Half Title
 Title Page
 Table of Contents
 The 1964 election: What kind of...
 "Johnson Nice": Social reforme...
 Economic troubles
 Poll wars: Fighting the popularity...
 Politics and elections

Digitization of this item is currently in progress.
LBJ and the polls
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00101412/00001
 Material Information
Title: LBJ and the polls
Series Title: University of Florida social sciences monograph ;
Added title page title: L.B.J. and the polls
Physical Description: xvii, 137 p. : ill. ; 24 cm.
Language: English
Creator: Altschuler, Bruce E
Publisher: University of Florida Press
Place of Publication: Gainesville, Fla
Gainesville, Fla
Publication Date: c1990
Copyright Date: 1990
Subjects / Keywords: Public opinion -- History -- United States -- 20th century   ( lcsh )
Opinion publique -- Histoire -- États-Unis -- 20e siècle   ( rvm )
Öffentliche Meinung   ( swd )
Politics and government -- United States -- 1963-1969   ( lcsh )
Politique et gouvernement -- États-Unis -- 1963-1969   ( rvm )
Politics and government -- United States -- 1963-1969   ( lcsh )
USA   ( swd )
United States -- Johnson, Lyndon B -- (Lyndon Baines), 1908-1973
Genre: bibliography   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references (p. 127-131) and index.
Statement of Responsibility: Bruce E. Altschuler.
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 21522501
lccn - 90036674
isbn - 0813009960 (alk. paper)
Classification: lcc - E847.2 .A38 1990
ddc - 973.923/092
System ID: UF00101412:00001

Table of Contents
    Half Title
        Page i
        Page ii
    Title Page
        Page iii
        Page iv
        Page v
        Page vi
    Table of Contents
        Page vii
        Page viii
        Page ix
        Page x
        Page xi
        Page xii
        Page xiii
        Page xiv
        Page xv
        Page xvi
        Page xvii
        Page xviii
    The 1964 election: What kind of mandate?
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
    "Johnson Nice": Social reformer
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
    Economic troubles
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
    Poll wars: Fighting the popularity gap
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
    Politics and elections
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
Full Text

LBJ and the Polls

University of Florida Social Sciences Monograph 77

44t ^

IL1 .

a|?~ G>

C`~-' .



and the Polls

Bruce E. Altschuler

University of Florida Press / Gainesville

Editorial Committee
University of Florida Social Sciences Monographs

George E. Pozzetta, Chairman
Professor of History

James Button A.J. Lamme
Professor of Political Science Associate Professor of Geography

Frederick O. Goddard Paul J. Magnarella
Associate Professor of Economics Professor of Anthropology

John C. Henretta Dorene Ross
Associate Professor of Sociology Associate Professor of Education

Copyright 1990 by the Board of Regents of the State of Florida
Printed in the U.S.A. on acid-free paper @

The University of Florida Press is a member of University Presses of
Florida, the scholarly publishing agency of the State University System
of Florida. Books are selected for publication by faculty editorial commit-
tees at each of Florida's nine public universities: Florida A&M University
(Tallahassee), Florida Atlantic University (Boca Raton), Florida Interna-
tional University (Miami), Florida State University (Tallahassee), Univer-
sity of Central Florida (Orlando), University of Florida (Gainesville), Uni-
versity of North Florida (Jacksonville), University of South Florida
(Tampa), University of West Florida (Pensacola).

Orders for books published by all member presses should be addressed
to University Presses of Florida, 15 NW 15th St., Gainesville, FL 32611.

An earlier version of chapter 5 appeared in Public Opinion Quarterly 50
(Fall 1986): 285-99. Copyright 1986 by the American Association for
Public Opinion Research. Published by the University of Chicago Press.

Library of Congress Cataloging-in-Publication Data

Altschuler, Bruce E.
LBJ and the polls / Bruce E. Altschuler.
p. cm. (University of Florida monographs. Social sciences;
no. 77)
Includes bibliographical references and index.
ISBN 0-8130-0996-0
1. Johnson, Lyndon B. (Lyndon Baines), 1908-1973-Public opinion.
2. United States-Politics and government-1963-1969. 3. Public
opinion-United States-History-20th century. I. Tite.
II. Series.
E847.2.A38 1990 90-36674
973.923'092-dc20 CIP

To Mom and Glenn


Acknowledgments ix

Introduction xi

1. The 1964 Election: What Kind of Mandate? 1

2. "Johnson Nice": Social Reformer 17

3. Vietnam 38

4. Economic Troubles 61

5. Poll Wars: Fighting the Popularity Gap 69

6. Politics and Elections 81

7. Conclusions 100

Notes 111

Bibliography 127

Index 133



BECAUSE MUCH OF THIS BOO K is based on documents in the Lyndon
B. Johnson Library, it could not have been written without the help
of those who made access available. I thank the library for the access
and all the archivists, especially Nancy Smith and Linda Hanson, for
their help in finding the material used here. Anyone who has ever
attempted to negotiate a way through the thousands of cartons of
documents in a presidential library knows the value of such assist-
ance. Thanks go also to the Lyndon B. Johnson Foundation for its
Moody Grant to help defray research expenses.
An earlier version of chapter 5 appeared in the Fall 1986 issue
of Public Opinion Quarterly. Thanks to Albert Gollin, Eleanor Singer,
and the anonymous referees for their useful suggestions about that
As for this book, it was improved considerably by the comments
of George Pozetta and the anonymous reviewers of the University
of Florida Graduate School Social Science Monographs Committee
acting for the University of Florida Press. Above all, my most im-
portant constructive critic was my brother Glenn. Without his thor-
ough reading of the original manuscript, the arguments and focus
of the book would have been far less strong. As my brother, he could
be more direct than anyone else in telling me where changes were
needed. And as a fine writer and historian in his own right, he proved
an excellent judge of what did and did not work in the earliest manu-


opinion polls, but Lyndon Johnson was the first to retain a private
polling firm on a regular basis and to assign an aide to watch and
analyze both these and the published polls.' For subsequent presi-
dents, polls have become a crucial source of information. As two aides
to President Ronald Reagan put it, "Polls continue to grow in their
influence on presidential decision making and they are more influen-
tial than conventionally thought." Political scientist Herbert Asher
has written that "polling has become an integral part of political
events at the national, state, and local levels. There is seldom a major
event or decision in which poll results are not a part of the news
media's coverage and the decision makers' deliberations." Yet, accord-
ing to J. Ronald Milavsky, "We do not yet know much about how
these administrations used such information."2
We know so little mainly because of the reticence of elected offi-
cials. Their preference for being seen as principled decision makers
rather than followers of the whims of public opinion leads them to
limit scholars' access to the internal decision-making process neces-
sary to understand the uses of polls. Such access can best be gained
by studying a presidency whose internal memoranda have been
opened to the public. No source of information on any presidency
provides as much documentation on the use of polls as the Lyndon
Johnson Library, which has made hundreds of memoranda, as well

LBJ and the Polls

as most of the polls and polling reports, available to scholars. Thus,
there are two major reasons for examining the use of polls in the
Johnson administration: his expansion of this use and the availability
of documents. Having these documents, I will explore Johnson's use
of polls and how they affected some of the major decisions of his
When politicians first began to pay serious attention to polls, it
was with the hope that they could improve political decision making.
The publisher of the Washington Post, one of the first newspapers to
publish Gallup Polls regularly, believed that politicians' views of pub-
lic opinion were "frequently inaccurate and irresponsible" because
they "responded to pressure and mistook it for opinion." Polling
"seemed an approach, at least, to a really scientific method."3 Instead
of listening to the leaders of interest groups, other politicians, or
unrepresentative mail, presidents could measure directly the issues
that most concerned the public. Even more important, a president
could measure trends to determine which policies were succeeding
and which were not. As two leading experts on public opinion see
it, the public's main competence is that it is "sensitive to the direction
and adequacy of policies being pursued by its leaders." Public opinion
may take some time to jell, but once it has decided that a policy has
failed, "the public will desert its leaders."4
One major claim for polls, then, is that they permit a president's
advisors to tell him when a policy is failing with the public-
ordinarily, a difficult task for an advisor. But the apparent objectivity
of the polls' numbers permit separation between advisor and mes-
sage. The advisor is presenting only facts, not opinions on the merits
of the policy. "In the course of our wanderings through the political
hinterlands," Louis Harris wrote about his days as a polltaker for
political candidates, "We have had to tell the candidates that they
were considered spineless, arrogant, stuffed shirts, loudmouths, cold,
poor speakers, or were just plain unknown." Even Benjamin Gins-
berg, a critic of polls, agrees that polls can sometimes serve "as anti-
dotes for false spokesmen and as guides to popular concerns that
might never have been mentioned by individuals writing letters to
legislators or newspaper editors." Although there are other measures
of public opinion, "whenever poll results differ from the interpreta-
tion of public opinion offered by some other source, almost invariably
the polls are presumed to be correct."5



Polls, then, may appear to be a solution for what George Reedy,
Johnson's press secretary, sees as the most serious difficulty of the
modern president: isolation from the public. The aura of victory, the
nearly unanimous public approval of the early honeymoon period
and the deference and protection of his advisors insulate a president.
When criticism comes, it is "something a president reads in the news-
paper or views on TV," making it easy to discount. Or, as President
Gerald Ford put it, "Few people, with the possible exception of his
wife, will ever tell a President he is a fool . yet the President-any
President-needs to hear straight talk."6
Thus, polls seem to be the cure for presidential isolation. Theo-
dore Lowi has pointed out, "The polls have themselves become a
vital aspect of direct representation in government" as "pollsters in
the inner councils of the White House continually take polls and
sift resulting evidence." Norman Bradburn and Seymour Sudman
state even more directly, "Only the polls provide a believable alterna-
tive to relying heavily on close friends to provide the voice of the
people."' When a president is in trouble with the public, the adminis-
tration's polltaker should rush in with evidence that the public has
lost faith in his policies. The scientific nature of the evidence will
both prevent the messenger being seen as a critic of those policies
and provide the necessary jolt to make the president reevaluate his
position. What I will demonstrate is that the opposite occurs. Rather
than bringing the president bad news, polls served as an additional
shield. When published polls announced declines in presidential pop-
ularity, Johnson's advisors used private polls-not to emphasize the
seriousness of the decline but to provide ways of demonstrating that
the president was more popular than the published polls indicated.
Rather than providing a direct link to a critical public, polls provided
a barrier to reality that only served to increase presidential isolation.
We will examine how this could happen and what was wrong with
the hopes of the advocates of the "scientific method" of measuring
public opinion.
Lyndon Johnson was most interested in polls measuring public ap-
proval of his administration. This interest is logical, for, as Richard
Neustadt wrote in his classic study Presidential Power, "The Washing-
tonians who watch a President ... have to think about his standing
with the public outside Washington. . Because they think about
it, public standing is a source of influence for him, another factor


LBJ and the Polls

bearing on their willingness to give him what he wants."8 The in-
creased importance of the mass media in the twenty-five years since
Neustadt wrote those words has, if anything, strengthened his point.
The importance of public approval as a source of power, coupled
with the desire of presidents (who, as candidates, have previously
sought such approval at the polls) to be appreciated by their constitu-
ents, has led presidents to devote considerable time and energy to
influencing public opinion.
When Johnson's popularity was high, he was able to use it as a
formidable weapon. Polls provided evidence for him to show those
whose support he needed that what he was asking of them was what
the public was seeking. Knowing how popular he was, the president
could use direct appeals to the public as pressure on Congress to
pass his legislative program. The polls provided useful information
about how to tailor those appeals-what arguments were most effec-
tive, what policies the public most wanted, and which constituencies
could be counted on for support. During the period between taking
office in 1963 and the end of 1965, Johnson had a series of triumphs,
including his landslide victory in the 1964 election and passage of
such major pieces of legislation as the 1964 Civil Rights Act, a tax
cut, federal aid to education, Medicare, The Voting Rights Act, the
establishment of the War on Poverty, and a number of Great Society
Programs. These achievements resulted primarily from Johnson's
skills and his high public support-not from the polls themselves,
though, as we shall see, the polls were a helpful tool. Beginning in
1966, however, the increasing costs of the Vietnam War and the sour-
ing economy took their toll on Johnson's popularity, dulling the edge
of his major weapon. The polls should have signaled him to regroup,
but instead they made him dig in more deeply. The shock of the
Tet offensive and the opposition of Eugene McCarthy and Robert
Kennedy in the contest for the presidential nomination finally dem-
onstrated how unpopular he had become. At the end of March 1968,
Johnson, having finally reassessed his position, dramatically an-
nounced both a Vietnam de-escalation and his decision not to seek
Before we can discuss the efforts of presidents to improve their
public standing, we need to understand how polls measure it. Gallup
first asked the public in 1939 whether it approved of the president
but did not repeat the question on a regular basis until 1951, when



