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.
Source Description
This file contains:
Public poll. RE: The approval/disapproval of Nixon's job as president. 1 pg. [Subject: Campaign] [Other Document], no date
Public Poll. RE: The approval/disapproval of the way Nixon is handling the Vietnam situation. 1 pg. [Subject: Foreign Policy] [Other Document], no date
Public Poll. RE: Voter opinion on whether the operations in Laos will shorten, lengthen, or won't make any difference in the Vietnam War? 1 pg. [Subject: Foreign Policy] [Other Document], no date
Public Poll. RE: Whether the American public feels the operations in Laos will shorten, lengthen, or not make any difference in the Vietnam War? 1 pg. [Subject: Foreign Policy] [Other Document], no date
Public Poll. RE: Approval/disapproval of President Nixon's policy of incursion into Laos as a means to continue the American troop withdrawal from Vietnam. 1 pg. [Subject: Foreign Policy] [Other Document], no date
From Gordon Strachan. RE: Trial Heat and Approve/Disapprove: Registration (Intend to Register) 18-20 Year Old Vote, leaner question. 3 pgs. [Subject: Campaign] [Memo], 4/1/1971
Opinion research surveys of January 28 and February 4, 1971. Results centered on the question of whether or not the Americans approve/disapprove of the way Nixon is handling his job as President. 1 pg. [Subject: Campaign] [Other Document], 3/23/1971
From Gordon Strachan to Trial Heat and Approve/Disapprove: Registration (Intend to Register) 18-20 Year Old Vote, Leaner Question. 3 pgs. [Subject: Campaign] [Memo], 4/1/1971
The Public Opinion Polling Industry and American Politics, by Peter H. Rossi. RE: The origins and subsequent development of the modern polling system. 28 pgs. [Subject: Domestic Policy] [Other Document], no date
RE: "Kennedy Leads Muskie and Humphrey but All Three Trail President Nixon in Latest Test Election." 1 pg. [Subject: Campaign] [Memo], no date
Scholar Source Context
Document identity
localId
26145519
label
WHSF: Contested, 11-10
core
doc
dtoType
document
citationUrl
pageCount
1
Source metadata
id
26145519
sourceUrl
contentType
document
title
WHSF: Contested, 11-10
description
This file contains:
Public poll. RE: The approval/disapproval of Nixon's job as president. 1 pg. [Subject: Campaign] [Other Document], no date
Public Poll. RE: The approval/disapproval of the way Nixon is handling the Vietnam situation. 1 pg. [Subject: Foreign Policy] [Other Document], no date
Public Poll. RE: Voter opinion on whether the operations in Laos will shorten, lengthen, or won't make any difference in the Vietnam War? 1 pg. [Subject: Foreign Policy] [Other Document], no date
Public Poll. RE: Whether the American public feels the operations in Laos will shorten, lengthen, or not make any difference in the Vietnam War? 1 pg. [Subject: Foreign Policy] [Other Document], no date
Public Poll. RE: Approval/disapproval of President Nixon's policy of incursion into Laos as a means to continue the American troop withdrawal from Vietnam. 1 pg. [Subject: Foreign Policy] [Other Document], no date
From Gordon Strachan. RE: Trial Heat and Approve/Disapprove: Registration (Intend to Register) 18-20 Year Old Vote, leaner question. 3 pgs. [Subject: Campaign] [Memo], 4/1/1971
Opinion research surveys of January 28 and February 4, 1971. Results centered on the question of whether or not the Americans approve/disapprove of the way Nixon is handling his job as President. 1 pg. [Subject: Campaign] [Other Document], 3/23/1971
From Gordon Strachan to Trial Heat and Approve/Disapprove: Registration (Intend to Register) 18-20 Year Old Vote, Leaner Question. 3 pgs. [Subject: Campaign] [Memo], 4/1/1971
The Public Opinion Polling Industry and American Politics, by Peter H. Rossi. RE: The origins and subsequent development of the modern polling system. 28 pgs. [Subject: Domestic Policy] [Other Document], no date
RE: "Kennedy Leads Muskie and Humphrey but All Three Trail President Nixon in Latest Test Election." 1 pg. [Subject: Campaign] [Memo], no date
citationUrl
collections
Richard M. Nixon's Returned Materials Collection
Contested Materials Files
imageCount
1
hasImages
yes
source
import
hasTranscription
no
Source extras
naId
26145519
levelOfDescription
fileUnit
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
94428539319abe89
ocrText
Richard Nixon Presidential Library
Contested Materials Collection
Folder List
Box Number
Folder Number
Document Date
No Date
Subject
Document Type
Document Description
11
10
Campaign
Other Document
Public poll. RE: The approval/disapproval of
Nixon's job as president. 1 pg.
11
10
Foreign Policy
Other Document
Public Poll. RE: The approval/disapproval of
the way Nixon is handling the Vietnam
situation. 1 pg.
11
10
Foreign Policy
Other Document
Public Poll. RE: Voter opinion on whether
the operations in Laos will shorten, lengthen,
or won't make any difference in the Vietnam
War? 1 pg.
11
10
Foreign Policy
Other Document
Public Poll. RE: Whether the American
public feels the operations in Laos will
shorten, lengthen, or not make any difference
in the Vietnam War? 1 pg.
Monday, October 18, 2010
Page 1 of 3
Box Number
Folder Number
Document Date
No Date
Subject
Document Type
Document Description
11
10
Foreign Policy
Other Document
Public Poll. RE: Approval/disapproval of
President Nixon's policy of incursion into
Laos as a means to continue the American
troop withdrawal from Vietnam. 1 pg.
11
10
4/1/1971
Campaign
Memo
From Gordon Strachan. RE: Trial Heat and
Approve/Disapprove: Registration (Intend to
Register) 18-20 Year Old Vote, leaner
question. 3 pgs.
11
10
3/23/1971
Campaign
Other Document
Opinion research surveys of January 28 and
February 4, 1971. Results centered on the
question of whether or not the Americans
approve/disapprove of the way Nixon is
handling his job as President. 1 pg.
11
10
4/1/1971
Campaign
Memo
From Gordon Strachan to Trial Heat and
Approve/Disapprove: Registration (Intend to
Register) 18-20 Year Old Vote, Leaner
Question. 3 pgs.
11
10
Domestic Policy
Other Document
The Public Opinion Polling Industry and
American Politics, by Peter H. Rossi. RE:
The origins and subsequent development of
the modern polling system. 28 pgs.
Monday, October 18, 2010
Page 2 of 3
Box Number
Folder Number
Document Date
No Date
Subject
Document Type
Document Description
11
10
Campaign
Memo
RE: "Kennedy Leads Muskie and Humphrey
but All Three Trail President Nixon in Latest
Test Election." 1 pg.
Monday, October 18, 2010
Page 3 of 3
Presidential Materials Review Board
Review on Contested Documents
Collection: H. R. Haldeman
Box Number:
232
Folder:
[Gordon Strachan Misc. Memos 1971]
Document
Disposition
37
Retain
Open
38
Retain
Open
39
Retain
Open
40
Retain
Open
41
Retain
Open
42
Retain
Open
43
Retain
Open
44
Retain
Open
45
Retain
Open
46
Retain
Open
47
Retain
Open
48
Retain
Open
49
Retain
Open
50
Retain
Open
51
Retain
Open
52
Retain
Open
53
Retain
Open
54
Return
Private/Political Poll Results, 1971.
55
Retain
Open
56
Retain
Open
57
Retain
Open
58
Retain
Open
59
Retain
Open
60
Retain
Open
Presidential Materials Review Board
Review on Contested Documents
Collection: H. R. Haldeman
Box Number: 232
61
Return
Private/Political Memo, Strachan to Files, 4/1/71.
62
Retain
Open
63
Retain
Open
64
Retain
Open
65
Retain
Open
66
Retain
Open
67
Retain
Open
68
Retain
Open
69
Return
Private/Political Survey Results, 3/23/71.
70
Return
Private/Political Memo, Strachan for Files, 4/1/71
71
Return
Private/Political "The Public Opinion Polling INdustry and
72
Return
Private/Political
American Politics," Peter H. Rossi, 11/70.
