bal-
linger. 1984). pp. 71-85: Amartya Sen. et al., "Ethics and Economics," Social Philosophy &
Policy. vol. 2 (1985): George J. Stigler. "The Ethics of Competition: The Friendly Econo-
mists," The Economist as Preacher and Other Essays (Chicago: University of Chicago
Press. 1982). pp. 14-26. and "The Ethics of Competition: The Unfriendly Critics," ibid., pp.
27-37: Manfred Tietzel. "Moral und Wirtschaftstheorie." Zeitschrift fuer Wirtshafts-und
Sozialwissenschaften. vol. 106 (1986). pp. 113-137.
3 On the rational discussion of ethical principles. see Hans Albert. Treatise on Critical
69 David Hume, A Treatise of Human Nature (Oxford: Clarendon Press. 1967). p. 537.
Reason (Princeton. NJ: Princeton University Press. 1985). pp. 71ff.
32
1
way or another, discuss ethics or morality as an aspect of human be-
environment. One factor which will tend to make a person's moral
havior, i.e., as part of the empirical reality studied by economists. Such
dispositions somewhat less responsive to environmental changes is the
contributions are essentially concerned with the explanatory status and
emotional reaction which we describe as "bad conscience," a reaction
significance which theoretical economics, as a positive science, is to
which apparently is learned concomitantly with the process in which
attribute to ethics or morality in its analysis of human behavior.
rule-governed behavior emerges. Once a general behavioral pattern has
This essay belongs in the second category. Its purpose is to analyze
been conditioned in a person, deviations from the rule typically do not
how ethics or morality can be incorporated as a potential explanatory
leave the person emotionally unaffected but, rather, tend to cause a
variable into economic theory. Section II specifies the particular issue to
feeling of uneasiness. As a learned response, however, these emotional
be addressed, and Section III discusses some common conceptions of
reactions may also be unlearned, and they will be if a person experiences
how economics might account for a variable "morality." Sections IV, V,
directly or indirectly that deviations remain without external sanctions.
and VI propose an alternative conception, based on the logic of Gary
That is to say, internal sanctions cannot be considered an autonomous,
Becker's and George Stigler's reformulation of the theory of consumer
independent source of moral enforcement. They are a source that ul-
choice.4 Section VII compares the approach advocated here to the game
timately derives from, and ultimately remains dependent on, the direct
theoretical argument on recurrent PD games and to David Gauthier's
or indirect experience of external enforcement. In the absence of any
arguments concerning "rational morality." Some implications are dis-
external reinforcement, people can hardly be expected to develop a
cussed in Section VIII and a conclusion is drawn in Section IX.
conscience. And an individual's conscience will typically not remain
unaffected if changes in relevant environmental characteristics occur. If
such changes occur, individuals will presumably not respond with in-
stant adjustments in their "moral strategies." Neither, however, can they
be expected to permanently preserve a moral disposition irrespective of
systematic changes in the relevant payoffs.6
What is often described as a process of "moral erosion" in modern
industrial societies may-to the extent that it actually occurs-simply
reflect a rational adaptation of people's moral strategies to systematic
changes in their social environment. The most obvious changes that
occurred in these societies over, say, the past two centuries can be sum-
marized in analytical terms as a progressive replacement of small-group
interaction by large-group interaction, with the typical consequences
which result from the fact that the mechanisms of informal enforcement
which constrain the former are not effective for the latter A par-
ticularly significant aspect of this process has certainly been the change
in the structure and role of the family. It is a person's learning history
within his family of origin which obviously has a crucial impact on what
one may call his moral development or, in more behaviorist language,
his moral conditioning. To the extent that the changing role of the
family and the general weakening or even elimination of informal en-
67 Scott points out. on pp. 106f. of Internalization of Norms, that people's "alertness" to
environmental changes will vary with the "strength" of their moral commitment: "Moral
learning. like all learning. is a matter of degree. Extinction or termination of a response is
slow when learning is relatively complete. But learning varies in its completeness.
There is. a continuum of moral learning. which may be represented as running between
two poles: complete calculation and complete commitment."
4 Gary S. Becker, The Economic Approach to Human Behavior (Chicago and London: The
See. e.g., Roland N. McKean. "Economics of Trust. Altruism. and Corporate Respon-
University of Chicago Press. 1976): George J. Stigler and Gary S. Becker. "De Gustibus
sibility." E.S. Phelps. ed., Altruism. Morality: and Economic Theory (New York: Russel
Non Est Disputandum," The American Economic Review, vol. 67 (1977). pp. 76-90.
Sage Foundation. 1975). pp. 29-44.
2
31
does not mean that individual morality is entirely uncalculated. The
real issue is not whether the individual calculates or not, but at what
level benefits and costs are compared. The "calculus of morality" is, as I
explained above, a constitutional calculus, the very essence of which is
to "compare" the overall benefits and costs of adopting a moral routine
with those of case-by-case choices.
II. Morality and Moral Behavior
In contrast to what is sometimes suggested in sociological treatments
of the subject, people's morality or commitment to "internalized
The notion of ethics or morality apparently implies the idea of some
norms" is never unconditional. It is a disposition which individuals will
normative standards or rules of behavior, the general observance of
not acquire if their direct or indirect experience does not tell them that
which is typically considered to be conducive to the existence of a bene-
moral behavior in general "pays," and which they will not preserve if
ficial and properly functioning social order.⁵ More specifically, one of
their experience systematically changes. Or, in other terms, "moral
the important characteristics of moral rules seems to be that they ask for
commitment requires sanctions for its maintenance as well as for its
a kind of behavior that is considered socially beneficial-i.e., beneficial
origins."6 Whether or not individuals have reasons to learn that moral-
to all individuals in some defined social community-but that appears
ity is an "efficient" strategy essentially depends on the character of their
to be in conflict with the immediate or direct interests of the actor
environment. An individual's social learning over a considerable initial
himself. Moral rules impose "impartial constraints" on the pursuit of
time span typically takes place in small group settings (notably the
individual interests, constraints which are socially desirable in serving
family), a crucial characteristic of which is that individual behavior is
interests that individuals share as members of a social community.6 This
closely monitored and rule enforcement is quite effective. In this type of
may not be perfectly descriptive of all kinds of rules which everyday
environment there are generally good reasons for an individual to ac-
language classifies as "moral." But it is certainly true for a relevant
quire a general disposition to respect "the rules of the game," and indi-
subset of what is commonly perceived as rules of "morality," and it is
viduals tend to carry this disposition with them as they enter into other
this subset which is of interest for the purposes of this essay. The rules
social settings. This tendency to generalize. to transfer a behavioral
that require us "to keep promises" or "not to cheat," for instance, will
strategy that has proven to be "successful" in one context into a new
typically be considered standard examples of moral rules, while a social
context, is, however, typically tentative and conjectural. It is controlled
convention, such as driving on the right side of the road, will probably
by the opposite tendency to discriminate between those kinds of en-
not be, presumably because the latter does not involve (at least not
vironment in which a certain activity or. strategy is in fact advantageous.
under standard circumstances) a conflict between individual and com-
and those environments where this is not the case. That is, depending
mon interest.⁷
on the relevant characteristics of different social settings, people's moral-
ity may well be restricted to specific environments only.66
5
See. e.g.. Stigler's "Unfriendly Critics." p. 35: "What we call ethics. is a set of rules with
Just as people will tend systematically to discriminate among social
respect to dealings with other persons. rules which in general prohibit behavior which is
settings in which morality is or is not experienced as an efficient be-
only myopically self-serving. or which imposes large costs on others with small gains to
havioral strategy, they can also be expected to respond to changes in
oneself. General observance of these rules makes not only for long-term gains to the actor
but yields outside benefits." If it is intended to be a definition of ethics. Stigler's statement is
their environment which affect the relative payoffs from morality on the
certainly burdened with a number of factual assumptions which can be and. in this essay.
one hand and expedient case-by-case choices on the other. Although
will be separated from purely definitional considerations.
morality, like all rule-governed behavior, is to some extent necessarily
6
In Morals by Agreement (Oxford: Clarendon Press. 1986). David Gauthier defines moral
principles as those "that constrain the actor pursuing his own interest in an impartial way"
unresponsive to environmental variation (perfect responsiveness would.