he began asking it every two months except during presidential elec-
tion years. He simply asked, "Do you approve or disapprove of the
way (name of President) is handling his job as President." A different
approach, used by Louis Harris, expands the scale into more catego-
ries, such as excellent, good, fair, and poor. A less common method
lets respondents rate their feelings on a "feeling thermometer" on
a scale of 1-100. These questions can be asked about overall perform-
ance or about specific policy areas, which was done frequently about
the Vietnam War. Despite considerable criticism of all these methods,
widespread reporting of the results has created nearly instant refer-
enda on the president and his policies. Most observers will accept
the general accuracy of the trends portrayed even if they are not
sure just what any single measurement means.9
Much of this book is about the ups and downs ofJohnson's popu-
larity. As with most recent presidents, Johnson's approval ratings de-
clined as his tenure in office advanced. Figure 1 shows just how sharp
and steady that decline was, although it also shows the temporary
rallies that the White House sought to portray as signs of a long-term
recovery. From a high of 80 percent in January 1964, Johnson's Gal-
lup ratings declined to 35 percent in March 1968, surged back to
49 percent after his withdrawal speech later that month, slipped to
35-43 percent for the rest of his term, and returned to 49 percent
only in early 1969 as he was about to leave office.
What can a president do to restore lost popularity? Charles Os-
trom, Jr., and Dennis Simon suggest two basic strategies for improv-
ing popularity ratings. A long-term approach seeks to improve the
quality of life of the public. While offering the eventual possibility
of success, such a strategy presents problems for a president whose
popularity is already in decline. Because it requires an innova-
tive, coordinated, and comprehensive policy that would significantly
change the status quo, this strategy needs the assistance of relevant
political actors. However, the reduction in presidential popularity
makes it more difficult to gain their approval, generating a vicious
circle of popularity decline. Instead, presidents have turned to such
short-term approaches as televised speeches or trips abroad. Such
behavior, as Ostrom and Simon point out, "has a consistent, albeit
short-lived, impact on presidential popularity," becoming "a quick fix
for a president looking to enhance his standing with the public."'0
I will show that, rather than providing evidence for a president







1 2 3 4 5 6 11 2 3 5 6 7 8 11 11 12 2 3 3 3 4 5 6 6 7 8 9 12 2 6 7 7 8 9 11 11 12 1 2 3 4 9 10 11 1 Month
5 26 2 9 30 5 22 5 29 5 26 Day

1964 1965 1966 1967 1968 1969 Year

Fig. 1. LBJ Gallup approval ratings


whose popularity is in decline to force him to reconsider his policy
failures and inaugurate the long-term approach, polls provided John-
son with the illusion that short-term approaches could succeed in
reversing the decline. Even worse, they were used to reinforce his
own rationalizations that there was no real decline. His advisors told
a president eager to believe them that something was wrong with
the published polls-sampling methods, question phrasing, or tim-
ing-or the polltakers themselves. Or they seized on every piece
of positive news, no matter how minor, claiming it was the start of
a revival. The obvious inconsistencies of such arguments-how could
there be a revival if there was no real decline?-passed without notice.
The hopes of advocates of polls as a way for presidents to learn
of their successes and failures are not completely misguided, but the
obstacles to objectivity are far greater than they imagined. In my
concluding chapter I will try to explain why Johnson and his aides
were so blind to these warnings. Perhaps the tale of the failure of
the Johnson White House to heed the polls' warnings will help future
administrations read poll results more objectively.



The 1964 Election:

What Kind of Mandate?

THE ELECTION OF 1964 was the final stage in Lyndon Johnson's
conversion to a belief in the importance of polls. During his unsuc-
cessful 1960 campaign for the Democratic presidential nomination,
he had so little faith in them that he did not even hire a pollster,
preferring, as he had in the past, to follow his own instincts about
public opinion. A major reason for Johnson's skepticism appears to
have been a Joseph Belden poll predicting that Adlai Stevenson
would carry Texas in the 1952 presidential election with 53 percent
of the vote. When Eisenhower received that percentage, the ridicule
that ensued cemented Johnson's opinion. However, the importance
of polls to John Kennedy's presidential campaign began to change
Johnson's views. As Kennedy's running mate, he received survey re-
sults from campaign pollster Louis Harris and from Oliver Quayle,
vice-president of Harris's firm. Johnson was impressed by their accu-
racy. While Belden was predicting that Richard Nixon would carry
Texas, Harris correctly forecast a Democratic victory in Johnson's
home state. Soon after the election, Harris stopped polling for politi-
cal candidates and quickly became one of the two leading public poll-
sters (George Gallup being the other). Quayle formed his own firm
in 1963. Hoping to poll for Kennedy in 1964, he sent Johnson a
survey assessing the ticket's prospects in Oklahoma. Kennedy's assas-
sination made Johnson president, facing an election less than a year

LBJ and the Polls

away. Having been impressed by Quayle, Johnson decided to hire
him for the 1964 campaign.'
As will be shown, Quayle's polls, coupled with high popularity rat-
ings in all polls in the first years of his presidency, gave Johnson
an appreciation of the usefulness of polls. But the nature of his con-
version was to set limits on how he used them that would prevent
him from utilizing them to their potential. Harris said that Johnson
was "the truest believer of polls, but only when they tended to support
what he was doing."2 High poll ratings were used to demonstrate
his popularity to himself and those whose support he wanted; bad
news was ignored, discounted, and attacked. According to Larry Ber-
man, "Johnson's tendency to discredit the messenger imbued his
operating style."3 Such an attitude would make it impossible to use
polls as objective information for decision making. Still worse, it made
Johnson's advisors reluctant to inform him of the full implications
of poll results or, as will be shown, made them anxious to get favor-
able rather than objective data and analysis. Uses of polls other than
demonstrating popularity were sporadic at best and often put aside
in favor of catering to Johnson's obsession with his public standing.
In order to gather information for the upcoming election quickly,
Quayle was commissioned to take a series of state polls in early 1964.
Even though he took numerous surveys for Johnson, the cost was
kept low because Johnson's questions were tacked on to surveys paid
for by others. Bill Moyers claimed that "Quayle never 'made money'
on the President; in 1964 his costs to us for tack-ons were $25,000.
Quayle's surveys that year were about $400,000." These surveys con-
tinued after the election. Moyers estimated that in 1965 and the first
half of 1966 "he completed no less than 79 survey reports at no cost
to us-an average of four a month or one a week." Despite such
bargain rates, Johnson did not rely solely on Quayle for polling infor-
mation, hiring such other polltakers as Alex Louis and Robert
Bower.4 In addition to these private polls, the White House collected
and analyzed any others they could get their hands on, including
the major published polls and polls forwarded by party leaders,
candidates for office, friends of the president, and ordinary citizens.
The analysis of these polls was coordinated by Hayes Redmon and
Richard Nelson during the first years of the administration, later by
Fred Panzer.5

The 1964 Election

The Prenomination Campaign

The first problem facing Johnson was the entry of George Wallace
into the Wisconsin, Indiana, and Maryland primaries. This challenge
posed no threat to Johnson's certain nomination, but a substantial
Wallace vote in these states could indicate a white backlash against
Johnson's civil rights initiatives, which could hurt him in the Novem-
ber election. Because Johnson had chosen to stay above partisan bat-
tles by avoiding overtly political appearances to concentrate on gov-
erning, he could not directly oppose Wallace.6 His continuing high
standing in all polls gave him no reason to change strategy. In late
March, George Gallup told Nelson that even Republican county
chairs expected to lose the election. Louis's early April poll of Texas
gave the president a more than 50 percent lead over potential Repub-
lican candidates Richard Nixon and Barry Goldwater, while a series
of Quayle polls showed results nearly as good in several other states.7
Nevertheless something had to be done to oppose Wallace. It was
decided to place prominent Johnson supporters on the primary ballot
in each state-Wisconsin Governor John Reynolds, Indiana Governor
Matthew Welsh, and Maryland Senator Daniel Brewster. Polls appear
to have played no direct role in this decision. Before the primaries,
Quayle took surveys in Indiana and Maryland but asked no questions
about a Wallace candidacy. Probably because these were tack-ons, rel-
atively few questions were asked about the presidential race. In Indi-
ana, Quayle ran a series of trial heats showing large Johnson leads
against Goldwater, Nixon, Nelson Rockefeller, and William Scranton.
However, despite these leads and high job ratings, Johnson fared
poorly on most major issues. A two-wave Maryland survey (Septem-
ber 1963 and February 1964) showed even better job ratings and
trial heats coupled with slightly favorable ratings on issues such as
57-43 percent positive on race and 55-45 percent positive on jobs.8
A careful analysis of these results would have indicated that John-
son's popularity was primarily personal and not due to issues, making
the stand-in candidates vulnerable. Failing to see this danger signal,
the national campaign provided little help to these candidates. Reyn-
olds received $16,000 from a Citizens for Johnson Committee but
nothing from the state or national party, while Welsh got $10,000
from his state's party-not enough, it soon became clear, for the ex-

LBJ and the Polls

pected big victories. Wallace received 34 percent of the vote in Wis-
consin and 30 percent in Indiana. Less than three weeks before the
Maryland primary, Quayle polled Baltimore and its and Washington's
suburbs, using the results to make an "informed guess" of a statewide
Wallace vote of 30-35 percent. Brewster spent $70,000 from per-
sonal, party, and interest group sources as well as receiving, according
to Herbert Alexander, "surreptitious help of staff members and poll-
ing services from the Democratic National Committee and the ser-
vices of a public relations man from Indiana who had helped Gover-
nor Welsh in his campaign against Governor Wallace."9 These efforts
were too late to stem Wallace's increasing strength. Wallace received
43 percent of the vote, losing to Brewster by only 50,000 votes.
Worries about this result caused the White House to ask Quayle
for another Maryland poll, a full survey rather than a tack-on. It
showed that the primary had damaged Brewster but that Johnson's
popularity remained intact with little damage from the backlash. Ac-
cording to Theodore White, "Johnson read the poll overnight and
was delighted," so he asked for more, receiving similar results from
Indiana and Wisconsin.10 Quayle further enhanced his status by coin-
ing the term "frontlash" to show that Johnson's gains among Republi-
can moderates were greater than his losses of backlash Democrats.
These polls played such a large role in restoring optimism to the
White House that Eric Goldman concluded, "Lyndon Johnson was
not only converted but close to transfixed. He went around with his
pockets stuffed with polls, always ready to pull them out for a stentor-
ian reading.""
Despite the significant use of surveys, Johnson had failed to use
them to their potential. The polls provided information on the com-
ing danger which was not heeded until after the results in Wisconsin
and Indiana had made it obvious. It was only then that more complete
use of surveys was made, and it was done primarily to demonstrate
that Wallace's unexpectedly high vote did not present a serious dan-
ger to Johnson in November. We shall see this theme repeated over
and over in Johnson's presidency. In this case, the danger proved
short-lived, so the failure to utilize polls to understand bad news did
little harm. When Johnson's popularity began to decline later in his
administration, he would not be so fortunate.
Johnson's most difficult problem was his choice of a running
mate.12 Early speculation focused on Attorney General Robert Ken-

The 1964 Election

nedy and Senator Hubert Humphrey. It is clear now that Johnson
never had any intention of selecting Kennedy. "With Bobby on the
ticket, I'd never know if I could be elected on my own."'3 There was
also considerable opposition to Kennedy in the South and in the busi-
ness community. Particularly disturbing to Johnson was a vice-
presidential write-in campaign for Kennedy in the New Hampshire
primary. However, when Kennedy offered to disown the effort if
Johnson asked him to, the president demurred, not wanting to appear
to be forcing Kennedy out. Kennedy then issued a statement which,
while not rejecting the possibility of the nomination, said that he
"wishes to discourage any efforts in his behalf in New Hampshire
and elsewhere" so that Johnson could freely select his own running
Because the certainty of his own nomination made the selection
of a running mate the only source of the suspense needed to generate
interest in the Democratic convention, Johnson sought to encourage
speculation about his choice by bringing in as many candidates as
possible. As Humphrey, the chief victim of this game, later wrote
of Johnson, "It was the kind of situation he delighted in: floating
a trial balloon, deflating it, suggesting different names. He held all
the cards and played them as the whim struck."4 Polls taken for John-
son reflect this desire to include as many possibilities as feasible-if
not for serious consideration by the president, at least for media atten-
tion and to flatter those mentioned. Johnson showed a number of
White House visitors a list of eleven vice-presidential possibilities with
their names included, leading some observers at the time to conclude
that there was a list of ten to which the visitor's name was conveniently
Several of Quayle's polls, taken about the time of the Republican
convention, attempted to test the popularity of possible candidates.
A Maryland survey gave Kennedy a lead over Humphrey but also
indicated that Humphrey hurt the ticket less. A second poll asked
voters about a large list of vice-presidential candidates, including
nearly everyone who had been mentioned except Senator Eugene
McCarthy. Kennedy was preferred by 41 percent, Humphrey by 20
percent, and Robert McNamara by 9 percent, with little support for
any of the others. However, when trial heats were run against a Barry
Goldwater-William Scranton ticket, all of the top three hurt the
ticket, McNamara by 2 percent, Humphrey and Kennedy by 3 per-