Memo, "Kennedy Leads Muskie
"
73
Retain
Open
74
Retain
Open
75
Retain
Open
76
Retain
Open
77
Retain
Open
78
Retain
Open
79
Retain
Open
80
Retain
Open
File- Narch6-7poll
Page 1
DO YOU APPROVE OR DISAPPROVE OF THE WAY RICHARD NIXON IS HANDLING HIS JOB AS PRESIDENT?
PERCENTAGE
DON'T
BASE
APPROVE
DISAPPROVE
KNOW
TOTAL PUBLIC
1058
51
36
13
MEN
508
51
37
12
WOMEN
523
52
34
14
18 - 20 YEARS OF AGE
55
47
41
12
21 - 29 YEARS
225
55
37
8
30 - 49 YEARS
426
52
34
14
50 YEARS AND OVER
350
49
37
14
8TH GRADE UR LESS
106
53
32
15
HIGH SCHOOL INCOMPLETE
142
37
49
14
HIGH SCHOOL GRADUATE
384
55
33
12
SOME COLLEGE
422
56
35
9
UNION FAMILIES
270
47
41
12
NONUMION FAMILIES
788
53
34
13
WHITE
935
56
31
13
NONWHITE
116
22
67
11
PROTESTANT
618
53
33
14
CATHOLIC
254
53
37
10
UNDER $5,000 INCOME
171
44
42
14
$5,000 - $15,000
564
52
35
13
OVER $15,000
231
56
36
B
NOW REGISTERED
891
51
37
12
WILL REGISTER
122
54
36
10
REGISTERED 08 WILL REGISTER
1013
52
36
12
WON'T REGISTER
45
41
33
26
TOTAL REPUBLICAN
338
82
10
B
LEAN REPUBLICAN
126
78
15
7
TOTAL DEMOCRAT
536
35
52
13
LEAN DENOCRAT
164
36
53
11
INDEPENDENT
121
50
30
20
TOTAL CONSERVATIVE
505
61
26
13
LEAN CONSERVATIVE
240
57
29
14
TOTAL LIBERAL
361
39
54
7
LEAN LIBERAL
156
43
52
5
(N BETWEEN
121
39
34
27
NIXOM VOTERS
447
72
18
10
HUMPHREY VOTERS
228
29
63
8
WALLACE VOTERS
39
43
47
10
NONVOTERS
205
47
39
14
EAST
260
49
41
10
MIDWEST
315
46
39
15
SOUTH
312
58
29
13
WEST
171
51
3175
12
HEARD/READ ABOUT LAOS
871
52
37A
11
0001
the
14
18
Page 2
DO YOU APPROVE OR DISAPPROVE OF THE WAY NIXON IS HANDLING THE VIETNAM SITUATION?
PERCENTAGE
DON'
BASE
APPROVE
DISAPPROVE
KNOW
TOTAL PUBLIC
1058
41
47
12
MEN
508
44
46
10
WOMEN
523
39
48
13
18 - 20 YEARS OF AGE
55
21
67
12
21 - 29 YEARS
225
39
48
13
30 - 49 YEARS
426
41
47
12
50 YEARS AND OVER
350
45
44
11
8TH GRADE OR LESS
106
40
46
14
HIGH SCHOOL INCOMPLETE
142
28
61
11
HIGH SCHOOL GRADUATE
384
45
43
12
SOME COLLEGE
422
46
45
9
UNION FAMILIES
270
36
49
15
NONUNION FAMILIES
788
43
47
10
WHITE
935
45
43
12
NONWHITE
116
15
72
13
PROTESTANT
618
46
43
11
CATHOLIC
254
36
51
13
UNDER $5,000 INCOME
171
37
51
12
$5,000 - $15,000
564
40
49
11
OVER $15,000
231
53
41
6
NOW REGISTERED
891
43
46
11
WILL REGISTER
122
31
54
15
REGISTERED OR WILL REGISTER
1013
41
47
12
WON'T REGISTER
45
29
55
16
TOTAL REPUBLICAN
338
66
27
7
LEAN REPUBLICAN
126
59
32
9
TOTAL DEMOCRAT
536
27
59
14
LEAN DEMOCRAT
164
32
60
8
INDEPENDENT
121
48
40
12
TOTAL CONSERVATIVE
505
49
41
10
LEAN CONSERVATIVE
240
43
48
9
TOTAL LIBERAL
361
31
59
10
LEAN LIBERAL
156
36
53
11
IN BETWEEN
121
32
48
20
NIXON VOTERS
447
63
28
9
HUMPHREY VOTERS
228
23
68
9
WALLACE VOTERS
39
39
55
6
NONVOTERS
205
27
58
15
EAST
260
37
50
13
MIDWEST
315
38
50
12
SOUTH
312
47
42
11
WEST
171
38
50
12
HEARD/READ ABOUT LAOS
871
44
47
9
Page 4
DO YOU THINK THESE OPERATIONS IN LAOS WILL SHORTEN THE VIETNAM WAR,
LENGTHEN THE VIETNAM WAR, OR WON'T IT MAKE ANY DIFFERENCE?
A. PER CENT ASKED
THIS QUESTION
1. SHORTEN THE VIETNAH WAR
2. LENGTHEN THE VIETNAM WAR
3. WON'T MAKE ANY DIFFERENCE
4. DON'T KNOW
PERCENTAGE
BASE
A.
1.
2.
3.
4.
TOTAL PUBLIC
871
34
25
22
19
MEN
447
41
25
24
10
HOMEN
405
27
24
21
28
18 - 20 YEARS OF AGE
34
26
46
22
6
21 - 29 YEARS
193
27
30
26
17
30 - 49 YEARS
360
36
23
23
18
50 YEARS AND OVER
282
37
21
19
23
8TH GRADE OR LESS
74
33
24
19
24
RIGH SCHOOL INCOMPLETE
99
27
28
27
18
HIGH SCHOOL GRADUATE
305
37
21
22
20
SOME COLLEGE
390
35
28
23
14
UNION FAMILIES
227
35
22
25
18
NONUNION FAMILIES
644
34
26
21
19
WHITE
787
35
24
21
20
NONWHITE
77
24
35
34
7
PROTESTANT
512
38
24
20
18
CATHOLIC
209
33
26
24
17
UNDER $5,000 INCOME
126
32
27
16
25
$5,000 $5 - $15,000
471
35
25
23
17
OVER $15,000
211
39
23
27
11
NOW REGISTERED
757
35
24
22
19
WILL REGISTER
87
31
30
23
16
REGISTERED OR WILL REGISTER
844
35
24
22
19
WON'T REGISTER
27
11
37
26
26
TOTAL REPUBLICAN
291
49
14
17
20
LEAN REPUBLICAN
112
48
17
16
19
TOTAL DENOCRAT
429
24
30
26
20
LEAN DEMOCRAT
150
26
32
25
17
INDEPENDENT
105
36
23
24
17
TOTAL CONSERVATIVE
435
41
19
23
17
LEAN CONSERVATIVE
205
37
22
26
15
TOTAL LIBERAL
300
30
37
20
13
LEAN LIBERAL
127
30
35
22
13
IN BETWEEN
91
21
21
31
27
NIXON VOTERS
388
50
15
17
18
HUMPHRLY VOTERS
191
21
37
23
19
WALLACE VOTERS
35
40
23
17
20
NONVOTERS
148
23
31
28
18
EAST
223
35
29
17
19
MIDWEST
255
31
22
29
18
SOUTH
250
35
24
22
19
WEST
143
35
26
20
19
HEARD/READ ABOUT LAOS
871
34
25
22
19
0004
Page 5
DO YOU THINK THESE OPERATIONS IN LAOS WILL SHORTEN THE VIETNAM WAR,
LENGTHEN THE VIETNAM WAR, OR WON'T IT MAKE ANY DIFFERENCE?
A. PER CENT ASKED
THIS QUESTION
1. SHORTEN THE VIETNAM WAR
2. LENGTHEN THE VIETNAM WAR
3. WON'T MAKE ANY DIFFERENCE
4. DON'T KNOW
PERCENTAGE
BASE
A.
1.
2.
3.
4.