(p. 3) and "the conception of morality as a set of rational, impartial constraints on the
in fact, be incompatible with the very notion of a rule). morality is a
pursuit of individual interest" (p. 6).
matter of an individual's adaptation to relevant characteristics of his
The argument can be stated more precisely in game-theoretical terms: typically those rules
seem to be classified as moral rules which provide solutions to recurrent Prisoner's Di-
lemma (PD) situations as opposed to recurrent coordination-type situations. to which
65 ibid., p. 103. See ibid., pp. 92f.: "A norm is learned or internalized when the actor has been
conventions like "driving on the right side" provide general solutions. See Viktor Vanberg.
sufficiently conditioned by sanctions that his behavior conforms to the norm at a spatial or
"Spontancous Market Order and Social Rules-A Critical Examination of F.A. Havek's
temporal distance from sanctions. Such learning is. however. never complete: it remains
Theory of Cultural Evolution." Economics and Philosophy vol. 2 (1986). pp. 91ff. The
ultimately dependent on subsequent sanctions. even if the prospect is remote."
potential conflict between common interest and individual interest which characterizes
66 See the aforementioned notion of "honor among thieves."
PD rules. in contrast to the largely self-enforcing character of coordination rules. seems to
30
3
Discussions of ethics or morality as a potential determinant of human
behavior apparently are concerned with how moral behavior can be
explained. That is, why do people behave in accordance with moral
rules-if and to the extent that they actually do so?⁸ Our everyday
interpretations of human behavior, as well as standard social theoretical
accounts, typically refer to some kind of expected sanctions as factors
VIII. Morality:
which explain why a person obeys moral rules. Those sanctions may
range from the more obvious formal sanctions imposed by some special
Individual Adaptation and Collective Choice
enforcing agency, over the sometimes less visible informal sanctions
people spontaneously impose on each other, to the totally intangible
In sociology, a common objective against a rational choice interpreta-
internal sanctions people's consciences impose on them (including the
tion of moral conduct is that "social order" would not be possible if
imagination of sanctions expected from some transcendental entity).
people were to actually calculate the costs and benefits of rule obedience
There seems to be, in principle, no reason why the economic model of
and rule violation in each and every choice situation, because there are
rational choice should not incorporate such expected sanctions as ex-
too many situations in which the benefits from "cheating" would clearly
planatory variables, viewing them as part of the opportunity costs and,
outweigh the costs. Therefore, the standard argument goes, one has to
consequently, among the determinants of people's choices. Yet, de facto,
assume that people typically tend to follow social norms in a habitual,
economic analysis apparently concentrates-to the extent that it pays
noncalculating manner. Such a general inclination towards uncalcu-
attention to the social-institutional context at all-on formal sanctions,
lated rule obedience is typically explained by the assumption that in a
paying less attention to informal enforcement mechanisms and being
process of socialization, individuals internalize the social norms of their
extremely hesitant to refer to intrapersonal sanctions as explanatory
community.
variables. It is particularly the role of such intrapersonal behavioral
What, in sociology, is labeled "internalization of norms" obviously
determinants, however, to which arguments on ethics or morality typ-
corresponds to what, in the context of this essay, is viewed as the process
ically seem to refer.⁹ And the specific issue on which the following analy-
in which individuals learn or acquire the dispositional trait morality.
The foregoing analysis suggests that, if the level of individual "constitu-
tional" choice is taken into account, a rational choice perspective can
be central to our common notion of moral rules. This is not meant to imply. however that
all principles which make actors in PD situations choose the "cooperative" alternative
indeed explain the kind of uncalculated rule obedience at the level of
qualify per se as moral principles See. e.g., Edna Ullmann-Margalit. The Emergence of
particular situational choices which sociologists have stressed as a cru-
Norms (Oxford: Clarendon Press. 1977). pp. 41ff.
cial ingredient of social order.
8 A separate issue which will not be discussed here is why certain kinds of rules-rather than
some conceivable alternative rules-are considered moral rules in particular social groups.
To analyze the phenomenon of "internalization" in terms of an indi-
If, as indicated above, moral rules are viewed as providing solutions to recurrent problems
vidual's "constitutional choice" will certainly appear to most so-
(in particular of the PD type) in social interaction, it seems plausible to assume that moral
ciologists as an inadequate rationalization of what, in fact, may be a
rules will share (through history and across cultures) certain common basic characteristics.
to the extent that the general structure of typically recurrent interaction problems is similar
largely unintentional process. Actually, this should not be a crucial is-
in all social groups. Such an assumption seems to be a central underlying premise of the
sue. Whether the acquisition of the dispositional trait morality is mod-
"individualistic-evolutionist conception of social institutions" that I describe in "Spon-
eled as a process of rational constitutional choice or as a process of
taneous Market Order." the modern versions of which can be traced back to David Hume
and Carl Menger. See Menger. Investigations into the Method of the Social Sciences with
unreflected behavioral conditioning.64 an individual can be expected to
Special Reference to Economics (New York and London: New York University Press.
acquire and to preserve this disposition only if it is experienced as gener-
1985), pp. 228f.: "Certain conditions resulting from general human nature and thus appear-
ating sufficient advantages or rewards. What is crucial for the general
ing everywhere produce similar institutions of law everywhere by their nature. while tribal
differences and variety of external conditions. result in differences in law." For an excel-
research strategy-i.e., for the issue of where one looks for an explana-
lent discussion of Hume's theory of "moral institutions." see Hartmut Kliemt. Moralische
tion-is not whether the process is described in terms of rational choice
Institutionen-Empiristische Theorien ihrer Evolution (Freiburg/Muenchen: Karl Alber.
or of unreflected habit formation. What is crucial is whether or not the
1985).
9
See, e.g., John L. Mackie's Persons and Values (Oxford: Oxford University Press. 1985), p.
relevant costs and benefits to the individual are viewed as the central
159, which considers the "internalization of norms" to be a crucial ingredient of "moral-
explanatory variable. That individual moral acts may be uncalculated
ity": "The association of moral sentiments with the practices. and a sense of guilt about
one's own transgressions. is a major part of such internalization. Only when this stage is
64 For an excellent discussion of the notion of norm internalization in the terms of behavioral
reached can we speak properly of a morality."
psychology. see Scott. Internalization of Norms.
4
29
particular actions in one case, routines in the other-and the kind of
sis will focus concerns the nature and potential explanatory significance
expected consequences that are of relevance in making a choice-spe-
of those intangible, intrapersonal determinants of choice which account
cific outcomes in one case, patterns of outcomes in the other. In con-
for a person's moral conduct, in addition to and separate from actual
trasting his concept of "constrained maximization" with the
threats of formal or informal external sanctions. 10
"traditional" notion of "straightforward maximization" Gauthier, on
Before commencing the discussion of this issue, a point of conceptual
the other hand, seems to suggest that more is involved than just apply-
clarification is in order. Moral behavior and morality, at least as the
ing the same notion of rationality to different levels of choice. It is the
concepts are used here, should be carefully separated. Moral behavior is
additional claims which Gauthier attaches to his concept of "con-
understood here in a purely descriptive sense, as behavior that is in fact
strained maximization" which mark the essential difference between his
in accordance with moral rules-irrespective of what the actor's actual
approach and the one taken here. Measured against the latter, Gauthier's
motivation might be. Morality, on the other hand, is understood as a
approach appears to try to prove too much or, in any case, more than is
dispositional trait which accounts for a person's moral conduct.
needed in order to show that there can be a rational basis for morality,
Whether a person's behavior qualifies as moral behavior in this sense is
understood as a general disposition to act in accordance with moral
properly a matter of observation and classification. Whether a person
rules.
acts out of morality is an explanatory issue. To use, as is sometimes
done, the term "moral behavior" in a sense that makes "morality" one
of its definitional attributes is to blur the distinction between what is to
be explained (moral behavior), and how it is to be explained (by refer-
ence to morality or to other factors). As it is used here, the notion of
morality does not refer to any observable behavior as such, but to some
intrapersonal, dispositional variable which may account for observable
moral behavior.