LBJ and the Polls

cent, leading Quayle to recommend McNamara.16 Another poll
showed that the public viewed Adlai Stevenson as best qualified but
a poor vote getter, while Kennedy, the best vote getter, was not widely
regarded as well qualified. Humphrey placed a strong second in both
areas. 17
The finding that no candidate added strength to the ticket gave
Johnson an even freer hand while refuting the argument that Ken-
nedy was needed to win the election. When the Republicans chose
a ticket of Goldwater and the little known William Miller, every poll
gave Johnson a huge lead, no matter who else was on the ticket. The
polls thus provided Johnson with scientific evidence that he could
choose any candidate he preferred.
Rather than announce directly that Kennedy would not be se-
lected, on July 30, Johnson told the press that no one from his cabinet
would be chosen. At about the same time he encouraged Humphrey
to demonstrate popular support. Humphrey began organizing; he
commissioned polls, lined up important leaders in politics, labor, and
business to support his candidacy, and increased the number of his
public appearances. His plans to open a campaign headquarters at
the convention site were too much for Johnson, who told him through
a third party to slow down, but he had shown the required support.
Quayle's polls were used to confirm this fact. An Indiana survey in
late July and early August showed Humphrey the choice of 63 per-
cent of those questioned, against 10 percent for McCarthy, 9 percent
for New York City Mayor Robert Wagner, 5 percent for California
Governor Pat Brown, and 1 percent for Senator Edmund Muskie.
The poll appears designed for Humphrey to win, given the weak
field he was put against. When Kennedy's name was added, he de-
feated Humphrey two to one, but neither Kennedy, Humphrey, nor
McNamara affected the results against Goldwater.'8 By this time it
was becoming increasingly obvious that Humphrey would be chosen,
yet Johnson continued to drop more names to maintain the suspense.
He even brought Senator Thomas Dodd, an unlikely possibility, to
Washington on the same plane with Humphrey, hoping to create some
last-minute doubts before finally revealing as the convention began
that it would be Humphrey. From Johnson's point of view the conven-
tion was a success, beginning a campaign that would result in a land-
slide victory.

The 1964 Election

Staying in Front

Because Johnson was so far ahead from beginning to end, there was
little need for high-level strategic decision making. As a result, cam-
paign spending was kept relatively low, but a series of polls was com-
missioned while others were acquired from outside sources.19 These
polls were used for several purposes: to ensure that Johnson re-
mained well ahead so that no changes in strategy were needed; to
evaluate the impact of specific events; to analyze issues; and to evalu-
ate the chances of other Democratic candidates. The more favorable
polls Johnson received, the more he wanted to see. Since the cost
was low, more polls were taken to make the president feel good. Some
were leaked to the press, but, from the earliest stages of the campaign,
only the margin of Johnson's victory was in doubt.
Quayle took a series of state polls subsequent to the convention
(table 1). Fairly large samples were used, partly because some used
screens to eliminate those unlikely to vote. The results pleased John-
son. Except in Georgia, where he was slightly behind, and in North
Carolina, where he was equally slightly ahead, he had huge leads.
Preconvention polls in some of these states had shown Goldwater
even further behind, but Quayle attributed this difference more to
Goldwater's actually receiving the Republican nomination than to any
slipping of Johnson's popularity. While such leads would surely
Table 1. Quayle State Polls, 1964
Interview Sample Trial heat results (%)
State dates size Johnson Goldwater
Georgia 9/30-10/5 675 40 44
Indiana 9/15-9/20 1,270 50 36
Iowa 9/8-9/16 901 46 30
Kentucky 9/2-9/10 766 53 29
Michigan 10/4-10/10 621 56 26
New York 9/11-9/15 1,200 59 22
North Carolina 10/15-10/18 897 45 41
Ohio 9/28-10/3 904 67 21
Pennsylvania 8/29-9/4 901 63 20
West Virginia 9/8-9/12 767 61 20
Source: Office Files of Fred Panzer, Lyndon B. Johnson Library, Boxes 28
(Georgia), 169 (Indiana), 170 (Iowa, Kentucky), 172 (Michigan), 175 (New
York, North Carolina), 176 (Ohio, Pennsylvania), and 178 (West Virginia).

LBJ and the Polls

shrink, small Goldwater gains would make few waves against the
Johnson tide. For example, a follow-up poll taken in Kentucky
showed that even right after the Walter Jenkins affair (discussed later
in this chapter), Johnson's margin had declined by only 4 percent.20
Most of these polls were thorough, including not only trial heats
but also Johnson's popularity ratings, demographic analysis, backlash
and frontlash analysis, and a series of open -and close-ended questions
about twenty or more issues. In many states other important elections
were also discussed either because of White House interest or because
another candidate had paid for the survey to which Johnson's ques-
tions had been added.
Despite his pleasure with these results, Johnson sought additional
confirmation. The states polled by Quayle did not include any from
the West or Southwest. Alex Louis had been hired to poll Louisiana
and Texas in April. Not surprisingly, he found Johnson so far ahead
in his home state that "we would expect no serious threat to Johnson's
carrying Texas in November." In neighboring Louisiana, on the other
hand, Louis found strong personal support for Johnson balanced
by opposition to his civil rights policy, resulting in even trial heats
with Goldwater.21 After the convention, Louis was asked to repoll
these two states. His August and mid-September polls showed John-
son maintaining his standing in Texas, but the situation in Louisiana
had deteriorated. In September, Johnson trailed Goldwater 45 to 38
percent, but there were grounds for optimism since the August mar-
gin had been 55 to 30 percent.22
How accurate were these polls? Comparing poll results with elec-
tion returns is a hazardous exercise. Because these polls were taken
a month or two before the election, subsequent events could change
voters' opinions. Many voters may not have begun to think seriously
about the election, yet their answers count as much as those who
have made a firm decision. One reason candidates take polls is to
see what they can do to gain more votes. Between the poll and the
election, each potential voter will be subject to a barrage of informa-
tion from candidate advertisements, news coverage, and discussions
with friends and acquaintances. And, unlike polls, election results
contain no undecided voters. How questions are asked can also be
a problem. Keeping this in mind, we can see that these polls were
reasonably accurate barometers of opinion. In all of the states in
which Quayle gave Johnson large leads, he won easily. In the two

The 1964 Election

relatively close states, Johnson won North Carolina 56 to 44 percent,
better than Quayle's poll, while Goldwater carried Georgia 54 to 46
percent, about twice Quayle's margin. Johnson carried Texas 63 to
37 percent, the exact percentage found by Louis's final survey, but
Goldwater won Louisiana 57 to 43 percent, making Johnson's Sep-
tember gains appear illusory.
Other sources covered some of the states not polled by Quayle
or Louis, including one poll of California voters paid for by Pauley
Petroleum in May and another in October for Pierre Salinger, who
had resigned as the president's press secretary to run for the Senate.
Just before the election, the Arizona Democratic Party sent the White
House two surveys of Arizona's most populous counties rating the
two candidates virtually tied in Goldwater's home state. One more
Johnson appearance, they pleaded, would carry the state.23 Their esti-
mate was accurate, as Goldwater won Arizona by less than 5,000 votes,
but this episode illustrates a danger of relying on polls taken for
others. The state party wanted Johnson to come to the state to help
their other candidates, so the purpose of sending him the results con-
tained elements of self-interest as well as enlightenment. Had there
been any information showing Goldwater well ahead, it would have
been in their interest to keep it to themselves.
Published polls provided another source of information that will
be discussed later. In addition to these regular analyses, Richard
Scammon undertook an analysis of Gallup Polls by region, which
he extrapolated to states "with a considerable amount of guesstimat-
ing." He concluded that all the East, the Midwest, except possibly
Nebraska and South Dakota, and most of the Far West would go
to Johnson. Only the South provided danger. Scammon found little
evidence of white backlash and speculated that the incidents in the
Gulf of Tonkin could help Johnson's future ratings.24
Not all methods of obtaining poll results were as orthodox as those
described. Some of Johnson's aides were not above spying on the
opposition. If Goldwater's own polls showed him as far behind as
everyone else's, a landslide was a certainty, so confidential sources
in the Republican Party were cultivated. Donald Cook of the Ameri-
can Electric -Power Company was a member of the National Indepen-
dent Committee for Johnson and Humphrey, a group of business
leaders, most of whom were Republicans or independents. Cook,
who, according to White, later declined an offer to become secretary


LBJ and the Polls

of the treasury, had developed contacts in Goldwater's polling firm,
the Opinion Research Corporation. On October 12 he informed the
White House that ORC's results were the same as those of other polls.
"Although these polls are not being talked about," he wrote to Walter
Jenkins, asking confidentiality, "they have produced a great deal of
pessimism in the Goldwater ranks." Two weeks later Tom Finney told
Johnson of a confidential report from Rockefeller's pollsters, Lloyd
Free and Hadley Cantril, on polls they had taken in late September
and October showing Johnson ahead in all regions. From these polls,
Finney predicted that Johnson would carry 46 states (he carried 44)
with 60 percent of the popular vote.25
Although all these methods would seem adequate to measure pub-
lic opinion, experienced politicians were asked to use their powers
of observation to ensure that the polls were correct. Party chairman
John Bailey traveled to New England to talk to political leaders and
concluded that four states were certain, "with a good probability of
the other two." John Bartlow Martin took a similar trip to the Mid-
west, and Charles Roche visited Indiana, Kentucky, Michigan, Ten-
nessee, and Wisconsin to examine the prospects of Democratic con-
gressional candidates in twelve selected districts. Bill Moyers kept a
file of information by states, each one with a road map, congressional
district data, election results for the last twenty years, an analysis of
the state's issues and history, names and addresses of party officials,
biographies of candidates, and mayors, and lists of newspapers with
their editors and publishers, presidential appointees, administration
achievements, and newspaper articles on political developments.26
Although Johnson's large lead made it unlikely that any event
would be catastrophic enough to endanger his reelection, prudence
dictated spotting any danger signs as soon as possible. Therefore,
polls were occasionally taken to measure the impact of specific events.
For example, Robert Bower and Irving Crespi were hired to take
a quick telephone survey to measure reaction to Johnson's speech
accepting the presidential nomination. Bower reported the reaction
to be generally favorable among all groups but without any real en-
thusiasm and so "unfocused as to suggest that little was added to
the campaign in the sense of whipping up enthusiasm or creating
a public focus on central issues or themes."27 In most campaigns such
a report would have caused some concern, but Johnson's large lead
meant that only threatening events, not merely bland speeches, were


The 1964 Election

cause for alarm or even serious reassessment. Since Bower's report
pointed to a basically favorable, if unenthusiastic, reaction, it was read
by campaign aides who appear merely to have noted it. Moyers had
suggested that if the survey proved worthwhile others would be com-
missioned,28 but none appears in the available documents. If any event
would be serious enough to worry about, questions could be added
to Quayle's state polls.
The only such event was the Walter Jenkins affair, which occurred
near the end of the campaign. The president's most trusted aide,
Jenkins was a man who kept out of the public eye while devoting
most of his adult life to working for Johnson. On October 7 he was
arrested in the men's room of the YMCA near the White House on
a charge of "disorderly (indecent gestures)," the police's euphemistic
description of homosexual behavior. Within a few days the Washing-
ton newspapers learned of it and of a similar arrest five years earlier.
When the story came out, despite efforts by Abe Fortas and Clark
Clifford to persuade the newspapers to withhold it on humanitarian
grounds, Jenkins resigned and was hospitalized to recover from the
strain. Quayle was quickly summoned to the president's New York
hotel suite late at night and asked to measure reaction immediately.
Questions about Jenkins added to surveys Quayle was about to take
in Kentucky, Westchester (New York), and North Carolina showed
no significant impact.2 Despite Republican attacks on the issue of
morality in the White House, Johnson gained two points in the next
Harris Poll, largely because international events-Nikita Khrush-
chev's forced resignation, China's first nuclear explosion, and the fall
of Harold Wilson's thirteen-year-old government-overwhelmed the
Jenkins resignation. Polls allowed the campaign to continue as before
without taking emergency measures to counter the scandal.