TOTAL PUBLIC
1058
78
26
19
17
15
MEN
508
84
35
21
20
9
WOMEN
523
73
20
18
15
20
18 - 20 YEARS OF AGE
55
60
16
28
13
3
21 - 29 YEARS
225
83
22
25
21
14
30 - 49 YEARS
426
80
29
18
18
14
50 YEARS AND OVER
350
77
28
16
15
17
8TH GRADE OR LESS
106
69
23
17
13
17
HIGH SCHOOL INCOMPLETE
142
70
19
20
18
12
HIGH SCHOOL GRADUATE
384
79
29
17
17
16
SOME COLLEGE
422
93
32
26
21
13
UNION FAMILIES
270
80
28
17
20
15
NONUNION FAMILIES
788
77
26
20
16
15
WHITE
935
80
28
19
17
16
NONWHITE
116
64
16
23
21
5
PROTESTANT
618
81
30
19
16
15
CATHOLIC
254
75
25
20
18
13
UNDER $5,000 INCOME
171
71
23
19
11
18
$5,000 -- $15,000
564
80
28
20
18
14
OVER $15,000
231
89
35
20
24
10
NOW REGISTERED
891
81
29
19
18
15
WILL REGISTER
122
68
21
20
15
11
REGISTERED OR WILL REGISTER
1013
79
28
19
18
15
RON'T REGISTER
45
55
6
20
14
14
TOTAL REPUBLICAN
338
82
40
12
14
16
LEAN REPUBLICAN
126
86
4).
14
14
17
TOTAL DEMOCRAT
536
76
19
23
20
15
LEAN DEMOCRAT
164
89
23
29
22
15
INDEPENDENT
121
84
30
19
20
14
TOTAL CONSERVATIVE
505
83
34
16
19
14
LEAN CONSERVATIVE
240
82
30
18
22
12
TOTAL LIBERAL
361
79
23
30
16
10
LEAN LIBERAL
156
78
23
27
17
10
IN BETWEEN
121
71
15
15
22
19
NIXON VOTERS
447
83
41
13
14
15
RUMPHREY VOTERS
228
79
16
30
18
15
WALLACE VOTERS
39
90
36
20
15
18
NONVOTERS
205
68
16
21
19
12
EAST
260
82
29
24
14
16
MIDNEST
315
76
23
17
22
14
SOUTH
312
77
27
18
17
15
WCST
171
78
28
20
15
15
HEARD/READ ABOUT LAOS
871
100
34
25
22
19
0005
Page 6
PRESIDENT NIXON HAS SAID THAT THE INCURSION INTO LAOS IS PART OF HIS PLAN TO
CONTINUE THE AMERICAN TROOP WITHDRAWAL FROM VIETNAM. DO YOU APPROVE OR
DISAPPROVE OF THIS INCURSION INTO LAOS?
PERCENTAGE
DON'T
BASE
APPROVE
DISAPPROVE
KNOW
TOTAL PUBLIC
1058
43
38
19
MEN
508
48
38
14
WOMEN
523
37
38
25
18 - 20 YEARS OF AGE
55
33
54
13
21 - 29 YEARS
225
42
42
16
30 - 49 YEARS
426
44
39
17
50 YEARS AND OVER
350
42
34
24
8TH GRADE OR LESS
106
39
32
29
HIGH SCHOOL INCOMPLETE
142
31
45
24
HIGH SCHOOL GRADUATE
384
49
34
17
SOME COLLEGE
422
46
46
8
UNION FAMILIES
270
43
35
22
NONUNION FAMILIES
788
42
40
18
WHITE
935
45
35
20
NONWHITE
116
26
59
15
PROTESTANT
618
45
36
19
CATHOLIC
254
44
36
20
UNDER $5,000 INCOME
171
35
41
24
$5 000 - $15,000
564
45
38
17
OVER $15,000
231
48
39
13
NOW REGISTERED
891
43
38
19
WILL REGISTER
122
40
41
19
REGISTERED OR WILL REGISTER
1013
43
38
19
WON'T REGISTER
45
35
42
23
TOTAL REPUBLICAN
338
58
23
19
LEAN REPUBLICAN
126
57
28
15
TOTAL DEMOCRAT
536
34
47
19
LEAN DEMOCRAT
164
32
54
14
INDEPENDENT
121
47
32
21
TOTAL CONSERVATIVE
505
49
33
18
LEAN CONSERVATIVE
240
44
37
19
TOTAL LIBERAL
361
40
49
11
LEAN LIBERAL
156
46
45
9
IN BETWEEN
121
36
37
27
NIXON VOTERS
447
55
23
22
HUMPHREY VOTERS
228
29
55
16
WALLACE VOTERS
39
33
49
18
NONVOTERS
205
37
46
17
EAST
260
41
4)
18
MIDWEST
315
39
38
23
SOUTH
312
46
36
18
WEST
171
44
41
15
HEARD/READ ABOUT LAOS
871
44
40
16
0006
Hallup
April 1, 1971
MEMORANDUM FOR FILES
FROM:
GORDON STRACHAN
SUBJECT:
Trial Heat and Approve/Disapprove:
Registration (Intend to Register)
18-20 Year Old Vote
Leaner Question
Registration
Gallup*
a. In all Trial Heat questions Gallup asks "Ase you now
registered to vote?"
b. Gallup does not ask "do you intend to register to vote in
either trial heat nor approve/disapprove questions."
c. Gallup probably has registration information for approve/
disapprove but he does not publish it.
Derge
a. On both Trial Heat and Approve/Disapprove Derge asks:
"Are you now registered to vote?"
b. If the interviewed says no, Derge asks: "Do you intend to
register for the 1972 presidential election?"
c. Derge began asking the registration question in December
1970 (Study #9575)
Harris**
a.
He has always asked a registration question on trial
heat questions. We do not know if he asks an intend
to register question.
b.
Harris does not ask a registration or intend to register
question on any approve/disapprove questions.
18 - 20 YEAR OLD VOTE
Gallup*
Gallup has been including the 18 - 20 year old votes in both
trial heat and approve/disapprove questions since January 1, 1971.
Harris**
In approve/disapprove questions Harris has always included
18 - 20 year olds.
Trial Heat questions by Harris have included 18 - 20 year olds
since January 1. 1971.
Derge
He has been including 18 - 20 year olds since the December
28 - 30, 1970 study.
Leaners
Gallup
The lead in trial heat question is: To get some idea of the
national political situation at this early stage, suppose the
Presidential election were being held today. If Richard Nixon
were the Republican candidate and if (Edmund Muskie) were the
Democratic candidate, which would you like to see win?
If the answer is "don't know," this leaner question is asked:
"As of today do you lean more toward Nixon the Republican, or
Muskie the Democrat (or toward Wallace, the third party
candidate)?
Derge
The lead in question is: In 1972 there will be another Presidential
election. Suppose this election were being held today and the
candidates were Richard Nixon and Edmund Muskie, which one
would you vote for?
Now suppose the candidates were Richard Nixon, Edmund Muskie,
and George Wallace as a third party candidate, which one would
you vote for?
If the answereis "don't know", Derge asks" "Would you say
that you lean more toward Nixon or more toward Muskie,
(or more toward Wallace)?
Harris***
The lead in question is: "If the election for President in 1972
were being held today and you had to decide would you vote for
Senator Edmund Muskie for the Democrats, President Richard
Nixon for the Republicans, (or Governor George Wallace as an
Independent)?"
*
Dr. David Derge, March 31, 1971
**
Charles W. Colson, March 31, 1971
***
Harris release, February 1, 1971
RECEIVED MAR 2 3 1971
COMPARISON OF OPINION RESEARCH CORPORATION SURVEYS OF JANUARY 28 AND FEBRUARY 4, '1971
Do you approve or disapprove of the way Richard Nixon is handling his job
as President?