Authors who argue in favor of incorporating an intrapersonal deter-
minant of behavior like morality into economic explanations generally
justify their approach by pointing out that there is ample evidence that
people respect moral rules even in situations in which-so far as an
observer can discover-the external incentive structure does not seem
sufficiently to motivate such "moral behavior." Examples of such
moral conduct, which seem to reflect the genuine impact of some intra-
personal. moral force. range from more trivial instances like people not
littering when unobserved at a remote beach, over more significant in-
10 To be sure. the distinction between "external" and "internal" sanctions or behavioral
determinants is not totally unambiguous since. strictly speaking. expected sanctions are the
directly relevant variable for a theory of individual choice. And a person's expectations of
"external" sanctions are. in principle. no less intrapersonal and subjective than what is
explicitly classified here as intrapersonal determinants. However, if one rules out. as is
typically done in economics. the possibility that people permanently and systematically
misperceive relevant characteristics of their environment, it can be reasonably assumed
that people's expectation on the whole reflect the "actual" threats of formal and informal
sanctions which the observer-analyst can discover in the respective setting. This does not
require postulating that people never misperceive their environment. It only requires
ruling out that such misperceptions are a relevant source of systematic differences and
changes in people's behavior.
11 Richard B. McKenzie, "The Economic Dimensions of Ethical Behavior." Working Paper
Series (Blacksburg. VA: Center for Study of Public Choice, 1976, Mimeographed). pp. 7f.:
McPherson. "Limits on Self-Seeking." p. 77: Robert H. Frank, "If Homo Economicus
Could Choose His Own Utility Function, Would He Want One with a Conscience?"
(Ithaca, NY: Cornell University. 1985. Mimeographed). p. 3.
28
5
stances like honest behavior towards strangers, to the most spectacular
So far, Gauthier's argument is perfectly analogous to the argument put
instances like people heroically sticking to certain principles despite
forward in this essay. The crucial difference between the two approaches
considerable costs to themselves.
arises from Gauthier's supposition that the rationality of a constitu-
Certainly, for each and every one of those particular examples one
tional moral choice directly carries through to the level of particular,
may suppose that a more careful analysis should be able to disclose
situational choices.61 No such claim is made by the present approach.
some subtle or hidden "external" determinants which would allow for a
All that is claimed here is that it can be rational-provided that certain
satisfactory explanation without any need to bring in a special intraper-
conditions are given-to make a constitutional moral choice, i.e., to
sonal variable like morality.¹² But as long as those "hidden external
adopt a moral routine and renounce the attempt to make an "optimal"
determinants" are not identified and shown to provide a satisfactory
choice in each and every particular choice situation. In arguing that a
explanation, it is legitimate and worthwhile to ask whether and how an
moral routine can be rational, what I argue is that it can be rational not
intrapersonal variable morality can be incorporated consistently into
to calculate in particular choice situations whether or not to follow a
economic theory. Which one of the two alternative strategies-to search
moral rule, but rather routinely to act morally without any such situa-
for "hidden external determinants" or to incorporate morality-will
tional calculation. I definitely do not mean to imply that, once a consti-
turn out to be the theoretically more fruitful one is a matter of fact that
tutional moral choice is made, it must be rational for a person to act
cannot be decided on a priori grounds.
morally in each and every choice situation, even if he should identify a
particular choice situation as one in which-when all consequences are
properly accounted for-the immoral choice appears to be the "max-
imizing" one.
The point is subtle, but crucial. In part, Gauthier also argues as if the
rationality of not calculating at the level of particular choices is what he
considers to be the relevant implication of a rationally adopted moral
disposition. 62 But he clearly seems to make a stronger claim when he
argues that, once a rational choice to adopt a moral disposition is made,
rationality requires a person to comply even if, in her judgment, this
means "acting in a way that results in real disadvantage to herself."63 In
making such a claim, Gauthier seems to give to his concept of rational
moral choice-which he calls "constrained maximization"-an inter-
pretation which not only shifts the notion of rational choice from the
situational to the constitutional level, but also changes the meaning of
the concept of rationality itself.
The distinction between situational and constitutional levels of
choice, as it is used in the approach advocated here, does not require any
presumption that the meaning of "rational choice" is somehow different
in both contexts. Rationality means, in both cases, to choose the alter-
native which is expected to yield the largest payoffs. What is different is
the character of the alternatives among which a choice is to be made-
straightforward maximization, and of adopting constrained maximization, as his disposi-
tion for strategic behaviour."
61 ibid., p. 187: "Our argument identifies practical rationality with utility-maximization at the
12 It should be obvious that the morality issue as defined here is different from the issue of
level of dispositions to choose, and carries through the implications of that identification in
informal versus formal sanctions, though the two are sometimes confused. That people
assessing the rationality of particular choices."
often respect moral rules in the absence of formal enforcement is a trivial and uncontrover-
62 This aspect is stressed when Gauthier argues on p. 167 of Morals by Agreement that a
sial observation. What is controversial is not whether informal sanctions matter (who
"constrained maximizer" is disposed to cooperate "without considering whether some
would want to deny this?), but whether moral behavior occurs in the absence of (sufficient)
individual strategy would yield her greater expected utility."
external sanctions of any kind, formal or informal.
63 ibid., p. 169.
6
27
Gauthier's approach and the one adopted here are much closer in
their central thrust, and the differences between them are much more
subtle. Like the present approach, Gauthier's approach puts its main
emphasis on the distinction between a "constitutional" and a "situa-
tional" choice level. As Gauthier stresses, in contrast to an interpreta-
tion which "identifies rationality with utility-maximization at the level
III. Morality: Preference or Constraint?
of particular choices,"⁵ his own approach identifies "rationality with
utility-maximization at the level of dispositions to choose."⁵ The choice
The personal attribute of morality can be characterized as a disposi-
to be moral is analyzed as "a choice about how to make further
tional variable. It reflects a person's general disposition or propensity to
choices, as "a choice among dispositions to choose." And the "ra-
act in accordance with moral principles, relatively independently of the
tionality of morality" is viewed as a matter of rational dispositions, not
specific incentive structure inherent in particular choice settings.
of rational actions;⁵ the capacity to choose one's dispositions is consid-
Attempts to account for a dispositional variable like morality in eco-
ered "an essential part of human rationality."57
nomic explanations seem to be at odds with the standard economic
According to Gauthier, the crucial reason why a disposition to act
classification of explanatory variables into either preferences or con-
morally can be rational is that it allows a person to get access to cooper-
straints. As normally understood, this classification corresponds to the
ative arrangements, and thus to potential gains from cooperation, from
"subjective vs. objective" distinction: preferences are the subjective, in-
which a person lacking such a disposition would be excluded. 58 Though
trapersonal determinants of choice, and constraints are the objective,
in a particular choice situation violation of a moral rule may be advan-
external determinants.¹³ Given this system of reference, introducing
tageous compared to moral conduct, the disposition to act morally may
morality into economic explanations as a variable would seem to re-
benefit a person more than a disposition which allows for such trans-
quire us either to classify morality as a subjective, intrapersonal prefer-
gressions. 59 What is relevant to Gauthier's "calculus of morality" is the
ence-variable or as an objective, external constraint-variable.
comparison of the overall benefits that can be realized from alternative
Since classifying morality as an external constraint apparently is ruled
dispositions, rather than the comparison of the benefits from particular
out by the fact that morality has been explicitly defined here as an
alternative choices. 60
intrapersonal, dispositional variable to be strictly separated from exter-
nal sanctions, the only alternative seems to be to view morality as a
52 ibid., p. 182.