Issues and Ads

Quayle's polls asked a series of questions about approximately twenty-
five issues. Although he found that white backlash hurt in a few states,
in most it was more than matched by what Quayle termed "frontlash,"
defections by moderate Republicans uncomfortable with Goldwater.
While specific issues rarely had great impact, a few, such as the econ-
omy and Social Security, had such potential. Voters liked Johnson
while viewing Goldwater as dangerously inconsistent.s0 Quayle sug-


LBJ and the Polls

gested a very "presidential" campaign in most states with the use of
only a few issues.31
The campaign turned out to be a mixture of "high road" speeches
and aggressive advertisements attacking Goldwater.32 Goldwater's po-
sitions on nuclear war and Social Security were the major targets.
Each of two nuclear war ads, run only once, created the most contro-
versy. One, the "daisy girl" spot, is the best known political advertise-
ment ever produced. While a little girl picked the petals off a daisy,
the announcer began a countdown culminating in a nuclear explo-
sion. Even though Goldwater's name was not mentioned, the message
was clear enough to generate a storm of Republican protests. The
ad was withdrawn, but the controversy meant that it was shown for
free on every news program. The Social Security ad showed a Social
Security card being ripped up as the voice-over told the audience
that Goldwater's "voluntary plan would ruin the system." Other ads
showed Johnson in the White House and a graphic illustration of
Goldwater's comment about sawing off the eastern seaboard.
Moyers was a strong proponent of this strategy. Although he fa-
vored projecting a positive picture of the president's record, he
wanted to start with these attacks on Goldwater to prevent him from
projecting a more moderate image. Pleased with the results of the
"daisy girl" ad, he urged Johnson to follow it up with more positive
"pro-Johnson, pro-peace, prosperity, preparedness spots." Hayes
Redmon, however, was worried that these ads had gone too far. Be-
cause people were already worried about Goldwater, it would be bet-
ter to combine negative with positive themes: "We will not destroy
Social Security. We will fight for health insurance for the elderly.""33
Once more, we can see the limited use of the polls, not to find
the correct way to use issues or to reassess the campaign's issue strat-
egy but to confirm strategy already chosen. Such confirmation is not
necessarily an incorrect use of polls, but if it is the primary one,
the probability of rationalization-interpreting the polls only in ways
that confirm while ignoring or discounting those that contradict-
increases greatly. With Johnson so far ahead his limited use of polls
did no harm, but it would set a pattern for the rest of his administra-
tion. In 1964, Johnson's campaign followed the classic front-runner
strategy-stress consensus issues, stay vague on more controversial
issues, and contrast the opponent's extremism and your centrism.


The 1964 Election

Johnson sought to retain the traditional Democratic coalition while
adding moderate Republicans. As John Kessel put it, "Through citi-
zens' organizations, television spots, and the President's own speeches
the word had gone out: there must not be any marked departure
from established practices in domestic policy. To be sure, there were
some who were concerned that Mr. Johnson's all-things-to-all-men
approach left the precise contours of the future a little vague." More
succinctly, Johnson told a crowd in Providence, Rhode Island, "We're
in favor of a lot of things and we're against mighty few."4 Once this
strategy was in place, the major use of polls was to make sure it was
working. Quayle's development of "frontlash" was a masterstroke, for
it brought Johnson's attention to what Johnson most wanted to see-
his success in attracting voters. Johnson's high job ratings and contin-
uous overwhelming leads in trial heats added to the message he so
enjoyed. The 1964 election provided no real test of how Johnson
could use polls to help overcome setbacks because the polls showed
only problems so minor that they could be swept under the rug.
One specific recommendation illustrates this use. In early Septem-
ber, Quayle's poll in Kentucky showed that the only issues that mat-
tered were race and economics, the latter an excellent one for John-
son. Quayle recommended the elimination of negative advertising
since the voters were already scared, preferring informal chats by
the president whom the voters believed to be down-to-earth, frank,
and warm. Instead, Moyers recommended heavy use of the Social
Security spot on the basis of a memo from Lee White which empha-
sized the alarm Kentucky voters felt about Goldwater's position on
the issue, concluding that the Democrats "should hit this hard." Al-
though this recommendation was hardly in line with Quayle's desire
for a more positive approach, the poll did show Goldwater to be
weak on Social Security, even if the issue was a secondary one.35 As
long as Johnson retained his large lead, such selective use of polls
would not be noticed.
In a few cases, low ratings on secondary issues were used to suggest
repackaging Johnson's message but, even on these lesser issues, never
to looking inside the package. In early October speechwriter Douglass
Cater analyzed the issue questions from four Quayle surveys taken
since August. Although the president was rated highly on peace, help
for older people, and defense, Cater believed that the administration


LBJ and the Polls

was underrated on jobs, lowering taxes, and waste in government.
A few days later, Cater wrote to Horace Busby that the president
wanted his and Humphrey's speechwriters to take note of this analy-
sis. Cater argued that the administration's record on these issues was
good and needed better explanation, especially on the issue of reduc-
ing waste in government spending, which was high in voter interest
but showed falling ratings for Johnson."
By showing Johnson so far in front, polls also enabled him to reject
deals he might have had to accept in a closer race. On October 24,
Henry Wilson reported that Louisiana Governor John McKeithen
had told Representative Gillis Long that he would endorse Johnson
in exchange for a favorable Tidelands decision. Wilson believed that
McKeithen's endorsement was needed to win the state but that a deal
would strengthen the conservative governor and "wipe out the na-
tional Democrats in Louisiana for the foreseeable future." Labor
leader Vic Bucci took a more optimistic view, citing the Alex Louis
polls mentioned earlier, which showed Johnson cutting into a large
Goldwater lead.37 Johnson's large national lead enabled him to reject
McKeithen's offer.

Allocating Resources

Because the Electoral College makes a presidential election the equiv-
alent of fifty state elections, most candidates need to decide where
best to utilize their limited resources. By providing information about
candidate standing, polls can tell a candidate which states are so hope-
less or so certain that they require few additional resources, which
need shoring up to retain the lead, and which present the possibility
of catching up or pulling ahead. Johnson was so far ahead nearly
everywhere that there was little need for such information, again lim-
iting his use of polls. However, the large lead presented the opportu-
nity to use the president's coattails to carry in Democrats in close
congressional and gubernatorial races. By taking polls of important
elections that were rated close, Johnson could determine which candi-
dates to campaign for, hoping to use his resources to help pull them
in. Although polls were used to study some races, overall their use
was sporadic. Rather than polling all or even only the most important
races, a handful were selected, primarily because the states were of
interest to Johnson for other reasons, or because his polls were tacked


The 1964 Election

on to those of a candidate for another office, or because someone
forwarded a survey to the White House.
The Senate contest in Maryland provides a good example. Because
of Wallace's strong primary showing, that state was polled early, not
to assess the Senate race in which Senator Tydings was in little danger
but to determine what role Senator Brewster should play in the presi-
dential campaign after his stand-in candidacy. Quayle concluded that
Brewster's candidacy had so damaged his popularity that he should
play a "very modest" part in Johnson's campaign. He also confirmed
that Tydings was well ahead in his race, which he eventually won
easily, 63 to 37 percent.38
Johnson received poll results from southern Democrats about pri-
mary contests in which he would play no part because whichever
Democrats won them, they would face no danger from the Republi-
can candidate in the general election. He continued to follow some
of the contests, either out of personal interest or because his own
polls were tacked on to one of the candidate's.39 There were a few
states where Johnson took an early poll but failed to follow up. In
the Senate race in Delaware that loss of interest proved damaging.
In June, Quayle polled to determine whether Republican Senator
John Williams could be defeated. He concluded that despite Wil-
liams's very favorable job ratings (67 percent positive), Johnson's
64-23 percent lead over Goldwater, coupled with voters' perception
of Williams as more conservative than they were, made defeating
him possible though difficult.40 No extra effort appears to have been
made in Delaware, either because the numbers had more weight than
Quayle's conclusion or because other states had higher priority, but
Quayle's recommendation turned out to be quite sound, as Williams
narrowly won with less than 52 percent of the vote. Similarly, Quayle's
early September survey in Pennsylvania showed Genevieve Blatt be-
hind Republican Senator Hugh Scott but with a chance to win.41
Again, there appears to have been little follow-up, resulting in a nar-
row win by Scott with 50.6 percent of the vote, despite Johnson's
65 percent. There is no guarantee that additional effort by Johnson
would have changed the outcome of either of these elections, and
these minor disappointments made little difference as Johnson's land-
slide swept in an overwhelming Democratic majority in both houses
of Congress. Once more, the failure to utilize polls to their full poten-
tial had been hidden by Johnson's overall success.


LBJ and the Polls


The election of 1964 presents a paradox. On the one hand, Johnson's
conversion to a belief in the importance of polls appeared complete.
Compared to previous presidential candidates, even John Kennedy,
his polling operation was quite extensive. On the other hand, the
information that the polls did and could have provided was not used
in any systematic fashion. The primary use was to assure Johnson
that he was well ahead and could continue to do what he had been
doing. Minor setbacks such as Wallace's unexpectedly strong primary
showings and the Jenkins affair could be shown to have no significant
impact on the president's overall popularity. In other words, Johnson
liked polls because they gave him a free hand. Other potential poll
uses such as gauging the impact of issues, determining the public's
image of the candidates, or allocating resources were slighted. The
dangers that they occasionally suggested were so minor that they
could be safely ignored. Unfortunately, a pattern had been set. The
polls existed to tell Johnson either that he was very popular or that
setbacks were merely temporary inconveniences that could be over-
come by remaining steadfast. As long as he remained popular, this
approach did no serious damage, but as his popularity declined, it
was the messenger who got the blame. Polls, which might have helped,
were not allowed to do so.



'Johnson Nice":

Social Reformer

A MAJOR TEXTBOOK ON the presidency expresses a common evalua-
tion of Lyndon Johnson's administration: "He succeeded in putting
on the statute books measures long struggled for in American politi-
cal life and he took on new goals that earlier Presidents would never
have dared to entertain. . The drain of the Vietnam War and
modest appropriations relegated the Great Society program more to
the realm of promise than to here-and-now fulfillment." In his study
of Johnson's character, James Barber goes so far as to speak of two
Johnsons, "Johnson mean and Johnson nice," who coexisted within
the separate worlds of foreign and domestic policy: "To the outer
world of foreign policy, Johnson presented his hard soldierly visage.
To the deprived at home, he showed his kindlier face."'
Legislatively, especially during his first two years, Johnson's accom-
plishments are extraordinary-a series of major civil rights bills,
Medicare, aid to education, the antipoverty program, and the estab-
lishment of the new cabinet-level Department of Housing and Urban
Development and Department of Transportation are only the high-
lights. In 1965 Congress passed 80 of Johnson's 83 major proposals.2
In this chapter we will examine a number of important policy areas
to see what role was played by polls. As in chapter 1, we will examine
both the actual use of polls and their potential.
In the introduction we spoke of the importance of the president's
standing with the public as a major resource in persuading other

LBJ and the Polls

political leaders to support his goals. To convince Congress to pass
his legislative program, Johnson had to demonstrate that he had so
much public support that it would be risky to oppose him. He also
had to show that the public backed the legislative agenda he was pro-
posing. Thus, the perception Congress had of Johnson's public sup-
port was as important as any objective measure of that support. At
first, demonstrating popularity was not difficult-the 1964 landslide
had shown both how strongly the public backed the president and
had swept in a receptive congressional majority. However, the elector-
al mandate alone would not be adequate to demonstrate popularity
for Johnson's entire term. Polls could act as a continuing election,
demonstrating the president's popularity. Dennis Simon and Charles
Ostrom, Jr., wrote, "The president's approval score, as the foremost
indicator of prestige, is a vital piece of information that operates
to shape the perceptions of Washington decision makers and make
them susceptible to persuasion."' When Johnson's approval ratings
were high, he would make certain that Congress was aware of it.
In addition to published polls, he could use those taken for him by
Quayle, not only to demonstrate support in general but also to show
support for particular programs and strength in state or congres-
sional districts represented by legislators whose support he needed. Of
course, it takes more than simply demonstrating public support to
pass legislation. One study of the relationship between legislative suc-
cess and poll ratings showed that even though "public opinion is an
important source of presidential influence in Congress," the relation-
ship is too complex "to be captured by simple-minded quantifica-
tion."4 A president's skill, the nature of his legislative agenda, timing,
the effect of other events both expected and unexpected, the legisla-
tive environment, the influence of interest groups, and variations in
popular support of a president from one issue to another and one
constituency to another all play a part in determining how successful
a president is in achieving his legislative goals.
Polls have the potential of helping a president decide how best
to achieve those goals. For example, they can suggest which issues
most concern the public, which proposals have the greatest popular
support, which arguments are working best, and which constituency
groups can be mobilized to encourage legislators to support the presi-
dent's program. They can also tell a president when and on what
issues he can best appeal directly to the public, for presidents can


"Johnson Nice"

make just so many televised appeals and major speeches without dam-
aging their effectiveness. The limits of polls must also be noted. Politi-
cal reality is too complex and changing to be captured in a single
survey or even a series of surveys. Poll results need the analysis of
someone who knows their limits, understands their ambiguities, and
has the political sophistication to interpret them. On many issues the
public lacks information or interest, making their opinions too super-
ficial to be long-lasting. Simply taking a poll tends to exaggerate the
number of people who hold opinions, for many who have not thought
about an issue, not wishing to appear ignorant, answer anyway.5 In
this chapter we will see the extent to which Johnson made use of
polls to advance his legislative proposals.
While polls are frequently published on a variety of domestic is-
sues, their irregularity, variations in phrasing, and goals that are dif-
ferent from a president's limit their usefulness in decision making.
Unless a president or his staff is able to decide which subjects to ask
about and when, he runs the risk of seriously misjudging public opin-
ion. One critic of polls has argued, "The commercial polls' almost
total disregard for questions pertaining to civil rights, race relations,
and poverty before matters reached a violent flash point in the 1960s
sparked some controversy within the professional polling commu-
nity."6 Although evaluating such a claim is beyond the scope of this
book, it is clear that the agendas of public pollsters differ considerably
from those of presidents.
In order to obtain the information about public opinion that they
need, presidents since Franklin Roosevelt have commissioned private
polls. Johnson, however, was the first president to retain a private
firm on a continuing basis. Quayle made polls available to the White
House after the election in much the same manner as he had before,
sending an average of one a week. Because these were again tacked
on to surveys done for others, he did not charge the administration
for them, but this meant that, because of the election calendar, few
surveys were taken in early 1965, forcing the president to rely on
data from the election when first setting his agenda. Quayle's method-
ology remained similar to that used during the election, except for
a reduction in trial heats. First, respondents were asked to rate the
president. They were then asked a series of questions to determine
the issues that were most important to them, after which Johnson
was rated on each. In addition, specific issues or events were occasion-


LBJ and the Polls

ally asked about in more detail, although, except for Vietnam, not
on a regular basis. These polls provided little depth, but they did
give the White House a reasonable idea of what the public was con-
cerned about and which issues were causing problems for the presi-
dent. When such problems appeared or initiatives were under consid-
eration, more specific questions could be added. Keeping this in
mind, we now turn to specific policy questions.