January 28, 1971
February 4, 1971
Approve Disapprove
Don't
Approve
Dis-
Don't
Know
Approve
Know
TOTAL
59
28
13
48
35
17
Male
60
29
11
52
34
14
Female
58
28
14
45
36
19
AGE
18-20
53
38
9
45
49
6
21-29
52
39
9
47
37
16
30-49
62
24
14
49
36
15
50+
60
25
15
48
31
21
RACE
White
61
26
13
51
32
17
Black
31
54
15
24
56
20
********
LABOR UNION
Yes
50
34
16
44
40
16
No
62
27
11
50
33
17
RELIGION
Protestant
65
22
13
52
30
18
Catholic
52
34
14
44
41
15
kheyming
REGION
East
54
30
16
43
42
15
Midwest
54
34
12
48
35
17
South
67
21
12
53
27
20
West
59
29
12
45
43
12
INCOME
--5,000
51
31
18
45
33
22
5-15,000
61
27
12
48
35
17
15,000+
67
27
6
53
38
9
PARTY
Democrat
42
45
13
33
48
19
Republican
82
7
11
78
13
9
Independent
60
28
12
49
32
19
x9thmxx
1968 VOTE
Nixon
77
13
10
67
18
15
Humphrey
33
51
16
24
59
17
Wallace
51
34
15
54
29
17
Don't Know
Didn't Vote
48
37
15
45
39
16
IDEOLOGY
Liberal
49
40
11
40
45
15
Conservative
67
22
11
60
28
12
Other
55
27
18
39
35
26
ORC
April 1, 1971
MEMORANDUM FOR FILES
FROM:
GORDON STRACHAN
SUBJECT:
Trial Heat and Approve/Disapprove:
Registration (Intend to Register)
18-20 Year Old Vote
Leaner Question
Registration
Gallup*
a. In all Trial Heat questions Gallup asks "Age you now
registered to vote?"
b.
Gallup does not ask "do you intend to register to vote in
either trial heat nor approve/disapprove questions."
c. Gallup probably has registration information for approve/
disapprove but he does not publish it.
Derge
a. On both Trial Heat and Approve/Disapprove Derge asks:
"Are you now registered to vote?"
b. If the interviewed says no, Derge asks: "Do you intend to
register for the 1972 presidential election?"
c.
Derge began asking the registration question in December
1970 (Study #9575)
Harris**
a.
He has always asked a registration question on trial
heat questions. We do not know if he asks an intend
to register question.
b.
Harris does not ask a registration or intend to register
question on any approve/disapprove questions.
18 - 20 YEAR OLD VOTE
Gallup*
Gallup has been including the 18 - 20 year old votes in both
trial heat and approve/disapprove questions since January 1, 1971.
Harris**
In approve/disapprove questions Harris has always included
18 - 20 year olds.
Trial Heat questions by Harris have included 18 - 20 year olds
since January 1, 1971.
Derge
He has been including 18 - 20 year olds since the December
28 - 30, 1970 study.
Leaners
Gallup
The lead in trial heat question is: To get some idea of the
national political situation at this early stage, suppose the
Presidential election were being held today. If Richard Nixon
were the Republican candidate and if (Edmund Muskie) were the
Democratic candidate, which would you like to see win?
If the answer is "don't know," this leaner question is asked:
"As of today do you lean more toward Nixon the Republican, or
Muskie the Democrat (or toward Wallace, the third party
candidate)?
The lead in question is: In 1972 there will be another Presidential
election. Suppose this election were being held today and the
candidates were Richard Nixon and Edmund Muskie, which one
would you vote for?
Now suppose the candidates were Richard Nixon, Edmund Muskie,
and George Wallace as a third party candidate, which one would
you vote for?
If the answereis "don't know", Derge asks" "Would you say
that you lean more toward Nixon or more toward Muskie,
(or more toward Wallace)?"
Harris***
The lead in question is: "If the election for President in 1972
were being held today and you had to decide would you vote for
Senator Edmund Muskie for the Democrats, President Richard
Nixon for the Republicans, (or Governor George Wallace as an
Independent)?"
*
Dr. David Derge, March 31, 1971
**
Charles W. Colson, March 31, 1971
***
Harris release, February 1, 1971
The Public Opinion Polling Industry and American Politics*
Peter H. Rossi
Department of Social Relations
The Johns Hopkins University
November 1970
*The author has benefitted from the advice and information of a number
of people. I am particularly indebted to Jack Honomichi, who gave
freely of his time and files. I am also grateful to Sidney Hollander,
Mervin Field and Leo G. Shapiro for their help.
There is a populist strain in American political thought which
accords to public opinion a special place in the formation of public
policy. In this view, the ideal government is one in which the will of
the people is directly and faithfully reflected in public policy: A
public official properly fulfilling his role should not stray too far
ahead or too far behind the main currents of popular thinking; and, the
laws of the land are best when they express the broadest possible popular
consensus.
There are many defects in the populist view, not the least of which
is the elusive nature of public opinion. Bryce clearly saw this problem
-
"The obvious weakness of government by public opinion is the difficulty
in ascertaining it". * Without reliable and authoritative means of
gauging public opinion, each party to a political dispute can with
apparently equal legitimacy invoke the support of public opinion for its
stand and threaten the sanction of public wrath as punishment for its
mistaken rivals.
For the first hundred and fifty years of the republic elections were
the main mode through which public opinion was directly manifested,
although newspapers and periodicals purported then, as they do now, to
reflect the views of their readers. In-between elections, many public
figures carried on extensive correspondence with local notables who
relayed what they perceived to be the main opinion trends in their
particular locality. Noting that public opinion was only imperfectly
*James Bryce The American Commonwealth, New York: McMillan, 1888.
2.
reflected in election contests and in editorial writings, late nineteenth
century populists argued for the widespread use of public referenda as
a way of settling how the people felt on specific issues and even more
important as a device whereby the populace could express itself very
directly on important public issues. By the 1920's many states,
especially those in the Far West where populism had been strongest, had
adapted their state constitutions to make it easy to put contested issues
up for decision by popular referenda.
It is hard to judge whether today we are more less in awe of public
opinion than in the nineteenth century. We certainly know more about the
contours and balances of public opinion on a wide variety of issues. We
also know more about the processes of opinion formation and change.
Because of this increased knowledge it is more difficult these days to
invoke arbritarily the authority of public opinion to justify any
particular stand. We are also more aware of the imperfections of public
opinion: How strangely rigid in some respects and flexible to the point
of fragility in other respects. We also know how wrong public opinion
may be on occasion and how many mistaken beliefs are held by large
portions of the American elector. Yet we are still moved by populist
appeals in our political thought. We still expect public policy somehow
to reflect at least the main tendencies in popular thought and public
officials are still worried whether their stands on issues are within
the boundaries of consensus.
We know more about public opinion today because we have developed
techniques for "ascertaining it". A minor industry has grown up around
the measuring of public opinion, although ironically most of this new
3.
industry is. concerned not with political but with marketing opinions.
The public opinion industry today amounts to about 200 major firms and
possibly an additional 100 minor ones with an annual industry wide gross
income of between two hundred and two hundred and fifty millions. It
is difficult to estimate how much of the industry income is derived from
public opinion polling on political issues: A good guess is that
considerably more than four-fifths comes from marketing studies.
The Origins of Modern Polling:
The essential feature of a modern public opinion poll is the use of
standardized personal interviews administered to small but representative
samples of individuals, the results being projected to estimate the
distributions of opinions in the total population. In this form, the
public opinion polls have their beginnings in the '1930's when a number
of enterprising psychologists and market researchers began to sell the
findings of public opinion polls as syndicated services to newspapers and
magazines.
In retrospect, public opinion polling appears to be rather natural
extension of psychological testing, itself a development fostered by the
success of mass testing of Army recruits during World War I. During the
1920's psychologists had developed a variety of tests of human abilities,
traits and dispositions. Social psychologists had ventured to measure
attitudes although their attempts to do so had rarely carried them out-
side the classrooms. It was the effort to measure consumer preferences
which took the psychologists out of the classroom into the larger
community. Indeed, if anything the major impetus to public opinion
polling came from the advertising industry's attempts to measure the
attractiveness and hopefully the effectiveness of its products.
4.
No single mame is more identified with public opinion polling than
that of George Gallup. In his career, he exemplifies the trends that
came together to start up public opinion polling having been a professor
of journalism before becoming director of research of an advertising
agency. His. founding of the Gallup Poll in 1935 signals the start of
the industry as we know it today. Although several other public opinion
polling efforts started up around the same time, notably Elmo Roper's
Fortune magazine poll, the Gallup poll is the only one which has survived
to the present
The early polls were greeted with considerable skepticism and even
some sarcasm on the part of political officials and journalists. After
all, it seemed hardly likely that respondents would tell the truth to
the women who did the interviewing. Furthermore, the questions put to
the samples were manifestly silly. How could a person sum up his stand
on complicated issues like the Social Security Act with a simple
declaration of support or opposition? But shortly after the polls were
started, an excellent opportunity came by to establish their credibility
in the eyes of the public and among political figures.