53 ibid.
preference-variable. This view appears to be implicit in the dichotomy
54 ibid., p. 158.
between morality and self-interest which is quite common in discus-
55 ibid., p. 183.
sions of the relationship between ethics and economics. 14 The propo-
56 ibid., p. 182f.: "A disposition is rational if and only if an actor holding it can expect his
choices to yield no less utility than the choices he would make were he to hold any
nents of such a dichotomy seem to suppose that some modification of
alternative disposition."
the standard economic model of self-interested behavior is needed in
57 ibid., p. 183: "But there is a further significance in our appeal to choice among dispositions
order to account for morality, though they are generally not very explicit
to choose. For we suppose that the capacity to make such choices is itself an essential part
of human rationality."
about the specific kind of modification they have in mind. There seem
58 ibid., p. 162: "The disposition to keep one's agreement. makes one an eligible partner in
to exist two views which, though they are not totally different, can be
beneficial co-operation, and so is itself beneficial." See ibid., p. 183: "The essential point in
separated. The one view conceptualizes morality as a motivating force
our argument is that one's disposition to choose affects the situations in which one may
expect to find onself."
which is beyond and outside of any interest calculation, or, in other
59 ibid., p. 162: "A person disposed to violate his covenants. cannot rationally expect to reap
terms, as a factor that is somehow outside a person's preference function
the benefits available to co-operators. Even if his particular breaches of covenant would
benefit him, yet his disposition that leads him to such breaches does not."
13 As I mentioned earlier, the distinction between intrapersonal morality and external sanc-
60
With respect to "constrained maximizers," as he labels morally disposed persons, Gauthier
tions is relevant and meaningful, even if one acknowledges that an individual's (necessarily
states on p. 15 of Morals by Agreement: "Of course, constrained maximizers sometimes
subjective) expectation of sanctions is the direct determinant of his choice.
lose by being disposed to compliance.
Nevertheless. we shall show that under plausible
14 The dichotomy between morality and self-interest is reflected in statements such as that
conditions, the net advantage that constrained maximizers reap from co-operation exceeds
found in Stigler's "Friendly Economists," pp. 25f.: In "situations where self-interest and
the exploitative benefits that others may expect. From this we conclude that it is rational to
ethical values.
are in conflict.
most of the time.
the self-interest theory.
will
win."
be disposed to. internalizing moral principles." See ibid., p. 170: "To demonstrate the
And this dichotomy is equally implied when McPherson argues in "Limits on Self-Seek-
rationality of suitably constrained maximization we solve a problem of rational choice. We
ing," p. 72, that "people's willingness to be moral" raises difficult questions "for economists
consider what a rational individual would choose, given the alternatives of adopting
who are professionally predisposed to postulate a self-interested basis for action."
26
7
and is not captured by the economist's notion of rational choice. 15 The
other view conceptualizes morality as part of a person's preference func-
tion; i.e., it considers human interests to be broader than "self-interest,"
as this notion is commonly understood, and to include something like
an interest in being a moral person.¹⁶
The first view is problematic even at the conceptual level. If the prin-
VII. Alternative Approaches: Supergames and
ciple of rational, self-interested behavior is meant to imply that individ-
uals choose from among potential alternatives the one which they
Gauthier's Rational Morality
expect to serve their interests best, it is difficult to see how morality can
be conceptualized as a factor that makes individuals choose something
It may be useful to pause here for a look at two approaches to a
different from what they think is in their best interest. One would have
rational explanation of morality which are similar but not identical to
to assume, it appears, that man has "two natures," an interest-oriented
the perspective outlined above.
one and a moral one, and that sometimes the one and sometimes the
The one approach is characteristic of all those conceptions which, in
other determines his behavioral choices. Without a more general theory
one way or another, argue along the lines of the game-theoretical notion
which would integrate the two at some higher level and specify the
of repeated Prisoner's Dilemma games, in particular the notion of a
conditions under which the one or the other of the "two natures" will
cooperative solution to infinite-horizon supergames.⁵⁰ The other ap-
prevail, explanations of observed behavior in terms of such a model
proach is the one elaborated by David Gauthier in his attempt to show
would be totally arbitrary. If, on the other hand, such a higher order
"why an individual, reasoning from non-moral premises, would accept
theory is specified, the relevant issue would be whether or not this the-
the constraints of morality on his choices."⁵
ory is preferable to the self-interest model in terms of theoretical fruit-
The repeated PD game approach and the present approach have in
fulness.¹⁷ Until now, it seems, nobody has advanced a potential
common that they both stress the relevance of the "shadow of the fu-
candidate for such a theory which could actually be compared with the
ture" in people's choices. The central argument is that in settings where
model of rational, self-interested choice.
people are likely to meet again, they have good reasons to consider the
By contrast with the first view, the second interpretation of the di-
more indirect and long-term consequences of their choices, and not just
chotomy between morality and self-interest does not question the basic
the immediate and short-term effects. Properly taking into account
logic of the assumption that individuals will choose what they consider
those indirect and long-term consequences, however, will let many
to be in their best interest. According to this interpretation, the problem
moral choices appear as rational which otherwise may seem to be "self-
with the self-interest model is that the concept of self-interest as it is
sacrificing."
commonly understood suggests too narrow an interpretation of human
The iterated PD game approach is different from the approach advo-
interests. The idea apparently is that a more appropriate model of
cated here in that it does not draw the distinction between the "constitu-
tional" and the "situational" level of choice which is so central to the
15 This view is reflected, for example, in McKenzie's "Economic Dimensions." p. 9: "Explo-
ration of the Christian or Kantian motives for individual action. which makes the 'good-
present argument. The iterated game argument remains at the level of
ness' of the act itself an end distinct from the gain to the actor. may yield useful predictions
particular, situational choices. It aims to show that in particular choice
about behavior which can complement the economist's insights founded on 'economic'
situations in which-if they are analyzed as one-shot games-the defec-
rationality."
16 This view is expressed, for example, in Terry L. Anderson and Peter J. Hill. "Constitutions
tive choice would be dominant, the cooperative choice is the rational
and Ideology: Substitutes or Complements?" (Bozeman. MT: Montana State University.
one if due consideration is paid to long-term future consequences. In
1985, Mimeographed). pp. 7ff. The authors propose a modification of the standard utility
terms of the distinction explained above, it can be said that the iterated
function in order to incorporate an "ethical component." The modified approach explains
individual choices in terms of a "behavioral objective function" which includes a normal
PD game approach is about moral behavior, not about morality. It
preference function along with an "ethical function." allowing for substitutions between
explains why it may be rational to act morally, but it does not explain
"welfare" and "ethics."
why it may be rational to adopt a moral routine.
17 In this sense, what is actually at stake in controversies about the status of the "model of self-
interested behavior" is not whether we allow for more or less "self-interest" in our explana-
tions. but whether we preserve a unitary behavioral principle or allow for dual or triple
50 See. e.g., James W. Friedman, Oligopoly and the Theory of Games (Amsterdam: North
human "natures." Suggestions for changes in the self-interest model should be stated in
Holland, 1977): Andrew Schotter, The Economic Theory of Social Institutions (Cambridge:
terms of alternative unitary models of behavior which can be compared in their explana-
Cambridge University Press. 1981).
tory power to the self-interest model.