Civil Rights

Johnson's main use of polls during the election campaign was to mon-
itor his popularity, and he continued to think in the same terms after-
ward. Just how obsessed he was with popularity ratings is shown by
his assertion that history would remember him for "what I did for
the Negro and seeing it through in Vietnam. The Negro cost me
15 points in the polls and Vietnam cost me 20."7 Quayle's emphasis
on the public's ratings of the president for each issue was in line
with Johnson's own concerns.
After taking office, Johnson moved quickly on civil rights. As he
put it in his memoirs, "I decided to shove in all my stack on this
one."8 Because of his narrow margin of victory and limited support
among southern whites, John Kennedy had moved cautiously on civil
rights legislation. An early version of the bill that eventually became
the Civil Rights Act of 1964 was introduced in January 1963, but
the full proposal was not sent to Congress until June. The modified
version was stronger than the original and would be strengthened
still further before final passage.9 Its major provisions included ban-
ning race discrimination in public accommodations such as hotels and
restaurants involved in interstate commerce, authorizing the attorney
general to file school desegregation suits, withholding funds from
government-supported programs that discriminated, and strengthen-
ing the Equal Employment Opportunity Commission.
The new president made the passage of this bill, along with the
less difficult Kennedy tax cut, his major priority. A few days after
Kennedy's assassination, Johnson told ajoint session of Congress, "No
memorial or eulogy could more eloquently honor President
Kennedy's memory than the earliest possible passage of the civil rights
bill for which he fought so hard."
Why did Johnson push so strongly for this bill? As a senator from


"Johnson Nice"

Texas, he had hardly been a champion of civil.rights, although, as
majority leader, he had played a part in the passage of the relatively
weak legislation of the late 1950s. Johnson had to show the nation
that he was not the conservative southerner that many pictured him
to be. He also needed to provide reassurance that there would be
no dramatic break with Kennedy's policies. Therefore he retained
most of Kennedy's advisors and strongly pushed Kennedy's legislative
program through Congress, but it seems equally clear from the ac-
counts of most observers and the obvious passion he brought to his
speeches on the subject that Johnson's backing of civil rights was due
more to genuine commitment than political expediency.'0 Quayle's
polling for the president had not yet begun, so polls played little
Johnson deliberately allowed others to handle the day-to-day tac-
tics in Congress while he sought to mobilize the public through a
series of speeches and the involvement of sympathetic labor, religious,
and civil rights groups." The bill passed the House in mid-February
1964 but was slowed by a Senate filibuster. In order to gain the
needed Republican votes to overcome the filibuster negotiations were
conducted between Minority Leader Everett Dirksen, Majority Whip
Hubert Humphrey, and Attorney General Robert Kennedy. In ex-
change for slight modifications in the employment and public accom-
modations sections, Dirksen agreed to oppose the filibuster. With all
senators voting, cloture was invoked in June, and the final version
signed into law on July 2.
James Harvey argues that Johnson's role, "though significant, was
carried on largely behind the scenes in order to help maintain white
southern support for the rest of his program." Johnson knew from
speaking to friends that his support for civil rights "was destined
to set me apart forever from the South, where I had been born and
reared" but argued that his legislative involvement was kept low-key
so that others, especially Dirksen, could take credit, helping to assure
their support.'2 Although no Quayle polls were available while the
bill was pending, Alex Louis had polled Louisiana and Texas in April.
In the former, he found strong personal support for Johnson coupled
with considerable opposition to his civil rights policy. In Texas, he
asked what voters liked or disliked about Johnson. While likes out-
numbered dislikes, civil rights was the most disliked, cited by 20 per-
cent (compared to the 10 percent who liked it). Probing for more


LBJ and the Polls

detail, Louis found that voting and job rights were favored but not
the public accommodations proposal. Despite these problems, Texans
gave their native son an overwhelming 83 percent approval rating,
along with huge leads in trial heats against potential Republican op-
ponents, so there was little cause for worry in the November election.13
Nevertheless, Wallace's surprisingly strong primary showings and
Goldwater's opposition to the civil rights bill gave Johnson reason
to consider the South his most problematic region. A memo of May
4 from Richard Goodwin to Johnson reflects how seriously the White
House took the danger posed by civil rights in the South: "I believe
that this is worth an awful lot of energy and foresight, since it is
an issue that could dominate the campaign." After the 1964 Civil
Rights Act was passed, the president privately asked black leaders
to refrain from large demonstrations for the rest of the year to avoid
scaring white voters. In response, many of these leaders signed a
statement attacking Goldwater's position on civil rights and urging
concentration on voting in place of mass demonstrations. At their
convention, the Democrats adopted a platform that was relatively
vague on civil rights.'4
As discussed in chapter 1, Quayle, asked to measure the extent
of white backlash, found that Johnson's gains among Republican
moderates exceeded any backlash losses. Nevertheless, Goldwater's
nomination meant that civil rights would remain a potential, even
if not an actual, danger. Quayle paid particular attention to the issue
in southern and border states. A detailed July survey of Maryland,
after the primary but before and during the Republican convention,
showed a modestly positive rating for Johnson on racial issues, but
Quayle concluded that by the election it would be negative. Of the
47 percent who rated it the most important issue, 24 percent consid-
ered themselves pro-segregation, 9 percent pro-integration, and 14
percent undecided. Despite this, Johnson's strongly favorable job rat-
ings and advantages in image and other issues strongly outweighed
any backlash losses.15
As the election neared, Quayle found that the problem remained
latent except in the Deep South. A September poll in Kentucky
showed backlash problems in the eastern part of the state but enough
strength elsewhere that they were more than offset. Quayle estimated
a potential backlash of 18 percent but an actual one of only 6 percent.
Although voters favored "the civil rights law," more specific questions


"Johnson Nice"

showed strong opposition to the public accommodations provision
and proposals to ban housing discrimination. As in Maryland, John-
son was so far ahead on image and other issues, especially economic
ones, that he led the trial heats 73 to 25 percent. An Indiana poll
taken at about the same time had similar results but with a slightly
smaller lead in a more traditionally Republican state. Despite a large
potential backlash in Pennsylvania, little had actually occurred. That
state strongly favored the civil rights law, including public accommo-
dations, but opposed open housing. New York gave even more favor-
able results with a bare majority for open housing.16
The situation in the Deep South was not as good. An early October
survey in Georgia gave Johnson his lowest job rating in any state
yet polled. Quayle cited racial prejudice as the main reason for the
large number of Democratic defections. Civil rights was the leading
issue, and most voters favored racial segregation. On this issue, John-
son was rated negatively by a margin of 73 to 27 percent, with most
of his support from blacks.'7 In essence, this was all Goldwater had
going for him. On election day he won only his home state of Arizona
and five southern states.
After the election, Johnson asked Attorney General Nicholas Katz-
enbach to draft a voting rights bill. The polls had indicated that,
of all civil rights proposals, this one was the least controversial. Never-
theless, Johnson did not anticipate moving too quickly because he
wanted to press for other legislation and give the nation time to digest
the 1964 law. In his 1965 State of the Union address there was only
a brief reference to civil rights, a promise to enforce the law and
eliminate "every remaining obstacle to the right and opportunity to
vote." His inaugural speech included a general condemnation of race
discrimination but nothing specific about voting rights. Nor was an
administration voting rights bill sent to Congress in January or Feb-
With no national elections in 1965, few tacked-on Quayle polls
were available. An Alabama survey in early February merely told
Johnson the obvious-few voters in that state supported him on civil
rights.'9 Before more useful information came in, events began to
force Johnson to speed up his schedule.
In January, Dr. Martin Luther King, Jr., led a series of mass dem-
onstrations to protest barriers to black voting. The next month, the
Civil Rights Commission held hearings in Mississippi that resulted


LBJ and the Polls

in a report recommending legislation to stop the widespread discrimi-
nation against black voters that they had found. Most shocking to
the public was the March 7 police violence against marchers in Selma,
Alabama. Johnson realized that public opinion favored legislation but
that "it would probably not take long for these aroused emotions
to melt away."20 This intuitive judgment led him to take quick action.
On March 15 he spoke in favor of federal voting rights legislation
before a joint session of Congress, dramatically ending his speech
with the civil rights movement's anthem, "We Shall Overcome." Press
reaction was extremely favorable, one account stating, "Johnson was
never more powerful," another that he "swept up a packed audi-
ence,"21 but Johnson asked Quayle to take a quick survey to gauge
the reaction of the general public. Two days after the speech, 100
people in New York City and suburban Westchester who had been
interviewed in a previous survey were contacted again. The result
was the best Quayle had ever found: 87 percent of those familiar
approved compared to only 3 percent disapproving. 90 percent fa-
vored the proposed law that Johnson officially sent to Congress the
day of the interviews.22
Although such results must have encouraged Johnson, the need
for a quick poll caused questionable methods, such as the limited
geographical area covered (likely to be more favorable than most on
this issue) and the small number of interviews. The results could
hardly be considered a reliable measure of public opinion even in
New York, let alone the nation. As the bill was pending, additional
polls were taken. Quayle's April survey of Columbia, South Carolina,
indicated that a voting rights law "may not be an anathema" in all
of the South; a 48-43 percent margin was in favor even though three-
fourths of those questioned opposed integration. In Kentucky, civil
rights was not an important issue for most voters.23
Despite providing only the most fragmentary evidence of public
opinion, these polls at least suggested that opposition to Johnson's
bill was weaker than on other civil rights issues such as open housing
or school integration. The Alabama marchers completed their dem-
onstrations under federal protection while the administration pressed
its proposals in Congress. With both Senate party leaders among the
66 cosponsors, passage was simply a matter of time. The bill passed
the Senate on May 25, the House on July 9. After a disagreement
between the two houses over banning not only federal but also state


"Johnson Nice"

poll taxes was resolved, the president signed the bill into law on Au-
gust 6. The rapidity of events had caused Johnson to speed up his
original timetable based on his own judgment of public opinion. Polls
taken during the campaign had shown him that voting rights was
the least controversial reform, but there was no real follow-up. When
events caused Johnson to take some action, only a few unsystematic
measurements of public opinion were taken, but they confirmed the
correctness of his decision. Once more, polls freed Johnson to do
what he wanted to do. However, we can see how the decision to save
money by relying on piggy-backed polls deprived Johnson of system-
atic data on the views of the nation: surveys were taken only in those
geographical areas others were interested in, and there was no de-
tailed probing of the opinions of those questioned.
The passage of the Voting Rights Act slowed momentum for fur-
ther legislation. Quayle would do considerable polling for the 1966
elections but little during the rest of 1965. A series of polls in Ala-
bama gave Johnson his worst ratings in the nation, almost entirely
due to civil rights. In Georgia, Johnson's ratings were nearly as bad:
despite a 9 percent gain on civil rights, his policies were still opposed
by 64 percent of those polled. Polls in a few northern states, on the
other hand, showed approval for Johnson's civil rights policies, even
if that approval often was not overwhelming.24
These results presented Johnson with a dilemma: declining sup-
port in the North and continuing opposition in the South indicated
that the climate for civil rights initiatives was no longer favorable,
especially since the least controversial proposals had been enacted,
but if nothing was done in 1966 the climate would undoubtedly
worsen in later years. The huge Democratic majority elected in 1964
would certainly shrink in the next Congress; the Vietnam War was
draining the president's energies and budget; and blacks in the cities
were losing patience, precipitating riots. Additional pressure on the
president to act resulted from Alabama's failure to convict anyone
for the murder of civil rights marcher Viola Liuzzo. Johnson's actions
reflected these cross-pressures. His State of the Union address pro-
posed laws to prevent discrimination in jury selection, to protect
against violence meant to deprive people of their civil rights, and
to ban discrimination in the sale and rental of housing. However,
with Congress less receptive than it had been for his previous propos-
als, Johnson did not press as hard, failing to send a legislative mes-