The public opinion polls achieved credence in the eyes of the public
and political figures through their successes in forecasting the results
of presidential elections. Of course, election forecasting was nothing
new: Straw votes and mock ballots had been conducted by newspapers and
magazines all during the 1920's and early 1930's, the major survivor
being the very extensive, although not very accurate, city and state
straw votes run by the New York Daily News. Perhaps the most well known
was the national straw vote run by the Literary Digest, a weekly news
magazine in the format of TIME and NEWS Week.
5.
The Literary Digest mailed straw ballots to all telephone subscribers
in the United States, receiving returns from millions but still only from
a portion of all telephone subscribers who in turn over represented the
middle and upper classes. Using these returns the magazine correctly
predicted the outcomes of the 1924, 1928 and 1932 presidential election.
In 1936, however, with the electorate polarized along socioeconomic lines,
the Literary Digest incorrectly predicted a landslide victory for
Alfred Landon. Gallup and Roper, in contrast, correctly predicted a
landslide victory for Franklin D. Roosevelt. The Literary Digest folded
within a few months after the presidential election. Whether its demise
was due to this failure or to other factors the memory of the magazine
lives on mainly in texts on statistical methods as an example of the
wages of the sin of biassed sampling. A consequence of the 1936 election
was to fix Gallup and Roper firmly as authorities in the measurement of
public opinion.
Although judged by present day standards the early public opinion
polls were crude, nevertheless they did have two considerable advantages
over the Literary Digest straw vote and similar efforts. The first
advantage was the employment of a rational sampling plan constructed to
insure that relatively small samples (around 3,000 persons) were
representative of the total American electorate with respect to region,
age, sex and socioeconomic status. The second major advantage was the
use of personal interviewers who were guided by the sampling plan to
choose respondents who in the aggregate were representative of the total
electorate. The sampling plans helped to overcome the biases involved
in the use of such income related lists as telephone subscribers and the
use of interviewers helped to overcome the biases of self selection as
well as insuring that persons who were not able to answer paper and pencil
nuestionnaires were reached,
6.
By the end of World War II public opinion polling was well established
in the United States. In the period shortly after World War II its use
spread to most of the democratic countries of the world. Even the failure
of the polls correctly to predict the outcome of the 1948 US presidential
election was taken in stride serving more as a spur to technical improvement
than leading to any serious reduction in either business or public esteem.
Given the background of populism in American political thought, It
can hardly be viewed as accidental that public opinion polling developed
first in the United States. Indeed, the rationale for gauging public
opinion put forth by the pollsters in the early years was explicitly an
appeal to populism.
In any early volume published on public opinion polling, Gallup*
wrote that the polls would enable elective representatives to find out
quickly and systematically the will of the public and hence the
correspondence between legislators' votes and public opinion could be
made very close. Indeed, Gallup saw the possibility that in the republic
of the future, legislative bodies would be replaced by continuous soundings
of public opinion on major issues. The expectation that public opinion
polling would play a major role in the formation of public policy led
Harry Field to break away from the Gallup organization in the early 1940's
to establish the National Opinion Research Center (with the help of the
Marshall Field Foundation). Fearing the profit making public opinion
polling organizations could only be biassed in the conservative direction,
Field established NORC as a non-profit university affiliate to insure that
*G. H. Gallup and S. F. Rae The Pulse of Democracy, New York: Simon and
Schuster, 1940.
7.
controversial public issues would be studied impartially. It was his
hope that periodic NORC polls would clearly establish what the people
were "really" thinking so that national and local legislators could use
poll results as a guide in voting the will of the people.
It is perfectly clear today that polls will hardly supplant traditional
political processes. Like the referendum and the recall, polling has
developed into an accessory to politics rather than into a central
political device. There are many reasons for the failure of the populist
dreams of the early pollsters, but the major reason is that on most
substantive issues, public opinion follows public policy rather than having
a dynamics of its own. At least in the minds of politicans and professional
pollsters the purpose of public opinion polling has shifted from being
the guide for political figures to being a device for measuring the
effectiveness of political appeals.
Public Opinion Polling Today:
The published public opinion polls are the most visible part of the
industry. A large number of newspapers subscribe to the syndicated
services of the Gallup's American Institute of Public Opinion Research
and the Time-Life, Inc. commissions periodic surveys through Louis Harris
and Associates. In addition, a number of regional polls are supported
by local newspapers. The Los Angeles Times prints the results of sponsors
of the California Poll, conducted by Field Research of San Francisco.
Other important regional or local polls include the Texas Poll, conducted
by Joseph Belden Associates, the Minnesota Poll, sponsored by the
Minneapolis Star Tribune, and the Iowa Poll financed by the Des Moines
Register. In addition, occasional poll results on current public issues
8.
are released to newspapers by Sindlinger & Co. Most recently, during
the recent (1970) Congressional elections, Daniel Yankelovitch, Inc.,
conducted a poll for the New York Times of electoral contests in New York
State.
The high aspirations Harry Field held for the National Opinion
Research Center never materialized. NORC conducted its last study of an
election in 1952 and although it has conducted many surveys on matters
of current public interest the typical outlet for the results has not
been the newspapers but scholarly journals, books and limited circulation
reports. The Survey Research Center at the University of Michigan has
been studying presidential elections since 1948 and releases its findings
in similar ways long after (usually years) the final returns have been
counted and the winner has been in office for some time. In short neither
of the major university sample survey centers conduct what might be called
public opinion polling in the sense of widely reporting findings close
to the time the basic data are collected.
Some of the early public opinion polls have gone out of existence
as such. Elmo Roper, one of the early pioncers, essentially stopped
his syndicated service after the 1948 election. Archibald Crossley
brought his public polling to an end around the same time although he did
venture forth from retirement to conduct a poll for Nelson Rockefeller
in his bid to attain the Republican presidential nomination in 1968.
The published polls, dominated by the work of Gallup and Harris,
are but a small part of the political polling taking place in the country.
Most of the polls are taken on behalf of candidates and parties and are
never released formally to the public. Occasionally, the results of a
9.
private political poll are "leaked" to the newspapers, but the usual
private political poll is used by its sponsors for a variety of private
political purposes and its results are not widely circulated.
In turn, political polling is but a very small part of the total
sample surveying industry. Most of the sample surveys in this country
are undertaken for market research purposes and the largest market
research firms rarely undertake any political polling. * For the few firms
who undertake political polling for publication as syndicated newspapers
or magazine features, the activity can be regarded mainly as providing
publicity for the firms rather than as serving as a major revenue producing
activity. Indeed, this explains why firms like Elmo Roper and Associates
found it easy to drop this activity after the 1948 elections.
In short, public opinion polling in the usual meaning of the term which
involves release to the public through syndicated newspaper features or
through news magazines or television specials is not a very profitable
activity. It is a useful publicity generating adjunct to an ongoing
market research business but hardly generates enough revenue to keep a
national sample survey organization going.
It is unprofitable because the costs of properly conducting public
opinion polling are higher than the market for publication of results will
bear. When the public opinion polls were started in the 1930's a national
Indeed the largest market research firm (A. C. Nielsen and Co.) whose gross
accounts for more than a third of the total industry income obtains most of
its revenues from conducting periodic inventories of products in large
national samples of retail stores providing up-to-date information to
consumer goods manufacturers on the movements of their products on the last
leg of their journeys into consumer homes.
10.
poll consisting of fifteen minute personal interviews with a sample of 3,000
individuals cost well under $10,000. A properly conducted poll of
approximately the same sample size and interview length today would cost
around $100,000. In fact most national surveys properly conducted usually
involve longer interviews and smaller samples and cost appreciably more.
Most of the major national surveys which are conducted according to the
highest survey standards are not concerned with political opinions or what
candidate preferences. The largest and probably the most accurate of all
sample surveys is the monthly survey of the labor force conducted by the
Bureau of the Gensus in which close to 50,000 houscholds are questionned
concerning the employment status of household members. The Current
Population Survey, as the survey's official title goes, is the basis for
monthly estimates of unemployment, for annual estimates of consumer income
and for other inter-censal mcasurements of population movements. Most of
the other properly conducted national surveys are also supported by the
government through contracts with or grants to such sample survey
organizations as NORC, the Survey Research Center, National Analysts, etc.