51 Gauthier. Morals by Agreement, p. 5.
8
25
high.49 And, typically, a person will have the more reason to care about
human behavior ought to allow for other interests-such as moral pref-
his reputation the more stable and the less anonymous the relevant
erences-which people pursue along with their more narrow self-inter-
social environment, and the more his own orientation in this environ-
ests. That is, self-interest is viewed as a specific subcategory of all inter-
ment is of a long-term character.
ests which an individual might pursue, a subcategory which exists along
with other, presumably more "noble" interests.
The interpretation of morality as an element in an extended prefer-
ence function cannot be accused of the conceptual inconsistencies
which characterize the first view. But there are other reasons why econo-
mists would typically be reluctant to embrace such an interpretation,
reasons that make them generally prefer explanations in terms of con-
straints over those in terms of differences and changes in preferences. To
explain intra- and interpersonal variations in moral behavior in terms of
variable moral preferences is, it would typically be argued, not a more
satisfactory account than explaining variations in people's fitness ac-
tivities in terms of their different preferences for exercising. Any expla-
nation of particular behavioral patterns in terms of specific prefer-
ences-whether moral or other-is, according to this view, redundant
or ad hoc so long as no evidence for those preferences can be provided
which is independent of the behavior that is to be explained. And the
supposition is that once one allows for such intrapersonal factors as
explanatory variables, there will be no limit to inventing special prefer-
ences or motives whenever observed patterns of behavior resist standard
explanations.
The attempt to incorporate "morality" into economic theory seems
thus to lead into a dilemma. One seems to be forced into the choice
either to deny that something like morality as an internal, dispositional
determinant of behavior exists, or to set aside the methodological objec-
tions against explanations in terms of preferences. Before choosing one
of the two horns of this dilemma, it may be useful to examine whether
there might not be some other way of incorporating morality into eco-
nomic theory, a way which is more consistent with the standard meth-
odological orientation of economics. The most challenging approach to
examine for this purpose is probably Gary S. Becker's reformulation of
the theory of consumer choice, in which the economists' argument for
relying as much as possible on constraints and as little as possible on
preferences for explanations of differences and changes in behavior has
been carried to its logical conclusion. Becker's approach is based on the
premise that no interpersonal differences or intrapersonal changes in
49 To be sure, the potential benefits from "moral reputation" may invite mimicry and at-
preferences exist and, therefore, differences and changes in behavior are
tempts to pretend morality and to realize gains from moral reputation without actually
bearing the costs of genuine morality. The advantages from genuine rule-following be-
to be explained exclusively in terms of (income and price) constraints. If
havior, however, by their very nature, cannot easily be realized by mimicry. To be un-
an intrapersonal, dispositional variable morality can be shown to find a
masked as a hypocrite is the most damaging thing that can happen to one's moral
consistent place within a Beckerian framework, this should prove most
reputation, while on the other hand, it is most beneficial to one's reputation if one is
observed acting morally in situations where one could not reasonably expect to be
effectively that such a variable can be systematically incorporated into
observed.
economic explanations.
24
9
The remainder of this essay will discuss how a Beckerian perspective
understood as a general label for all inalienable sources of a person's
may indeed be used to arrive at an interpretation of morality which can
productivity. And deliberately forgoing potential gains from rule viola-
be more fruitfully incorporated into economic theory than alternative
tion may be considered an investment in the human capital of a good
accounts. Following a discussion of Becker's approach (Section IV), sec-
reputation. Whether or not the reputation of being a moral person
tions V and VI will offer an interpretation of morality which provides an
actually is an asset, and whether or not it is worthwhile to invest in it,
alternative to the dichotomy between morality and self-interest. Instead
obviously depends on the character of the relevant environment and on
of viewing morality as a motivational factor, outside or inside a person's
one's own position in it. Like all investments, investments in reputation
preference function, the conception to be advanced here will discuss the
mean current sacrifices for expected future benefits. Such sacrifices can
dispositional variable morality in the context of a person's production
be rational only to the extent that there is an "expected future." And,
function.
where there is a future, the expected payoffs from such investments may
vary significantly for persons in different positions. For instance, for a
person with an already established "negative reputation" the potential
returns from such investments may be quite small compared, e.g., to
those of a newcomer who still has to establish a reputation, or to those of
a person with an established positive reputation.
Depending on his experiences in certain environments, a person may
have more or less reason to expect that "honesty pays." It should be
noted that for moral learning, as for human learning in general,
vicarious learning-learning from indirect (observed, reported, etc.) ex-
perience-may play a significant role in shaping people's moral at-
titudes. And one has to take into account the fact that theories about
the payoffs from investment in reputation are typically about the long
run and therefore not as readily testable as theories about short-run
gains and losses from moral behavior. People may well learn sys-
tematically to distinguish between environments where they consider an
"honesty pays" theory appropriate and other environments where con-
siderations of reputation are considered irrelevant, a phenomenon to
which notions like "honor among thieves" refer. The relevant point in
the present context is that the more reason a person has to care about his
reputation, the more rational it may be for him to adopt the rule always
to act morally. because it may simply be too costly to try to identify
those situations in which rule violation would not be detrimental and/or
because the risk of misidentifying such a situation may appear to be too
48 See John Finley Scott, Internalization of Norms-A Sociological Theory of Moral Com-
mitment (Englewood Cliffs. NJ: Prentice-Hall. 1971), pp. 108f.: "What amounts to an
implicit belief that sanctions will follow deviant action has to be reinforced by sufficiently
persuasive and present evidence that they do in fact follow. So rewarding does a social
animal find a cognitive orientation toward his fellows that attention to what happens to
these
others. is probably impossible to suppress completely.
If he observes that others
are punished for deviation. his own learned commitment is thereby reinforced: but if he
sees that others are not punished for deviation. or are rewarded for it, his moral commit-
ment is weakened.
The curiosity of the moral about the immoral is notorious. The
degree
of moral behavior. is a function not only of the sanctions applied to a single
person. but also of the sanctions applied to the group of persons who may learn vicariously
from each other. Sanction one and you influence all: the good man is kept good not only
because of his own rewards. but also because he sees the punishment dealt to the evil."
10
23
why morality may be an efficient behavioral routine, since formal en-
forcement nowhere is-and, for reasons of cost, never can be-tight
enough as not to permit a broad range of situations where rule viola-
tions would remain the more advantageous alternative. Taking informal
enforcement into account will allow some narrowing of the range of
such "opportunities to cheat," but it is common opinion that there will
IV. Morality and Choice:
always remain a more or less significant set of situations in which formal
and informal external sanctions alone typically cannot be assumed
Theories as Subjective Constraints
effectively to eliminate the temptation "to cheat." The question of how
moral behavior in this kind of situation can be explained, and in par-
Becker's proposal for a reformulation of the theory of choice is moti-
ticular whether recourse to a dispositional variable morality can be an
vated by precisely the problem discussed above: that explanations of
appropriate and fruitful explanatory approach, is exactly the issue
differences or changes in people's behavior in terms of preferences are
raised at the beginning of this essay.
bound to be arbitrary since, as Becker supposes, "no useful theory of the
Can adopting a rule "always to act morally" be rational for an actor,
formation of tastes" exists in the social sciences on which one could rely
even if he expects that there will be situations in which nonmoral be-
in "choosing the appropriate taste proxies.