LBJ and the Polls

sage to Congress until the end of April. According to Harvey, "No
doubt he was thinking about the mid-term congressional elections
and the growing white backlash." Harvey cites Gallup and Harris
polls showing that a majority of whites thought Johnson was pushing
for integration too quickly. In contrast, Katzenbach argued that open
housing was of little importance to most voters, basing his arguments
not on polls but on the findings of members of Congress who "tested
it out with the public and it was really surprising."25 Interestingly
enough, Quayle's polls provided evidence for both views, showing
that large majorities opposed open housing laws but that it was not
a leading issue.
The bill turned out to be Johnson's first major civil rights failure.
Pressures from the real estate industry resulted in the exemption of
60 percent of housing. To gain support for even this diluted measure
Johnson had to agree to include an amendment against rioting. Even
though these compromises caused a debate among black leaders over
whether the measure was still worth supporting, they were not
enough to gain Dirksen's support to stop a filibuster. By September,
the bill had effectively died.26
Quayle's 1966 polls continued to show a deterioration in support
for Johnson's civil rights policies.27 Thus, the reasoning behind John-
son's statement at the beginning of this section is clear, but he did
make some effort in 1967, using his State of the Union address to
repeat his support for the previous year's proposals. However, the
reduced Democratic congressional majority, declining public support,
and Johnson's growing preoccupation with Vietnam led to a less
strenuous effort that resulted in yet another failure. Public support
for civil rights had turned into concern about urban riots. So serious
had the situation become that a Connecticut poll taken for the White
House by Joseph Napolitan underrepresented blacks "because we
had extreme difficulty getting interviewers to go into Negro areas,"
a problem the pollster said was becoming widespread.28 Several Dem-
ocratic politicians expressed their concern by sending the White
House polls showing how poorly their constituents rated Johnson
on civil rights. Johnson himself had concluded that the issue of open
housing "had become a Democratic liability."29
Fred Panzer suggested several times during late summer that polls
indicated Johnson should shift to other issues. In one such memo
he pointed to a forthcoming Gallup Poll finding that blacks and


"Johnson Nice"

whites agreed on "four ways to deal with racial problems"-compul-
sory youth training programs, gun control, a curfew for youths under
age sixteen, and interracial councils. Panzer's clincher was, "Senator
Bobby Kennedy is riding the gun control issue very hard. He's on
the front page of today's New York Times as a result." The day he
received this, Johnson sent a note to Marvin Watson telling him to
"get letters to the leaders on gun control legislation."30 However, not
much was done to follow up.
Such discouraging advice was not enough to stop Johnson from
again pushing for civil rights legislation in his 1968 State of the
Union address. Johnson's persistence was due largely to the strong
lobbying of black leaders, especially Clarence Mitchell of the NAACP,
a constituency he could not afford to alienate. This time the proposals
were separated in order to ease the passage of the least controversial.
A federal jury bill was quickly approved, but open housing would
not be as easy. For reasons still not entirely clear, Dirksen changed
his mind at the end of February, resulting in quick Senate passage."
The bill stalled in the House, but Johnson's withdrawal from the pres-
idential race and the King assassination gave it renewed momentum.
A compromise providing a three-stage phase-in over two years but
covering 80 percent of the nation's housing was signed into law a
week after the assassination.
As for the other side of the race issue, urban riots, Johnson had
appointed a Commission on Civil Disorders in July 1967. In March
1968 the commission recommended large-scale job creation and train-
ing programs, greater efforts to desegregate schools, increased
spending on educating disadvantaged students, increased welfare
payments, and considerably more government spending on low- and
moderate-income housing. For the commission, "The only possible
choice for America" was "a policy which combines ghetto enrichment
with programs designed to encourage integration of substantial num-
bers of Negroes into the society outside the ghetto." Faced with a
Bureau of the Budget estimate that these programs would cost $30
billion, Johnson, who was already asking Congress for a tax increase
to pay for the Vietnam War while resisting congressional pressures
to cut domestic spending, decided to ignore them.32
Harvey's evaluation of Johnson's record reflects the most wide-
spread view. "Overall more gains were made in desegregating some
of America's major institutions during the Johnson administration


LBJ and the Polls

than in any previous ones," he wrote, but, especially in the later
stages, inadequate funding and enforcement had weakened the ef-
fort. But his explanation that "one can only speculate that he gave
in to southern pressures in return for support of an increasingly
unpopular war" does not satisfy. The Vietnam War is crucial but not
simply because of the need to maintain southern support. As Bruce
Miroff has written, "Since Johnson's preoccupation with the Vietnam
War continued to mount, his involvement in the travails of racial poli-
tics diminished even further after 1966." The cost of the war and
Johnson's commitment to it drained his energy and his budget's
money away from civil rights. The war also alienated many liberals,
a strong civil rights constituency, causing them to be wary of any
proposal coming from the White House.33
Equally important is the role of public opinion. Johnson's initia-
tives of 1964 and 1965, especially his attacks on job and voting dis-
crimination, had widespread public support. Later proposals were
less welcome, and continuing urban riots eroded public support still
further. Polls showing this decline confirmed Johnson's instinctive
political judgment. Harvey also criticizes weaknesses in enforcing the
law, but this lack has been a chronic problem for presidents in many
areas of public policy. The public may clamor for new laws, but it
pays little attention to the details of their implementation, with the
exception of the people directly affected by those details. Thomas
Cronin put it, "The administrative side of the presidency is the least
glamorous and least envied aspect for presidents and public alike.
. . After a time a president usually chooses to spend his scarce politi-
cal capital on new policy initiatives rather than in implemental or
management strategies."34 Thus, it should come as no surprise that
Johnson's polls asked no direct questions about the implementation
of initiatives already passed.


Proposals for government health insurance in the United States have
a long history, but, until Medicare, the opposition of the American
Medical Association had prevented their enactment into law. When
Medicare was first introduced in Congress in 1952, President
Eisenhower was against it, as was Senator Johnson." Senator John
Kennedy, however, was a strong supporter, and he made it an issue


"Johnson Nice"

in his 1960 presidential campaign. By then the inadequacy of private
health care, the huge jump in hospital costs, the growing number
of older citizens, and the low income of many of them had strength-
ened the case for Medicare, which, according to published polls, was
supported by two-thirds of the public.3 Johnson had also begun to
change his position, evolving into a strong Medicare supporter by
the time he became president. In 1962 the Senate narrowly defeated
Kennedy's proposal. A Gallup Poll taken during the debate asked
a series of questions to determine what people knew about Kennedy's
plan, then asked them to choose between it and one based on private
insurance. This format reduced Medicare support to only 44 percent,
with 40 percent backing the private insurance plan preferred by the
Johnson, however, continued to back Medicare strongly. It was an-
other way to show continuity while pressing a program he wanted.
In his first speech to Congress, he asked enactment of "the dream
of care for the elderly," then told a press conference, "I can think
of no piece of legislation that I would be happier to approve." Al-
though the Senate attached a Medicare rider to the Social Security
bill, the conference committee was unable to agree, preventing pas-
sage in 1964.
Polls taken during the election campaign showed growing support
for Medicare. A September survey in Indiana showed a 13 percent
increase, but Quayle's question-"help for older people"-was
phrased too generally to be useful. He frequently used this wording,
then classified it as support for Medicare in his analysis, but no one
in the White House commented.37
Other polls painted a clearer picture of how much the public
wanted Medicare: 85 percent in New York, 82 percent in Pennsylva-
nia, and 73 percent in Michigan were willing to pay additional Social
Security taxes for it. Even in more conservative North Carolina 74
percent would pay more. Johnson's memoirs point to the many letters
he received during 1964 and 1965 which "described in terms of
human suffering the serious need for a bill to assure medical care
for the elderly. The volume of that mail made it clear that Medicare's
day, at long last, had dawned.""3 With such overwhelming support,
Johnson pressed the issue in his campaign.
Johnson's landslide not only swept into office large Democratic ma-
jorities in both houses of Congress; it also resulted in the defeat of


LBJ and the Polls

several prominent AMA supporters. Richard Harris estimated that
Medicare gained 4 Senate and 44 House votes in the election.9 In
his 1965 State of the Union address, Johnson declared that the pas-
sage of Medicare was his first priority. To counter the obvious support
for Medicare, the AMA and a group of House Republicans each an-
nounced alternative plans, which, they argued, corrected deficiencies
in Johnson's proposal. In a brilliant stroke, Ways and Means Commit-
tee Chair Wilbur Mills announced a new plan including all three
suggestions. Despite the additional cost, Johnson immediately agreed.
Appearing on television to describe the new bill, Johnson introduced
a group of Democratic congressional leaders along with the bill's
cosponsors. Turning to the Senate Finance Committee chair, conser-
vative Harry Byrd, he asked his view. A surprised Byrd had little
to say, so when Johnson asked whether Byrd had any objections to
quick hearings, the committee chair had to agree to prompt consider-
Quayle continued to find a public consensus behind Medicare.40
The combination of public opinion, a strong Democratic congres-
sional majority, andJohnson's persuasiveness led to the passage of Medi-
care, which was signed into law on July 30. After enactment, Quayle
continued to ask about "help for older people." The president's con-
tinuing high ratings were a welcome contrast to declines on Vietnam
and civil rights. As Johnson's overall popularity dipped, in some states
Medicare was the only issue on which he received a positive rating.4'
As with civil rights, polls did not provide the initial impetus for
Johnson's Medicare proposals. By showing him overwhelming sup-
port as well as a willingness by the public to pay for Medicare, polls
again freed Johnson to do what he wanted. Given the strong opposi-
tion of the AMA and the potential for doubt caused by the 1962
Gallup Poll, this support was quite significant. Beyond measuring
the strength of the public's support, however, little use was made of
polls. Once Medicare was passed, the polls only asked what the public
thought of the president's record on the issue in general. No attempt
was made to explore this issue more deeply, either to see the specifics
of what the public liked or to discover whether it favored changes
in Medicare or additional initiatives to help older people. Unlike Viet-
nam, the economy, or civil rights, Johnson's ratings on this issue re-
mained high throughout his presidency, so there was no danger hid-


"Johnson Nice"

den within the polls. Whether there was a missed opportunity will
remain unknown.

The Great Society

As important as the civil rights bills and Medicare were, Johnson's
ambitions for social reform were far greater. Beginning in March
1964, Goodwin inserted "Great Society" into Johnson's speeches. A
May college commencement was Johnson's first use of it as a main
theme, stressing that "an end to poverty and injustice" would be "just
the beginning." Doris Kearns pointed out the widespread public sup-
port for these ideas, writing that "opinion polls confirmed what most
men already knew: the people were pleased with their president and
shared his confidence in the almost limitless capacity of the American
nation." She emphasized that "all his achievements depended upon
this essential harmony between his acts and popular desires; that
without that all his skills and energies would have been futile."42
The main element of the Great Society was the War on Poverty.43
Before his assassination, Kennedy had asked aides to develop an anti-
poverty program, but no' specific plan had emerged before Johnson
took office. On his first day, Johnson was briefed about the program
by the head of the Council of Economic Advisers, Walter Heller. Re-
sponding enthusiastically, Johnson told Heller, "That's my kind of
program; I'll find money for it one way or another." As with civil
rights and Medicare, it provided continuity with the Kennedy admin-
istration and a chance for Johnson to demonstrate his liberal creden-
tials, but, in contrast to the other two, there were no Kennedy propos-
als pending in Congress so Johnson could present it as his own
program. He transformed the original conception of a few pilot proj-
ects into a bold nationwide attack on poverty for which the war meta-
phor was perfect.
By stressing the poverty issue at press conferences and speeches,
Johnson was able to focus media attention on poor people, leading
to a spate of stories exploring their problems. In his first State of
the Union address, he declared, "This administration, here and now,
declares unconditional war on poverty," but he also emphasized that
his program, by reducing the dependence of the poor on welfare,
would eventually benefit all Americans. Finding that the Bureau of