The appreciably increased costs of public opinion polling today arise out
of two factors: First, public opinion polling is a Labor intensive activity
and the wages of interviewers have more than tripled since the 1930's;
secondly, technical advances in the art of questionnaire construction,
interviewing and esepcially sampling have all acted to make public opinion
polling more than five times as expensive (corrected for the differences in
1930 and 1970 prices). Commercial firms interested in public acceptance of
their particular products might be willing to invest that much in obtaining
information which may give them a competitive advantage but there is no
comparable market for public opinion polling in the usual sense: Newspaper
11.
editors find that their sales are not much increased by running a column of
public opinion poll results that they are willing to pay enough to offset
the costs of a properly conducted poll.
A published public opinion poll has to stand up under the scrutiny of
technical experts. * Hence published polls have to be conducted according to
at least minimal acceptable current standards or else suffer being attacked
by opponents on technical grounds. The current practices of the published
polls are closer to the minimum acceptable standards than to the best
current procedures. The standards employed are sensitive to criticism:
Indeed, the major improvements in sampling occurred after the 1948 failure
of the polls to predict the victory of Harry S. Truman.
The pressures for high standards in private polling are considerably
less. The constraints imposed by possible public criticism are avoided by
the unpublished nature of the polls. Hence it is possible to obtain
political polls at prices comparable to the 1930 price levels. For example,
during the 1964 senatorial campaign in Illinois, NORC estimated that a
properly conducted opinion poll in that state would cost one of the
candidates approximately $60,000. He subsequently commissioned a private
poll priced at $6,000. During the past (1970) senatorial campaign in
Maryland, one of the candidates obtained a statewide poll for under $5,000,
in contrast to a rockbottom estimate of approximately $50,000 for a properly
conducted one.
*The New York Times recent use of Daniel Yankelovict and Associates during
the 1970 campaign illustrates this point very well. Yankelovich used
telephone interviews, a technique casily criticized because of the well
known bias of telephone usage toward the middle and upper income brackets.
The Times felt constrained in presenting its results to counter this
argument by stating that personal interviews were harder to obtain than
telephone interviews and hence that the apparent bias was being cancelled out.
12.
The private-po)ling industry manages to maintain lower prices by
lowering quality: Bost of. the private polls conducted on behalf of local
and state candidates and party organizations are so shoddy that they border
on fraudulence. Sampling methods long regarded as defective are used:
Sample sizes are ridiculously small; the questionnaires employed are hastily
and poorly constructed; and the interviewers are neither properly trained
or supervised. For any purpose more sophisticated than the most gross
estimate of popular standing, such private polls are worthless. Thus, 3.f a
particular candidate is clearly headed for a landslide victory, these polls
will probably reflect that fact. In the more usual case where the fate of
a local candidate is more equivocal, the defects of such polls invalidate
their use as good estimates of a candidate's standing.
Over the years there has been a tendency for the national parties and
candidates to use private polling of a more sophisticated variety and better
quality. Academic social scientists have been employed to provide advice
and guidance and the men now in the key advisory posts in national campaigns
seem to be more aware of the problem of assuring that their research efforts
are of better quality. *
*One can make an argument that for many practical political purposes highly
sophisticated polling techniques do not yield sufficiently greater amounts
and sufficiently better information than the less sophisticated procedures
to justify the much greater costs involved. With a limited budget and a
limited use projected for polling data, it is undoubtedly wiser to invest
only lightly in public opinion polling. However, the counter argument
is that it may be better to operate with no information than with grossly
incorrect information. For example, polls conducted for Senator Tydings in
the recent Maryland senatorial campaign (1970) showed him running far ahead
of his opponent, a factor which some observers feel led the Senator to
conduct his campaign differently than had he been shown to be trailing.
13.
For this reason, on the national level, * private polling efforts tend
to be more sophisticated and technically of higher quality. During the 1968
presidential campaign, the Republicans employed the services of political
scientist David R. Derge, of Indiana University, to plan and coordinate a
private polling operation largely run through the highly respected Opinion
Research Corporation. The Democrats apparently did not have as well coordinated
an effort dividing the effort among a number of small firms, ** including
Joseph A. Napolitan Associates, Olivert in Quayle and Company and Independent
Research Associates of Chapel Hill, N.C. Some of these firms specialize in
private political polling and tend to be of low cost and engage in work of
corresponding quality.
The Political Functions of Polling:
The published polls are mainly produced as another service to readers.
or viewers of the media in which they appear. In this sense they are
features similar to political columnists and comic strips which the editors
provide because they believe their readers or viewers find such materials
interesting. The perennial question is raised whether such published results
affect the outcome of elections, voters' preferences presumably being altered
by knowing that their proferred candidates are either likely to win or to
lose. In truth, it must be said that there is little evidence that there
*An excellent summary of the political polling conducted in connection with
the 1968 campaign was made by Jack Honomichl, upon which this account draws
very heavily (Jack Honomichl "Political Polling 1968" The Analyst Vol. 1,
#1, March 1969).
**It should be noted that some of these small firms consist mainly of one
or two persons, often operating out of offices in their homes, who act
mainly as research designers, subcontracting the field work to interviewing
services or other research firms.
14.
are any appreciable effects. Candidates who were shown in some polls to be
going down to defeat, e.g. Roosevelt in 1936 and Truman in 1948, were
elected. It is rare that winning candidates obtain much greater shares of
the final vote than they have been shown to be winning by in pre-election
published polls. * Apparently, most voters define the published polls as
just one more part of the campaign, not important enough to modify their
votes or even their intentions to go to the voting booth.
A much more Important question can be raised concerning the effects
of the published polls on main actors on the political stage -- public
officials, legislators, candidates, financial backers of candidates and
others with considerably more than ordinary amounts of interest in the
political life of the nation. Polling on particular issues provides these
groups with some estimate of how the electorate in general stands on a
partícular issue. For example, the steady deterioration of popular support
for President Johnson's conduct of the Viet Nam war as shown in the
published polls of 1966 and 1967 are reputed to be factors both in the
stiffening of opposition to the war in the Senate and in the decision of
the President not to run again for office in 1968. **
Perhaps the greatest popular attention is paid to published polls on
the relative standings of candidates for the Presidency. Starting with polls
on potential candidates for each of the major parties, the published polls
*Even repeated interviewing of the same persons has little effect on their
eventual voting behavior except to make them more interested in the election
and more likely to vote. See Paul F. Lazarsfeld, et. al., The People's
Choice, New York.
Johnson apparently commissioned a number of private polls on popular
appraisals of the Viet Nam war. At least newspaper stories at the time
referred to Johnson's use of results from unspecified polls as expressing
popular support for his actions in the war.
15.
carry out a monitoring up until the day before the election itself. The
final Gallup and Harris poll results were published in 1968 the day before
the election, reflecting interviews taken the weekend just ended. It is
difficult to estimate how much impact the condidate preference polls have
upon any part of the political process It is clear that the public and
the politically active are paying attention: How behavior is modified by
the appearance of the polls is hard to say.
The more important political functions are being played by the private
political polls, those conducted for specific clients and ordinarily not
released to the public. One important function of such polls is to provide
intelligence to public officials on how their policies are faring in the
eyes of the public. The use of private polls in this sense goes back a
long way in the short history of public opinion polling. Franklin D.
Roosevelt relied heavily on Hadley Cantril's Office of Public Opinion
Research at Princeton to conduct a series of polls in the period 1940
through 1942 on American attitudes towards aiding England and France.
Roosevelt apparently monitored very carefully the impact of his moves to
aid our future allies on public opinion holding up the announcement of
additional steps when polling results indicated that the public mood was
not favorable.*
*Roosevelt's use of public opinion polling was not very widely known, even
among behavioral scientists. Several researchers investigating the
relationship between events and changes in public opinion during the
immediately pre World War II period noted that public policy was to some
degree responsive to public opinion trends and speculated that the
correspondence was due to some unknown processes by which political
leaders were responsive to such changes. See especially Jerome S. Bruner
Mandate from the People.
16.
All during World Was II, several government agencies, notably the
Office of War Information and the Office of Price Administration,
commissioned polls on the state of American civilian morale and on popular
reactions to consumer goods shortages. Between 1946 and 1954, the State
Department commissioned a large number of polls on foreign policy issues,
all conducted by NORC, and financed out of the Secretary of State's
discretionary funds.* The NORC polls monitored the popular standing of
Secretary of State Dulles and public reactions to major Cold War moves on
the part of the United States and the USSR.