.or in formulating predic-
havior would be more advantageous? It is rational, if he perceives the
tions about the effects of these variables on behavior."¹⁸ The remedy
costs of identifying the relevant subset of situations or the risk of mis-
Becker suggests is to "treat tastes as stable over time and similar among
identifying a situation as high enough. Whether this is the case or not
people"19 and to explain differences or changes in observed behavior
will crucially depend on the actual structure of the social environment
exclusively in terms of differences in price and income constraints, con-
in which a person is acting, on his own "position" within this environ-
sidering that the relevant prices and incomes may take quite subtle
ment, and on his relevant theories about "the world." If an actor's rele-
forms and include more components than money prices and money
vant theories imply, for instance, that the essential remuneration for
income. What allows Becker to shift all of the explanatory burden to
moral behavior is transcendental, then to him morality can be a rational
price and income constraints is essentially a narrower definition of pref-
strategy whatever the actual structure of his social environment, al-
erences and the reinterpretation of the household (consumer) as a pro-
though the probability that an individual will adopt and maintain such
ducing unit. The "preferences" which Becker assumes to be "stable and
a theory will not itself be independent of his social environment. The-
identical" for all individuals are not specific tastes for particular goods,
oretically more challenging, however, is the case in which the theories on
as is suggested by the common use of the term. Rather, people just as
which the rationality of morality depends are at least in principle testa-
human beings are assumed to share certain fundamental preferences or
ble. In this case, if one rules out systematic and permanent mispercep-
"basic pleasures." The consumer is viewed as producing "com-
tions as a relevant source for differences in people's theories, such dif-
modities," i.e., ultimate goods (such as health, status, etc.) from which
ferences would have to be explained in terms of differences in the en-
he directly obtains utility. In producing these "final objects of choice,"
vironments which condition people's theories.
the consumer-producer is using as inputs his own time and resources,
Apart from immediate disadvantages resulting from formal or infor-
along with goods and services purchased in the market. An essential
mal sanctions, a main interest that may motivate an actor to act morally
implication of this conception is that the demand for market goods is to
is a concern about his reputation. To have the reputation of being a
be modeled as derived demand, comparable to a firm's demand for
moral person, i.e., a "man of principle." may considerably increase an
factors of production.²²
individual's productive efficiency, i.e., his ability to produce the ultimate
The consumer's (household's) utility (U) is a function of the com-
objects of choice. The reputation of being a moral person can be viewed
modities (Z), the "ultimate objects of choice":
in this sense as a kind of human capital, where "human capital" is
18 Becker, Economic Approach. p. 133.
p. 9: "As rational persons understanding the structure of their interaction. they recognize a
19 Stigler and Becker. "De Gustibus," p. 76.
place for mutual constraint. and so for a moral dimension in their affairs.
Agreed
20 Becker. Economic Approach. p. 145.
mutual constraint is the rational response to these structures." Gauthier's approach to a
21 ibid., p. 146.
"rational explanation of morality." which is similar but not identical to the approach taken
22 ibid., p. 134.
here, will be examined in more detail later in this essay:
22
11
= Zₙ).
expect a pattern of rule-constrained choices to serve their long-term
interests better than case-by-case choices made under "temptations" of
Since, by assumption, interpersonal differences or intrapersonal
the immediate choice situation.
changes in the utility function are ruled out, systematic differences in
That, for the reasons mentioned, rule-governed behavior may be pref-
people's behavior which cannot be accounted for in terms of con-
erable to a strategy of case-by-case "optimization" does not imply, of
ventional income and price constraints are to be explained in terms of
course, that just any rule may be adopted. There are many conceivable
differences or changes in people's production methods or, in Becker's
behavioral rules which obviously would produce a pattern of outcomes
terms, in their productive efficiency with regard to the Z-commodities.23
clearly inferior to what could be realized by deciding each case on its
The consumer's (household's) production function for the commodity
own. The preceding arguments, therefore, do not by themselves imply
Zᵢ is:
that it must be rational for an individual to adopt moral rules. An
individual's direct "moral interest" is not in his own obedience to rules
but, instead, in the obedience of others. His own obedience directly
where Xi refers to market inputs, tᵢ to the consumer's own time input,
benefits others, and an individual's own immediate interests would be
and E to "environmental variables" which, in particular, are supposed to
best served if he were allowed to ignore rules while others were bound by
reflect the production technology employed by the consumer-producer.
them. Therefore, to explain why morality can be an efficient behavioral
In the present context, the question about Becker's reformulation of
technology requires more than simply showing why adopting behavioral
the theory of (consumer) choice that is of interest is whether and, if so,
rules may be rational. What has to be shown is why it can be rational for
how a variable morality can be incorporated into a Beckerian frame-
an individual to adopt a rule which appears to imply self-sacrificial
work. 24 If morality is considered to be a variable which may account for
demands.
systematic differences in people's behavior, it obviously cannot be a
It seems obvious that morality can be considered an efficient be-
preference in the Beckerian sense. And since, for the reasons I men-
tioned earlier, it cannot be interpreted in terms of conventional income
havioral technology by an actor only if he perceives the relevant typical
choice situation not to have Prisoner's Dilemma characteristics. If an
and price constraints, the only category under which it can be sub-
sumed, if it can at all, is the environmental variable E. It is therefore the
actor does assume that in typical choice situations respect for moral
specification of the E-variable which is crucial for our present purposes.
rules is not in his best interest, it simply cannot be rational for him to
Becker's explanations of E (or its equivalent in the slightly different
adopt the rule always to act morally. That is, if morality is to be a
version of the household production function in the work of Stigler and
rational strategy, the actor must perceive the relevant typical choice
Becker) are not very extensive and systematic. It is obvious, however,
situation to be different from the setting that has been described above.
that although they are called "environmental," the E-variables are not
Our analytical attention has to be focused, therefore, on the reasons why
strictly external variables which the observer-analyst could discover in
an actor may have such a different perception.
the "environment in which the production takes place.' Actually, the
The obvious answer is that in ordered social communities. we will
notion that the environmental variables reflect "the level of technology
typically find some kind of enforcement mechanisms which are de-
of the production process"26 is ambiguous in this respect. In con-
signed to "correct" the expected payoffs from rule violation and rule
ventional production theory for the firm, technology is typically viewed
obedience so as to make the latter the rational choice. After all, the
as an external variable which is given and is the same for all relevant
standard argument for how a group of individuals may escape recurrent
firms. Technology, therefore, cannot account for systematic differences
social dilemma situations that take the form of a Prisoner's Dilemma is
in the behavior of firms. Such a view is justified since technology is, in
that they may agree jointly to submit to constraints that eliminate the
temptation to "cheat."+7 But this certainly is not sufficient to explain
23 ibid., p. 145: "One might argue that indeed all households have precisely the same utility
function. .that they all derive that utility from the same 'basic pleasures' or preference
function. and differ only in their ability to produce these 'pleasures."
See. e.g., J. Elster. Ulysses and the Sirens (Cambridge: Cambridge University Press. 1979):
24 The relevance of Becker's approach is also discussed. though with different conclusions
Thomas C. Schelling. Choice and Consequence (Cambridge and London: Harvard Univer-
from those which are drawn here. in Alexander Rosenberg. "Prospects for the Elimination
sity Press. 1984).
of Tastes from Economics and Ethics." Social Philosophy & Policy. vol. 2 (1985). pp. 48-68.
46 As James Buchanan argues in "Ethical Rules." the "moral dilemma" referred to here can
25 Becker. Economic Approach. p. 135.
be more precisely discussed as a "large-number dilemma."
26 ibid.