LBJ and the Polls

the Budget's proposals lacked the desired bold approach, in early
February he set up a task force under R. Sargent Shriver, who later
became the first director of the Office of Economic Opportunity, the
agency established to run the antipoverty program.
These plans were developed so quickly that there were no private
polls taken. When Eric Goldman was asked to survey intellectuals,
he found them tremendously enthusiastic.4 On March 16, Johnson
formally proposed the Economic Opportunity Act to Congress. Its
main features were the establishment of the OEO and a domestic
version of the Peace Corps, training and educational programs for
young people, community-based programs, and an attack on the
causes of poverty with a total cost deliberately estimated at just under
$1 billion.
Early polls indicated that although other issues were considered
more important, the public favored an antipoverty program. How-
ever, none of these polls asked more specific questions.45 After send-
ing his proposals to Congress, Johnson threw his energy into increas-
ing the public's concern for the poor and translating it into legislative
action. He traveled to areas of widespread poverty, sent letters to
virtually every important interest group, made a series of speeches
before a variety of audiences, and sent cabinet and other officials
around the country to discuss the issue. Organized labor was a partic-
ular target. Republicans tried, with little success, to attack the bill
as an election year gimmick. Barry Goldwater argued that "most peo-
ple who have no skill, have no education for the same reason-low
intelligence or low ambition."46
Of particular concern to Johnson were southern Democrats who
worried that many of the benefits would go to poor blacks and that
training facilities would be integrated. Johnson persuaded Georgia
Representative Phil Landrum to sponsor the bill in place of the nor-
mal sponsor, the flamboyant black chair of the Education and Labor
Committee, Adam Powell. A number of southern objections were
met by such compromises as adding a loyalty oath for Job Corps par-
ticipants and promising that liberal Adam Yarmolinsky would not
be involved in the new agency. The bill passed quickly and was signed
into law on August 20, 1964.
During the campaign, Quayle continued to ask about the War on
Poverty. New York was polled twice, using particularly large samples


"Johnson Nice"

of 1,200 or more each time. Both surveys showed more than 60 per-
cent in favor. Other states showed comparable or even more sup-
port-67 percent in Pennsylvania, 56 percent in West Virginia, and
74 percent in Kentucky. Once more, however, there was no additional
questioning about specific aspects of the antipoverty program.47 After
the election Johnson's legislative agenda was so full that antipoverty
efforts concentrated more on organizing OEO than on additional
Federal aid for education was the last of Kennedy's legislative initi-
atives tackled by Johnson. Campaign polls had shown that the public
wanted it,48 but the problem Kennedy had been unable to solve was
how to aid religious schools without violating the constitutional sepa-
ration of church and state. Kennedy's Catholic faith had complicated
the situation, a problem that the new president did not have. Johnson
wrote that he had three options: ignore the issue, try to pass
Kennedy's bill, or negotiate a compromise. He claims that he rejected
advice from members of Congress and some of his aides favoring
the first option because of his own experience as a schoolteacher and
the extra flexibility his Protestantism afforded him. He failed to men-
tion the significant support showed by his polls, but he did cite the
critical need to improve education and the lack of state and local
money to do so. Proposing the same legislation as Kennedy would
be "self-defeating," so he named a bipartisan task force to examine
the problem. It emphasized the urgent need for federal aid. Finally,
Johnson found a compromise formula that would indirectly aid pri-
vate schools by providing funds for their students. Before sending
it to Congress, he made certain it was acceptable to the major contend-
ing groups. Once this was done, a special message proposing $1.5
billion in federal aid to education was sent to Congress on January
12, 1965. In another one of Johnson's legislative miracles, it passed
in less than ninety days.49 Quayle was asked to see what the public
thought after passage. As with the War on Poverty, he found strong
With the Vietnam War draining resources away from domestic ini-
tiatives, the primary use of polls was occasionally to monitor the pub-
lic's views of existing programs. In Newark, New Jersey, a city proba-
bly selected because of its racial tensions, 88 percent favored the War
on Poverty. The White House also received a poll taken by Lloyd


LBJ and the Polls

Free to measure support for the welfare state. It found public support
for specific programs if they could be justified in "programmatic and
moral" terms based on human needs, but, paradoxically, most Ameri-
cans also agreed that all those able to work could find jobs and that
"the relief rolls are loaded with chiselers and people who just don't
want to work." Free's advice was to be careful about phrasing, avoid-
ing terms such as welfare, while working to educate the public about
the problems of the poor.5" By this time, however, the White House
was so preoccupied with the Vietnam War and civil rights and eco-
nomic problems that it lacked the time, resources, and energy to em-
bark on such a major educational effort. Free's poll was one of the
few that explored an issue other than Vietnam in depth, yet it went
largely unused, another example of the failure to utilize polls to their
full potential.
Rather than proposing new Great Society programs, Johnson's
main efforts were aimed at retaining those already enacted. With
the War on Poverty needing congressional extension, he defended
it in his 1967 State of the Union Message but also agreed to make
changes in light of experience. The White House would have been
better able to judge what changes to make if they had surveys of
public opinion on particular programs. The only ones available, how-
ever, came from agencies which had an interest in maintaining or
even expanding their owh programs. For example, William Crook,
the director of VISTA (the domestic version of the Peace Corps),
sentJohnson an OEO-commissioned Gallup survey, "Attitudes of Col-
lege Students toward VISTA," which showed that 90 percent had
heard about the agency and 52 percent were interested in joining.
Crook optimistically projected these percentages to mean three mil-
lion volunteers. Overall OEO spent about $200,000 for polls taken
by such leading pollsters as Gallup, Harris, and Daniel Yankelovich.52
Despite such efforts, Johnson was forced to accept weakening amend-
ments to gain the needed congressional support, resulting in passage
in November.
Because Johnson's polls indicated that the public still supported
the War on Poverty, he might have been able to salvage more than
he did. However, gaining financial and political backing for the Viet-
nam War was more important to him. Polls told him what he could
have done, but he chose another route. The economy would not per-
mit both guns and butter, as we will see in chapter 4.


"Johnson Nice"

Four-Year House Terms

A more short-lived issue was Johnson's proposal to extend House
terms to four years. Its presentation in his 1966 State of the Union
address was greeted with surprise and enthusiasm, according to one
account "by far the heaviest and most heartfelt applause of the eve-
ning."53 Although the proposal had been pending in Congress for
more than a year without a committee hearing, Johnson had never
indicated his support. What is particularly intriguing is that a Gallup
Poll had been taken a month before the speech showing 61 to 24
percent approval of a four-year term. These results were released
two days after the speech. Johnson initialed a memo with these results
the day before they were released, making it unlikely that he had
learned of them before his speech, but the coincidence is striking.
None of Quayle's polls had asked about the issue. Unfortunately for
Johnson, congressional hearings indicated considerable opposition
and little public enthusiasm, so the proposal was allowed to fade
away." In this case the Gallup Poll had failed to reveal the unimpor-
tance of the issue to most Americans. The intensity of the opposition
was far more important than the numerical advantage of the luke-
warm supporters.


While an exhaustive examination of all of Johnson's initiatives is be-
yond the scope of this book, the discussion provides enough evidence
to evaluate Johnson's use of polls. At the beginning of this chapter
a number of possible uses of polls were suggested: measuring the
popularity of the president either in general or on specific issues;
demonstrating that popularity to other political leaders whose sup-
port the president needs; determining which issues the public is most
concerned about; learning which proposals have the greatest (or least)
public support; seeing which arguments are working best; and find-
ing out which constituency groups can be best mobilized to support
the president's program and which are the most strongly opposed.
In this chapter we have shown that Johnson's use of polls was domi-
nated by the first two. Nearly all of the polls taken devoted most
of their questions to presidential ratings. During the first two years
of his administration, these ratings showed great popularity for John-


LBJ and the Polls

son and general support for his programs. Because the public ap-
proved of the direction the administration was taking, Johnson could
use the polls to justify his initiatives, giving himself a free hand even
when there were powerful interest groups in opposition, such as the
AMA on Medicare. Thus, the hope of the early advocates of polls,
discussed in the introduction-that polls would allow presidents to
distinguish between pressure and opinion-appear justified. This,
however, is only partly true, as will be demonstrated in later chapters.
Johnson used polls to make this distinction only on issues where he
disagreed with that pressure. When his popularity declined later in
his term, he used polls not to learn of public opinion but to try to
manipulate both that opinion and the perception of it. In other
words, polls served not so much as information to help the president
make appropriate decisions but as a weapon to back up decisions
already made. During the period of legislative initiatives, Johnson's
popularity, as measured by polls, gave him the freedom to do what
he wanted.
V. O. Key has described three types of consensus: permissive, sup-
portive, and decisive.55 In the most common, the permissive type, the
public generally supports a proposed policy but lacks strong feelings.
A supportive consensus backs current policy as well as those making
it but suggests that there are certain boundaries, often not clearly
defined, that should not be crossed. The decisive consensus, the rarest
of the three, occurs when a strong majority demands governmental
action to achieve a particular policy. In contrast to these consensus
situations is one of conflict where the public is too divided to support
any particular policy.
What polls did for Johnson was to blur the distinctions among
the first three consensus situations. Johnson's polls rarely measured
intensity of feelings or probed for the details of what the public sup-
ported or opposed. While it is true that the public is often too igno-
rant of the details of an issue to make an informed choice, probing
questions in a survey could highlight this ignorance and show
whether providing more information to the public would be a strat-
egy worth pursuing. Instead, Johnson's approach allowed him to pro-
vide the evidence that the public supported his legislative initiatives
without quibbling over the details. As long as his popularity remained
high and he did not cross the boundaries of the supportive consensus,
this approach worked. Unfortunately, it also overlooked the warning


"Johnson Nice"

signs when he was approaching the boundaries of consensus. Thus,
polls were not, as their proponents had hoped, able to provide the
scientific evidence that would convince Johnson that his policies had
lost public support. The remainder of this book will be an examina-
tion of how and why this phenomenon occurs.
Uses of polls other than measuring popularity were sporadic at
best. Quayle's surveys did ask respondents to rate the importance of
issues, but not much was done with this information. The fourth
use-measuring the proposals that have the most support-was done
for civil rights but only to back decisions already made, as when John-
son decided to press for the more popular voting rights legislation
before open housing. Medicare provides an example of the fifth use,
when polls showed overwhelming support for the program even if
it was necessary to pay higher Social Security taxes. It was possible
to ignore arguments that the program would be too costly since the
public had made it clear that it was willing to pay the cost. Once
Medicare was enacted into law, however, survey questions about it
became so general as to be of little use. As for constituency groups,
the way was eased for passage of the Voting Rights Act because it
was shown that opposition was less vehement in the South than many
expected. Quayle's backlash-frontlash measurements also showed
that, at least for the first two years of the administration, there was
little to fear from a white backlash. On the more positive side, polls
seem to have played little part in showing which groups were most
supportive of Johnson's policies and mobilizing them.
The irregularity of these uses shows how overwhelming was John-
son's need to demonstrate public approval. Sometimes this need led
to questionable techniques, such as in the survey of reaction to his
congressional voting rights speech. It limited the depth of polls be-
cause Johnson needed the fact of public approval more than the de-
tails or explanations. Unfortunately, this lack of attention to details
would prove a key element of his downfall.




WHEN HE FIRST BECAME PRESIDENT, LyndonJohnson could hardly
have imagined how much Vietnam would come to dominate his presi-
dency. During the 1964 campaign, Vietnam presented little threat
to his landslide. Oliver Quayle's polls showed that, despite potential
opposition to administration policies, foreign policy issues were not
of great importance to most voters.' Johnson's task was made rela-
tively easy by Goldwater's image as a threat to peace. As with other
issues, the president was able to remain statesmanlike by stressing
unexceptionable generalities, such as peace coupled with firmness.
The Gulf of Tonkin incidents in late August, together with over-
whelming congressional approval of the administration's resolution,
muted Goldwater's criticism that the president was not tough enough.
At the same time Johnson's speeches stressed his opposition to escala-
tion. Less noticed at the time were the qualifications such as "for
the moment" or "at this stage of the game" that allowed him eventu-
ally to approve the escalation that so many voters thought he had
ruled out.2
In this chapter we will examine the role played by polls in formu-
lating and implementing Vietnam policy. Because the story ends in
March 1968 with Johnson simultaneously announcing his decisions
to de-escalate and to withdraw from running for reelection, the most
important question is what the polls told Johnson about the potential
and actual dangers he faced in maintaining public support. In the


introduction, we discussed the hopes of some that polls could provide
a way of overcoming presidential isolation by informing a president
that he and his policies lacked support. Clearly this did not happen
for Johnson's Vietnam policy, at least not until it was too late. Was
it a failure of the polls? Johnson himself? His advisors? Or is there
a deeper problem that polls alone cannot solve?
To answer these questions we must divide Vietnam policy into three
stages: the early decisions to escalate made between February and
July 1965; the slow expansion from late 1965 until early 1968, cou-
pled with administration efforts to maintain the support of the public,
the press, and Congress; and the March 1968 decision to de-escalate
coupled with Johnson's withdrawal. During the first stage, the crucial
question to be examined will be whether the polls did or should have
contained warnings of potential danger and, if they did, why they
had so little impact. Since the second stage included a steady decline
in Johnson's popularity, the question is what impact this decline had
on Vietnam policy and why it took so long to change that policy.
In the third stage, the focus will be on why the* sudden shift took
place and what role polls played.