From all accounts, President Eisenhower showed very little interest in
the kind of intelligence that could be provided by sample surveys. With
government agencies prohibited from conducting any political polls after
1954, ** political polling to provide intelligence to public officials
appeared to have gone into a docline during the Eisenhower years.
The use of private polls for these purposes was started up again under
John Kennedy and has continued through the Johnson and Nixon administrations.
It is difficult to ascertain just how much private polling was undertaken
during this period on behalf of each of the three presidents involved.
*When the existence of such polls was made public in 1964, Congress reacted
negatively prohibiting the State Department and other federal agencies from
commissioning any polls on political issues, thus bringing to an end the
longest serics available on foreign policy issues.
**The prohibition against political polling extended only to the borders of
the United States. The US Information Agency and other government
departments have supported public opinion polling in other countries.
Some large part of the impetus for the spread of public opinion polling
throughout the non-communist world came from the existence of USIA and
later AID funds to be used for this purpose. Indeed, many American firms
established foreign subsidiaries or developed close working relationships
with foreign firms in order to be able to handle the contracts involved.
17.
During Johnson's presidency, the newspapers gave the impression that the
President was kept very much up-to-date on popular feelings about the Vict
Nam War. At least President Johnson was reported as carrying about with
him poll results showing such support which he liked to show to reporters
and columnists.
It is even more difficult to ascertain how such polling is financed.
The President does have discretionary funds available to him for which he
does not have to account in detail. It is also possible that funds for
these purposes are made available through private donors or even donated by a
sympathetic pollster as a service to the President. It is also difficult to
assess the uses to which such polls are put although one can infer from the
fact that the results of such polls do not 100m às important in either
insiders' accounts of presidential decision making or in the political media
that such intelligence is not very important,
Much more important to the American political process are the political
polls conducted in connection with electoral contests. As mentioned above
the 1968 presidential campaign saw private polls being conducted by both
the Democratic and Republic National Committees. The Republicans were
apparently more sophisticated in their use of polling, devising a method of
obtaining quick soundings of popular responses to candidate Nixon's speeches
and other salient events of the campaign. The Democratic Party effort was
less focussed and reputedly less useful to the candidate.
How many private political polls are conducted in connection with
lesser elections is hard to assess for many firms may be engaged in such
activities. Honomichl reported that Market Opinion Research Corporation of
Detroit was involved in more than 50 state wide contests, usually on behalf
18.
of Republican contestants. Oliver Quayle and Associates was involved in a
similar number of state level contests on behalf of Democratic candidates.
Other organizations endoubtedly handled a much larger total, leading to an
estimate of several hundred polling efforts conducted on behalf of one or
another candidate.
There are apparently several uses to which such polls are put. First
of all, in the carly stages of an electoral contest, polling results can
be used to drive opponents out of the contest. Thus, in the recent
Maryland gubernatorial campaign a poll conducted by the incumbent Marvin
Mandel showing him to be clearly leading over all other potential
Democratic candidates was used to convince Sargent Shriver not to enter the
primary campaign.
Secondly polling results can also be used to obtain financial support
being tendered to potential supporters as evidence of the soundness of
investing in the candidate's political fortunes.
Thirdly, a private poll may be used to influence members of the press
corps to treat a candidate more seriously or to otherwise influence the
treatment of a candidate. Thus during the 1960 primary campaign in West
Virginia, Kennedy staff members "leaked" results of a Louis Marris poll to
create a press coverage more favorable to the Kennedy candidacy. This
particular maneuver is credited with considerably increasing the saliency
of the Kennedy campaign and is reputed to have advanced Kennedy's chances.
For good reasons, this use of the private polls is particularly objected to
by members of the public opinion profession, especially those who run the
published polls. The latter fear that if private polls are used to influence
the electorate and media personnel, public regulation of all polling is
19.
more likely to result. There are other reasons for opposition as well: It
to usually the case that only parts of such surveys are "leaked", those
parts most favorable to the candidacy of the man in question. Furthermore,
it is difficult to evaluate the results of a private poll in which the
methods and techniques employed are not revealed.
Fourthly, the polls can be used as devices to monitor the effects of a
campaign itself. This is probably the most sophisticated use of private
political polling and one to which the quality of the usual private
political poll is ordinarily inadequate. To detect shifts in voter
preferences requires delicate and accurate instruments to which the usual
private political poll bears as much resemblance as a baseball bat to a
microtome.
The Public Interest in Public Opinion Polling:
The broadest purpose of public opinion polling is to provide accurate
estimates of the distribution and central tendencies of popular opinions on
matters of public policy. The techniques involved are partly art and
partly science. Questionnaire writing and interviewing are arts which can
be wielded with great skill and sensitivity or used in a clumsy and
insensitive fashion. The scientific aspect of polling derives from the
statistical theory of population sampling. Both the artistic and the
scientific aspects of public opinion polling can be taught and can be
evaluated. It is possible to tell a good public opinion poll, soundly
conducted, from a poor one taken by someone who is poorly trained.
Because public opinion polling can affect the outcome of the political
process, the public interest is great in knowing what sort of value to
place upon the information provided by a public opinion poll. The major
20,
public interest in this sense is in providing at minimum some way of
ascertaining the quality of a given poll and at maximum in assuring that
some standards of quality are maintained. Furthermore, the public interest
is stronger in the case of published public opinion polls and "leaked"
private polls than in the case of private polls, especially if the latter
are used primarily for intelligence purposes internal to the sponsor and
his cadres.
At the present time, 10 is difficult (and in some cases impossIble) to
ascertain enough information about how both the published and private polls
are conducted in order to make judgements of their quality. The well
publicized polling operations conducted by Gallup and Louis Harris are
perhaps easiest to learn about, but even in these cases it is difficult to
obtain precise information on critical items such as sampling and to obtain
copies of questionnaires. The descriptions of sampling techniques
obtainable until recently from Louis Harris and Associates were notable
mainly for obscuring rather than clarifying procedures actually used. Most
published polls do not report the numbers of interviews upon which their
results are based so that it is impossible to know whether the percentages
for example, referring to Negroes in a table, are based on interviews with
ten people or ten hundred.*
If this is the situation with respect to published polls where there
is at least some public pressure to disclose methods and techniques used,
then the situation with respect to the private polls must be considerably
*Even when the results are published in the more leisurely form of book
length monographs, where the time pressures for publication are less and
the space constraints are minimal, Louis Harris often does not reveal the
numbers upon which his percentages in tables are based. For example most
all of the tables have no case bases in William Brink and Louis Harris
Black and White, New York: Simon and Schuster, 1966.
21.
worse. It 10 not possible to obtain any information on the technical side
of private polling operations. The suspicion therefore grows that most such
private pollo are conducted within adequately written questionnaires. poorly
trained interviewers, and haphazard sampling plans.
Legislation has been introduced (althous as yet not acted upon) into
Congress to require that public opinion polling organizations deposit with
the Library of Congress information on the techniques employed by the
organization for polls which are published. A resolution along these lines
was also introduced into the California State Legislature in 1968, but
failed to come to a vote. The American Association for Public Opinion
Research has devised a code of ethics which calls for disclosure of
critical items concerning technique and sponsorship but the Code has no
method for enforcement and is sufficiently vague in critical respects.
Reacting to these criticisms, the major regularly published polls have
formed a new organization. The National Council on Published Polls, whose
membership includes many of the regional polls as well as the two major
national published polls. Up to this point the National Council appears to
be more concerned with heading off regulatory legislations than with setting
forth procedures by which the industry could police itself.
Furthermore, most of the suggested codes of ethics, and proposed
regulatory legislation, address themselves primarily to the published
political polls and not to private polls, where quality is more of an issue.
The arguments pro and con regulatory legislation are not clearly on
one or the other side. On the one hand, it is apparent that the public
opinion industry is reluctant to police itself through its own professional
22.
asseciations. It is also clear that there are undoubtedly some public
opinion polling organizations which by whatever minimum standards one would
apply ought to be drammed out of the profession. On the other hand,
regulatory legislation is not a particularly attractive route to take.