47 With reference to those kinds of social situations. Gauthier argues in Morals by Agreement,
12
21
different rule), the relevant comparison is, of course, not between spe-
fact, "out there" as "objective knowledge," in the sense of Karl Popper's
cific single outcomes but between patterns of outcomes over a sequence
theory of a "third world" of objective ideas.²⁷ It can be assumed that
of choices. That is, it is the overall performance of alternative strategies
firms will systematically tend to adopt efficient technologies once they
which counts, where "to decide each case on its own merits" can be
are "out there," if they are competing in markets in which the user of
considered one strategy. The potential benefits from adopting a rule for
less efficient technologies would not be able to survive.
making certain kinds of decisions can be seen to result from avoiding
Whatever one might assume about competition among consumers
the disadvantages of case-by-case decisions, for which there seem to
(households). they certainly do not compete as "producing units" in the
exist at least three potential sources:4¹ first, the decision-making costs
same sense that firms do. They produce for themselves, not for a market.
such as the costs of collecting and evaluating the information needed in
And different technologies can be expected to have a much greater
order to decide each case; second, the risk of mistakes resulting, for
chance to survive in such a context. In fact, there seems to be little
instance, from inadequate information; third, the tendency of case-to-
reason to expect a simple and uniform connection to exist between the
case decisions to be biased in favor of short-term interests at the expense
technologies "out there" and the technologies actually adopted by dif-
of long-term interests.
ferent consumer-producers. Rather, it seems to be perfectly reasonable
There exist a number of theoretical contributions on the "reason of
to assume that differences in adopted technologies may indeed account
rules"⁴² issue which stress the one or the other of these sources. The first
for systematic differences in consumer-producer behavior.
source figures prominently in A. Gehlen's theory of institutions as
Quite apart from this, the variable "production technology" has to be
cultural substitutes for instincts, which allow for routines and relieve
interpreted in a way which allows for systematic differences between
human actors from the burden of being constantly required to make
consumer-producers if it is to serve its explanatory purposes in Becker's
decisions. The second source is stressed in R. Heiner's pertinent and
approach. And Becker, in fact. specifies that the technology variable
informative essay on the "Origin of Predictable Behavior." Heiner re-
includes such things as consumer knowledge ("whether real or fan-
views a broad selection of relevant theoretical sources which all sup-
cied"), skills, and human capital in general, 28 things which obviously
port-in one way or another-the general notion that "rule-governed"
vary from person to person. Clearly, however, with such variables as
behavior, such as instincts, habits, customs, and norms, can be viewed
knowledge and human capital construed as constraints, the preferences
as an adaptive response to environmental complexities and uncertain-
VS. constraints dichotomy can no longer be assumed to correspond to
ties which make it difficult for an individual in recurrent choice situa-
the intrapersonal, subjective VS. external, objective distinction. Knowl-
tions to select the most preferred alternative: such complexities and
edge and human capital are genuinely intrapersonal, dispositional varia-
uncertainties make rule-governed behavior in recurrent situations an
bles, and in this respect, they are more like preferences than like
overall useful strategy. To adopt behavioral rules means to adopt "pat-
constraints in the conventional sense.
terns of behavior for which deviations exist that are preferred under
Becker's reformulated theory of choice does not substitute an "objec-
certain conditions, but which are nevertheless ignored because of uncer-
tive" constraint variable for a "subjective" (preference) variable. Rather
tainty in reliably interpreting information about when to deviate."⁴⁴
it shifts the burden of explanation from one kind of intrapersonal, dis-
The third source, finally, is the central theme in the so-called "theory of
positional variable to some other kind. Whether this is only a matter of
self-management, or self-control" which emphasizes that individuals
semantics and camouflage or an actual analytical improvement²⁹ ob-
may rationally choose to adopt a rule constraining their choices if they
viously depends on whether it can be assumed that the relevant disposi-
tional variable in the production function context is not-or, at least, is
41 It should be kept in mind that what I discuss here are the potential advantages to an actor
less-subject to the deficiencies that make explanations in terms of
of adopting a rule for himself. that is, advantages he realizes by his own rule obedience.
This is to be distinguished from the totally different issue of what are the potential advan-
preferences so unattractive to economists. In fact, Becker apparently
tages to an individual of other people's rule obedience. that is. the advantages from living in
assumes that for production function variables like knowledge and
a social community where certain rules are practiced.
human capital, we are in a much better position to identify appropriate
Geoffrey Brennan and James M. Buchanan, The Reason of Rules: Constitutional Political
Economy (Cambridge: Cambridge University Press. 1985).
43 Arnold Gehlen. Studien =ur Anthropologie und Soziologie (Neuwied a.R. und Berlin:
27 Karl R. Popper. Objective Knowledge-An Evolutionary Approach (Oxford: Clarendon
Luchterhand. 1963).
Press. 1972).
44 Ronald A. Heiner, "The Origin of Predictable Behavior," The American Economic Review.
28 Becker. Economic Approach, p. 145: Stigler and Becker. "De Gustibus," pp. 77ff.. 82, 84.
vol. 73 (1983). p.
29 Becker. Economic Approach, p. 144.
20
13
proxies and to formulate predictions about their effects on behavior
than we are in the case of preferences.³⁰
The human capital component in the consumer production function
can be said to reflect a person's inalienable resources which, to the extent
that they are not genetically inherited, have been acquired or learned
over the person's lifetime. If one ignores the possibility of relevant in-
VI. The Calculus of Morality
born differences, systematic differences in people's human capital en-
dowments must reflect systematic differences in their "history," pre-
Can it be rational to be always moral if it is not always rational to be
vious experiences, investment decisions, and so forth. By their very
moral? This, stated in a seemingly paradoxical way, is the issue to be
nature, human capital variables are, as a rule, not directly observable.
discussed in this section.
One typically has to rely on indirect evidence, either in the form of
An analysis of morality as it is understood here (i.e., as a general
observable indicators or in the form of information on relevant aspects
disposition to act in accordance with moral rules) has to draw a careful
of a person's history (e.g., on a person's education), together with some
distinction between a person's situational or particular choice (whether
general assumptions about how various aspects of a person's history
deliberate or not) in a particular situation to respect or not to respect
translate into his current human capital.
moral rules, and the "constitutional" choice to act morally in general.³⁸
It should be noted that, once one is willing to allow for "subjective
It is trivial to suspect that in particular situations it may be expedient for
constraints" and to rely on dispositional variables for which only indi-
an individual to act in accordance with moral rules. And it is equally
rect evidence can be obtained, the economist's standard methodological
trivial to suspect that, considering each situation individually, con-
objections against explanations in terms of preferences are no longer
formity with moral rules will typically not always be rational in terms of
conclusive. Methodologically, the problems faced by an explanation in
expected payoffs. 39 The interesting issue is whether it can be rational for
terms of subjective constraints (a person's theories) are not different, in
an individual to choose (deliberately or habitually) always to obey moral
principle, from those faced by an explanation in terms of preferences.
rules, knowing that there will be situations in which his payoffs would be
What are needed in either case are additional theoretical assumptions
greater from nonmoral rather than from moral behavior. The following
which specify independent evidence for supposed differences and
discussion of this issue will proceed in two steps. It will, first, address the
changes in the intrapersonal variable. To say that people demand certain
general issue of why it can be rational for an individual to adopt rules
market goods because they perceive them as being instrumental in ob-
rather than to decide each case individually, trying to choose the "op-
taining ultimate Z-goods focuses attention on people's theories about
timal" alternative. And it will, second, specify the reasons why this may
the relevant characteristics of goods and on their theory-learning in
also be true for the particular category of rules that is of concern in the
explaining those variations in revealed demand which cannot be readily
present context: moral rules.⁴⁰
explained in terms of conventional income and price constraints. It
If the benefits from following a rule are to be compared to the benefits
would seem to be theoretically equivalent to this to argue that people's
from deciding each case individually (or to the benefits from adopting a
preferences for market goods are derived preferences, derived from their
invariable, ultimate preferences. These derived preferences can be as-
38 The notion of an individual's "constitutional" choice is used here in analogy to the familiar
sumed to vary, depending on what individuals learn-by direct or indi-
notion of constitutional collective choice. In the latter context the term "constitutional"
rect experience-about the relation or association between market
refers to the rules for making collective decisions. In the same sense, an individual's
goods and Z-goods. Whether stated in terms of theory-learning or in
constitutional choice is his choice of rules for making certain kinds of individual decisions.
39 This may be different for a person who firmly believes in transcendental remuneration for
terms of preference-learning, the substantive theory that seems to be
moral behavior.
needed should not differ very much between the two approaches.