For example, some fear that regulations might expand from mainly technical
to substantive considerations with the end result that some topics may become
taboo
Some form of self-regulation or governmental regulation appears to be
just beyond the horizon. The 1972 presidential election will raise the
issue anew, if the 1970 congressional elections has not already done so in
some states In the long run the end result will be that the published
polls will be pushed to employ higher technical standards.* Raising the
standards for published polls may also have the effect of raising questions
about the private polls. After all, the same legislator who may vote for
requiring the Gallup Poll to disclose its sampling plan may also be moved to
ask what sampling plans are to be employed by the private pollster who
proposes to work with him in his campaign to become re-elected.
The Proper Place of Political Polling:
The populist rationale for public opinion polling has long ago been
sloughed off by both pollsters and political figures. The major reason for
discarding the view of public opinion polls as a device for setting public
policy was the discovery that the relationship between public policy and
currents of public opinion was a very complicated one. First of all, pre-
*It may well be that such a move would force the end of published polls as
raising standards undoubtedly would raise costs appreciably and hence
price polls out of the reach of the media who now purchase them.
23.
election presidential polls indicated that the greatest part of the
electorate was largely unaffected by the campaigns conducted by the
candidates. By and large, most of the electorate have made up their minds
long before the candidates are nominated. Elections are won and lost by
relatively small proportions of the electorate who decide to sit this one
out or participate by casting a ballot, as well as small proportions who
change their preferences in reaction to the specific issues in a compaign.
It turned out that voting was as much an expression of long standing
loyalties to political parties as an expression of agreement with particular
candidates or party platforms. The populist image of an electorate which
in alert to issues and platforms and which calculates its own interests as
well as that of the general commonweal was hardly validated in the polling
results.
Secondly, opinions on specific issues followed public policy as often
as it led public policy. Thus attitudes towards civil rights for Blacks
has shifted radically in the American population since World War II with
the critical turning point between the 1954 Supreme Court decision outlawing
segregation in the public schools. In 1942, NORC interviewers could not
find a single white Southerner in their national samples who approved of
whites and Blacks sharing the same schools. By 1965, a majority of white
Southerners approved of school desegregation.
Similarly during the buildup of our participation of the Viet Nam war,
majorities of the samples surveyed disapproved of each succeeding step of
involvement before the step was taken and approved of the step after it was
taken. It was not until late in 1966 that a trend towards disapproval of
the Viet Nam was began to appear in the polls. Even this seeming exception
could be interpreted as reactions to the growth of vocal opposition to the
war in the Senate.
24.
The currents and trends of public opinion on political issues thus
appears to be the resultant of a complex interplay between long standing
political divisions within the electorate and the way in which public debate
over the issues involved illuminates the connections between those basic
divisions and the various points of view on the issues in question. Thus it
is predictable which portions of the public are going to be more or less in
favor of an issue but the general level of support for particular positions
is influenced strongly by the course of public debate over that issue and
by the policies that may be adopted by legislatures and political leaders.
Thus the last few year's dramatic shift towards more general public support
for legalized abortions Lollows upon the opening up of public debate over
the issue and the graduate liberalization of abortion laws in several
states. Although Catholics as a group remain more opposed to liberalization
than other religious groups, the level of support among Catholics has risen
at almost the same rate as it has among other portions of the public.
The flexibility of public opinion in some areas is matched by its
rigidity in others. The prestige standings of occupations have not changed
appreciably since the first studies conducted in the middle 1920's. American
food preferences have remained virtually constant since they were first
studied around the same time. Similarly, Americans' regard for the
importance of particular public offices, e.g. the Presidency or the
governorship or mayoralty of a large city, has remained virtually constant
regardless of the currently held opinions concerning the incumbents of those
offices. Furthermore, the patterns of constancy and flexibility are not
easy to identify in advance: At one point, it was held that opinions on
issues related to the primordial concerns of family, kinship and ethnicity
would be less flexible than opinions on more remote concerns such as foreign
25.
affairs. But the experiences of the last decade which saw radical shifts on
issues such as desegregation, legalization of abortion, and optimum family
size have belied this generalization. A useful theory of public opinion
formation which is capable of making more or less accurate predictions about
future trends has yet to emerge.
The more we learn about public opinion through the polls the less
important public opinion appears to be as a primary element in the formation
of public policy. This finding has a double-edged implication: On the one
hand, we now know that political leaders can influence public opinion by
their stands on political issues. This frees public policy formation from
the dead hand of the past. On the other hand, it is not entirely clear
which new directions will be accepted by the public and which rejected,
which raises the uncertainty of policy formation, especially since
ultimately the acceptance of public policy by the public through the
electoral process is important to policy makers.
Aside from providing readers and viewers with editorial materials
through the published polls, the major functions that are played by public
opinion polling are similar to those played by market research for individual
firms. Candidates use polls to learn more about the "market" for their
candidacy and to test out the effectiveness of their campaigns in garnering
support from the electorate. The results of public opinion polling can also
be used to validate one's claim to a place on the party ticket and to
convince potential financial backers that their investment will be worth-
while. Public officials and public agencies use public opinion polls to
monitor the effects of their programs and to modify their administrative
actions in the light of the "market". Sample surveys have been used to
assess the effectiveness of programs such as Head Start, to monitor the
26.
effects of the Neighborhood Youth Corps and to provide data for making
decisions on consumer credit through studying the hard goods buying
intentions of the public.
Public opinion polling in the sense of political marketing research
has come to play an important part in the political process. It is not the
role that was envisaged by the early pioneers. Nor is it a role that
appears in anyway to be illegitimate: On the contrary, public policy and
political candidates may be all the better for having better information on
the preferences and opinions of the electorate. The major problem lies in
the accessability of such information and the quality of the information
*
itself,
Public opinion polling, even at the crudest level of competence,
is expensive and hence candidates who have more resources at their command
can obtain more information than others who cannot afford the services of
pollsters. As for quality, pollsters come in many models, sizes and prices.
Entirely too much of the political polling is of shoddy construction and of
dubious accuracy. Vitally connected with the problem of quality is the
difficulty that the consumer has in judging whether or not a set of "facts"
are worthy of attention. The public opinion industry has yet to work out
ways of policing its own ranks. If it fails to do so, we can expect to find
increasing demand from political figures for some sort of public regulation,
at least to the point of full disclosure of methods and techniques.
*Members of the public opinion industry are very much concerned with what
they term illegitimate uses of polls. For example, many of the better public
opinion polling firms require that their clients submit copy to them before
releasing results a move to prevent distortions and omissions in public
and quasi-public release of information. The practice of "leaking" polling
information from private polls to journalists is frowned upon, apparently
because such "leaks" are most likely to be subject to distortion and
omission. I consider these problems to be a subsidiary one to the general
problem of quality control which extends to the presentation of results as
well as to the conduct of the polling operations themselves.
27.
At the present time, no useful theory of public opinion formation and
change appears to be ready to appear on the scene. As a consequence the
field of public opinion appears to be at the level of naturalistic zoology.
Many "facts" are being collected out of which a model of public opinion
may be constructed. But for the time being, the "facts" are mainly used to
provide snapshots of "political markets", paid for by those who apparently
find the "facts" useful.
MEMORANDUM
THE WHITE HOUSE
WASHINGTON
Kennedy Leads Muskie and Humphrey But All Three Trail
President Nixon in Latest Test Election
Senator Edward Kennedy of Massachusetts registers the strongest
showing of three leading Democratic Presidential candidates in test
election against President Richard Nixon according to a recent nation
wide contest conducted by Opinion Research Corporation, Princeton, New
Jersey.
Senator Kennedy received 38% of the vote in a national trial heat but
trails President Nixon by 4 % points. Maine's Senator Muskie received 34%
training the President by 6 % points. Senator Humphrey tallied 35% of the
voters falling 9 percentage points behind the President.
These results were obtained by telephone interviews in a nation-wide
sampling of 1019 persons, ages 18 and under. The interviews were conducted
during the period of March 1 through March 3.
The question asked in each of the three trial heats was: "in 1972 there
will be another Presidential election. If the election were held today and the
candicates were Richard Nixon, (name of the Democratic candidate being
tested), and George Wallace as a third party candidate which one would you
vote for? 11
Following are the results of each of the trial heats:
Kennedy versus Nixon
Muskie versus Nixon
Nixon
42%
Nixon
40%
Kennedy
38%
Muskie
34%
Wallace
3%
Wallace
16%
Undecided
8%
Undecided
10%
Humphrey versus Nixon
Nixon
44%
Humphrey
35%
Wallace
14%
Undecided
7%