As I will discuss in more detail later. the issue addressed here is different from the one to
Though Becker and others assert that there is no useful theory of the
which game-theoretical arguments on iterated PD games are supposed to give an answer.
Those arguments show why in recurrent game settings the "shadow of the future" may
formation of preferences,³¹ there exists, at least, a behaviorist learning
make it rational to choose the cooperative strategy in a particular move. The issue dis-
theory developed by B.F. Skinner and others; by referring, like the pre-
cussed in this essay is why it can be rational for an actor to adopt a moral or cooperative
routine instead of calculating the expected benefits and costs in each and every choice
situation. Another way of stating the difference is to say that the repetitive-game arguments
30 ibid.; Stigler and Becker, "De Gustibus," p. 89.
are about situational choices. while the argument to be elaborated here is about constitu-
31 Becker, Economic Approach, p. 133.
tional choices.
14
19
what are "efficient technologies" to reflect their direct and indirect expe-
ceding argument, to a person's learning history, it provides some sys-
riences with potential alternative technologies.³⁶ And their theories can
tematic information on the formation of "secondary reinforcers," the
be assumed to be the more homogeneous the more similar their past
behaviorist's term for the kind of phenomena which in economics are
experiences and vice versa.³⁷ In particular, with respect to the issue of
traditionally called "preferences."³² Certainly, one may consider this
morality, it should be noted that one relevant dimension along which
kind of psychological theory to be unsatisfactory for the explanatory
people's theories of "efficient technologies" may vary is their testability.
purposes of economics. But it might not be so easy to demonstrate that
There are certainly technology theories which are more readily testable
these theories are in principle inferior to the "implicit learning theory"
than others, and therefore prone to be eliminated if they are incorrect.
on which Stigler and Becker necessarily rely for their own assumptions
For instance, theories about short-term payoffs from morality will be
about how a person's previous consumption of "music appreciation" or
typically more easily testable than those about ultimate, long-term
heroin translate into his current "productive efficiency."3 On the other
payoffs, and theories about what will produce more material wealth are
hand, adopting insights from behavioral psychology may provide eco-
certainly more readily testable than theories about what produces
nomics with a new grip on some of its notoriously puzzling theoretical
greater happiness. And it can be assumed that the less testable the rele-
issues, a new grip which is-as some of the following arguments are
vant theories are, the easier it will be for interpersonal differences to
supposed to indicate-perfectly in line with the core thrust of the eco-
persist (which is not the same as saying that less testable theories as such
nomic paradigm.³⁴
are more likely to persist than more readily testable ones). Theories
Where human behavior is to be explained, we cannot avoid coping
about whether morality "pays" may even be inherently immune to falsi-
with subjective determinants of choice. Simply to rule out, by defini-
fication-for instance, if the relevant remuneration for moral conduct is
tion, the possibility that preferences are variable does not eliminate the
expected from some transcendental entity.
subjective dimension. Its only consequence may be that it requires us-
The interpretation of morality as a theory component in a Beckerian
as is obviously the case with Becker's approach-to define the con-
consumer production function explains the choice of the title for this
straint-variables in a way that makes them include genuinely subjective,
essay. Whether one understands the Latin idiom "De gustibus non est
intrapersonal variables like people's theories about the world. There is
disputandum" as saying that differences in preferences cannot be ra-
no a priori reason, however, to assume that explanations in terms of
tionally disputed, or, in the Stigler-Becker interpretation, as implying
such subjective constraint-variables are somehow methodologically su-
that such differences do not exist, with morality things are different in
perior to those in terms of preferences. Whether the one or the other is
both respects. De moribus est disputandum: People obviously differ in
preferable is a matter of fact, not of principle, and it is easily conceivable
the extent to which they exhibit the behavioral disposition called "mo-
that for some purposes the one, and for other purposes the other, ap-
rality," and these differences can be rationally analyzed as reflecting
proach may prove to be more fruitful.
systematic differences in their (explicit or implicit) theories about what
However, although I have criticized the economist's standard view
are efficient ways to produce the ultimate objects of their preferences.
that explanations in terms of constraints are in principle superior to
those in terms of preferences. the following discussion will proceed
upon the presumption that, with respect to the issue of morality, an
interpretation in terms of subjective constraints-or in terms of people's
32 Behavioral psychology makes a worthwhile distinction that is quite compatible with
36 Taking into account the impact of people's past experiences on their present theories has
Becker's notion of stable. fundamental preferences: the distinction between primary rein-
implications for the concept of rational choice: people's choices can be assumed to be
forcers which are essentially the same for all human beings because of common genetic
rational, given their theories. People's theories can be assumed to be "rational" in the sense
characteristics, and secondary reinforcers which vary between persons and change over a
that they reflect the character of the environment in which they have been formed. This
person's lifetime because of differences and changes in learning experiences.
allows. however, for people's theories to be different and not equally adequate within the
33 Stigler and Becker. "De Gustibus." pp. 79ff.
same kind of current choice situation.
34 In "Rational Egoism versus Adaptive Egoism as Fundamental Postulate for a Descriptive
37 Conventional applications of economic analysis typically focus on sufficiently "homoge-
Theory of Human Behavior. his presidential address to the 1986 Public Choice Society
neous" populations or define the issue in a sufficiently limited way (e.g., "what happens if
Meeting. Dennis C. Mueller pleaded for modeling "economic man" more along the lines
the 'price' of activity A increases?") that relevant cultural or other differences in people's
suggested by behavioral psychology. In Die zwei Soziologien-Individualismus und Kollek-
theories can be ruled out. This is an essential condition for the relatively satisfactory
tivismus in der Sozialtheorie (Tuebingen: Siebeck (Mohr). 1975). I have discussed in some
explanations these standard applications can provide by relying exclusively on "external"
detail the theoretical affinities between a behaviorist and an economic. rational choice
constraints.
approach in social science.
18
15
theories-does, in fact, provide a more fruitful theoretical perspective
than one in terms of preferences. Asking what affects people's explicit
and implicit theories about the payoffs from morality seems to lead into
a more promising avenue of inquiry than asking what shapes people's
preferences for being moral.
V. De Moribus Est Disputandum
1
Morality has been defined above as a general disposition or tendency
of a person to act in accordance with moral rules, relatively indepen-
A
dently of the particular circumstances the person faces in various choice
situations. Morality is considered to be a dispositional variable which
can account for differences or changes in observable behavior which
appear to be unexplainable in terms of conventional "external" con-
straints. As I discussed above, what is needed in order for such a disposi-
tional, intrapersonal variable to allow for more than merely arbitrary ad
hoc explanations is some kind of connecting theory which establishes a
systematic link between assumed states of the dispositional variable and
some kind of evidence which is independent of the behavioral phe-
nomena under explanation.
To argue that morality is a determinant of behavior that is distinct
from external constraints is not the same as assuming that morality is
independent of any external influence. A person's future morality will be
affected (strengthened or weakened) by current constraints the person
experiences, just as his current morality has been shaped by previous
experiences. A person's morality can be said to reflect some gener-
alized-explicit or implicit-expectation or theory about the con-
sequences which may result from rule obedience or rule violation. 35 In
the consumer production function context these expectations can be
viewed as "technology" components or, more specifically, as a person's
knowledge or theories-whether explicit or implicit-about what kind
of behavioral patterns are conducive to producing the ultimate objects
of choice.
That the theory components of the consumer production technology
reflect the (necessarily subjective) theories of a person does not imply
that they allow for unlimited idiosyncracy. Although, as I mentioned
above, the competitive pressure that tends to homogenize firm produc-
tion technologies in a market context does not constrain consumer
production. choice of technology in the latter context is certainly not
arbitrary. If we presume that people are informed about their relevant,
more immediate environment. we should expect people's theories about
35 The phrase "explicit or implicit" is used here in order to indicate that the relevant "expec-
tations" or "theories" may be habitual and that the actor may not be able to articulate
them.
16
17