Early Escalation

As explained in chapter 2, there were few Quayle surveys taken dur-
ing early 1965, when crucial decisions were made on regular bombing
of North Vietnam and sending, then sharply increasing, American
ground combat forces in South Vietnam, but other polls were avail-
able to the president. The public polls frequently asked about Viet-
nam; other pollsters took surveys for the White House; and friends
and supporters forwarded polls. Johnson had no shortage of infor-
mation about the views of the public.
When Johnson took office, South Vietnam's president, Ngo Dinh
Diem, had just been overthrown and killed. The resulting period
of government instability led Johnson to conclude that the situation
in Vietnam "was considerably more serious than earlier reports had
indicated." Secretary of Defense Robert McNamara was dispatched
to Saigon in December 1963 to evaluate the situation. He reported
that immediate action was necessary to prevent a communist victory.
Johnson expanded the role of American military advisors but did
not approve the use of additional troops.3 This pattern was to be


LBJ and the Polls

repeated throughout the next year. Johnson's advisors, with little dis-
sent, would depict a worsening situation requiring an increased role,
but Johnson, while approving modest changes, would defer major
decisions. In March 1964, Johnson cabled Ambassador Henry Cabot
Lodge that he was putting off military action against North Vietnam
until the South Vietnamese government had stabilized, despite an-
other pessimistic report by McNamara and Maxwell Taylor.4 In June,
a new team sent to Vietnam replaced Lodge with Taylor as ambassa-
dor and made General William Westmoreland the new commander
of the American forces.
At about this time the resolution of support that was to be intro-
duced after the Tonkin Gulf incidents was proposed, but Johnson
again postponed action. Stanley Karnow argues that the major rea-
sons for this deferral were lack of a specific act by North Vietnam
to justify retaliation and the fear of alarming voters. George Kahin
also reported a consensus of Johnson's advisors that bombing the
North would require building public support, not best accomplished
during an election campaign.5 Johnson wished to do as little as possi-
ble before the election, and poll results showing that most Americans
paid little attention to Vietnam again helped free him to pursue his
desired course of action. As Doris Kearns has put it, the public
showed little interest in events in Vietnam, and "Johnson wanted to
keep it that way."6
Only the Gulf of Tonkin incidents changed this strategy. Despite
considerable debate over how serious these attacks on American ships
by North Vietnam were, or whether they even occurred,7 the adminis-
tration's resolution authorizing "all necessary measures to repel any
armed attack against the forces of the United States and to prevent
any further aggression" quickly passed both houses of Congress with
only two dissenting votes in the Senate. Together with its accompany-
ing military measures, this resolution effectively defused Goldwater's
charges that Johnson's foreign policy had not been firm enough,
charges that some of Johnson's polls suggested were hurting him,
albeit not very much.8 Again, however, Johnson resisted the urging
of advisors who wanted more retaliatory bombing. He also instructed
his aides to de-emphasize the incidents and called the Soviets on the
hot line to stress his desire not to widen the war.9 Once the election
was won, the hard decisions that had been postponed had to be made.
Judging what factors cause a president to make an important decision



is always difficult, but Johnson's style exacerbates the analytical prob-
lem. Larry Berman wrote that "Johnson's style incorporated a preoc-
cupation with secrecy, a need for controlling information both to and
from the White House, an extreme tendency to overreact to written
criticism, an emphasis on consensus and team play, and a tight rein
on the White House staff."'0 The record may seem to show Johnson
asking questions, receiving advice, then making up his mind, but it
may in fact be showing him manipulating the process to get his aides
to suggest the policy he had already decided on. Certainly we have
seen this pattern in his use of polls. It also begins to indicate why
the information that Johnson did not want to hear, even if couched
in the apparent objectivity of survey research, had so little effect.
Kearns argues that Dean Rusk, McGeorge Bundy, McNamara, and
Taylor were the strongest influences on Johnson because they shared
his assumption that Southeast Asia provided a crucial test of Ameri-
ca's ability to resist communism. Eric Goldman wrote that "Lyndon
Johnson became President without a foreign policy" but that he had
"a set of assumptions, attitudes and prejudices which, together with
his personality, amounted in practice to a foreign policy." He identi-
fied these as a belief that America was often misunderstood abroad
and because of this belief, it had to counter foreign aggression
quickly; a belief that communists were repeating the type of aggres-
sion begun by the Nazis; and concern for congressional approval of
foreign policy while he retained control of that policy himself. John-
son expected more problems from hawks than doves, "but whatever
the facts were, LBJ's judgment of public opinion played its role in
shaping his policy toward Vietnam." Johnson himself cited the exper-
iences of World War II and Korea to justify the main goal of Ameri-
can policy, helping South Vietnam "win their contest against the ex-
ternally directed and supported communist conspiracy.""
After the election, Johnson set up a National Security Council
Working Group to consider policy options.12 Three options were put
forward: a continuation of present policy with limited bombing repri-
sals; graduated military moves, starting with increased bombing; and
dramatic escalation, introducing ground troops and stepping up the
bombing. Debate over the next months centered on these options
with nearly all advisors favoring either dramatic or slower escalation.
Only Vice President Humphrey, NSC staffer James Thomson, and
Undersecretary of State George Ball opposed escalation. When Hum-


LBJ and the Polls

phrey and Thomson were removed from the debate, the message
to other advisors was clear. According to Berman, "Ball never had
a chance" because the burden of proof was always put on him as
the dissenter from the consensus.13 Given Johnson's opinions about
containing communist aggression, there was no way that Ball could
meet this challenge.
With no Quayle polls available, Louis Harris was asked to conduct
a series of surveys on Vietnam. According to Moyers, polls were taken
in November 1964 and January and February 1965, with Harris ask-
ing "that we not let anyone know he did it for us." Harris concluded
that "the American people feel that China is testing our will. The
majority favor the action the president has taken. They do not believe
we can walk away from South Vietnam." He did not mention the
relatively narrow margin of 48-40 percent favoring "sending a large
number of American troops to help save South Vietnam" or the 69
percent who agreed that fighting in Vietnam was "not very impor-
tant." On the other hand, someone who read the memorandum (pos-
sibly Johnson, to whom it was sent on February 16) circled the 83-5
percent margin by which the public thought it "more right than
wrong to bomb North Vietnam than it was earlier." The February
survey, from which this response was taken, was particularly impor-
tant because it was completed shortly after reprisal bombings of
North Vietnam. A Gallup Poll, released February 15, had shown
overwhelming support for these reprisals as well as a 3 percent in-
crease in Johnson's popularity rating (to 69 percent) and 60 percent
approval of his "handling of the situation in Vietnam."" The danger
signals were there, even if it took a careful look at the data to find
them, but neither Harris nor any of Johnson's advisors pointed them
out. Johnson got the message he was looking for, that the public
strongly backed a policy of escalation.
In addition to these ambiguities in public support for escalation,
Harris's surveys also showed that 75 percent favored negotiations
"to settle the war in Vietnam." Important figures such as France's
President Charles de Gaulle, U.N. Secretary-General U Thant, the
Senate majority leader, Mike Mansfield, and columnist Walter Lipp-
mann were urging negotiations which most White House advisors
opposed even discussing. Instead, a public relations effort was
mounted to demonstrate that the cause of the war was North Viet-
namese aggression.15 At the same time, Johnson did not wish to alarm



the public, so decisions to escalate were announced as quietly as possi-
ble. Such contradictions would eventually cause serious trouble.
At the time Johnson feared hawks more than doves. Criticism from
Richard Nixon and the China lobby had reminded him of Republican
charges that President Harry Truman had "lost" China. So seriously
did he fear a repetition that he later told Kearns, "Losing the Great
Society was a terrible thought, but not so terrible as the thought of
being responsible for America's losing a war to the Communists.
Nothing could possibly be worse than that."'6 In March and April,
Quayle's polls began to arrive, providing confirmation for Johnson's
fears. An early March survey in New York City showed 59 percent
disapproval of Vietnam policy with "more potential 'hawks' than
doves,'" warned Quayle. At the end of March, Quayle sent a second
poll that was primarily about a special election for Congress in South
Carolina but also included a short separate report for Johnson on
national issues to help him judge the political mood of the South.
On Vietnam, this mood was hawkish, with 95 percent against with-
drawal and 83 percent agreeing that "we should step up our military
effort even more and win the war." However, even among these
hawks, 88 percent opposed bombing China and 66 percent were
against bombing Hanoi. Nor was this district particularly fertile terri-
tory for Johnson, who, despite his national 1964 landslide, received
only 34 percent of its vote." Like Harris's surveys and the public
polls, these not only contained data to support Johnson's preconcep-
tions; they also pointed to considerable public confusion and ambigu-
ity that could lead to several, possibly conflicting, interpretations. Un-
fortunately, these ambiguities were largely ignored by the polling
reports as well as by Johnson and his advisors.
Johnson's eventual decision was to endorse the middle course
among the three options presented to him. After the American bar-
racks in Pleiku were attacked in early February, he angrily told his
advisors, "I've had enough of this," announcing his intention to take
strong measures.18 Within a few days he decided upon more sustained
bombing, but the situation continued to deteriorate, leading to the
approval of Westmoreland's request for two battalions of Marines
to guard the Danang air base. By early April, the use of an additional
20,000 Marines, now authorized to take offensive action, had been
approved. This policy of gradual escalation, according to Kearns, was
a "predictable choice" because it "represented the moderate path be-


LBJ and the Polls

tween the competing extremes of massive destruction and total with-
drawal."'1 In his memoirs, Johnson frequently portrays himself as
taking this middle course. Of the three options presented by the
working group, he had chosen the one in the middle. As for the
April troop increase, he compromised between the military who
"wanted to move fast and in strength" and "a few who opposed any
significant involvement in the ground war" by agreeing with those
advisors "who thought we should proceed, but more deliberately."20
Polls helped assure Johnson that the public supported these deci-
sions. His high approval ratings combined with support for strong
military measures meant the public would support escalation. On the
other hand, its simultaneous support for negotiations backed what
he saw as his restraint in rejecting even stronger military action. The
main reasons for Johnson's decisions were his views of the world and
of events in Vietnam. The use of polls followed the now well-
established pattern, demonstrating the public support needed to free
him to carry out policies he had already decided on.
A series of escalations followed during the next few months, most
notably the late July decision to increase the number of American
troops in Vietnam to 180,000-more than double-by the end of
1965, with possibly another 100,000 in 1966. Johnson again saw this
as a middle course, having rejected the "extremes" of withdrawal and
a virtual war footing which would mean asking Congress for "great
sums of money," calling up the reserves, increasing draft calls, and
declaring a state of emergency. When Johnson asked his military advi-
sors whether the public would support escalation, Army Secretary
Stanley Resor replied, "The Gallup Poll shows people are basically
behind our commitment." As Herbert Schandler describes the
decision-making process, there was no examination of alternative pol-
icies or overall costs once the first major commitments had been
made. Instead "the issues that were addressed and the decisions that
were made were always tactical in nature." It was assumed that the
cost to the enemy could be made too high, but never considered was
the possibility that the cost to the United States would be too high.21
Public opinion was considered to the extent that, as Johnson's ques-
tion to Resor indicated, its support was considered necessary for the
original commitment to large-scale troop involvement.
A closer examination of the polls would have shown how tenuous
that support was. The ambiguity of the public was indicated by its



favoring both strong military measures and negotiations. In essence,
because of their lack of knowledge about Vietnam, people were will-
ing to support any course of action that would work, deferring to
the president's office and information. This deference can result in
brief surges of support after critical events such as the Gulf of Tonkin
attacks, but, in the long run, the public's support for large-scale
American involvement could only be sustained if it could see signifi-
cant progress in return for the cost.
Such public uncertainty makes responses to polls very sensitive
to the wording of the questions,22 a constant problem for Quayle.
His July 1965 survey of West Virginia found 82 percent agreeing
that we should "do as we are, keep military pressure on but seek
negotiation."2 Because this statement contained something for both
hawks and doves, as well as pointing out ("do as we are") that this
was the president's policy, the overwhelming agreement with it dem-
onstrated far less support than the 82 percent seemed to indicate.
Quayle's Vietnam analysis contrasts sharply with his more cautious
interpretation of his poll on the Dominican Republic intervention,
when he advised the White House to "tell the President that while
this is highly favorable, he should not kid himself into thinking it
was overwhelming," citing 90 percent approval for President Ken-
nedy after the Cuban missile crisis.24

More Troops and Less Support

During the steady escalation of the war after the original decision
had been made-the second stage, from late 1965 until March
1968-Johnson's job ratings, both overall and on Vietnam, were in
decline, despite a few periods of temporary resurgence. There are
three possible ways polls could have been used at that time. First,
they could have been designed to measure public approval of specific
policy alternatives. Unfortunately, public opinion was simply too con-
fused to do this. For example, a poll taken in early 1966 by Lloyd
Free found the public evenly divided about whether to extend a pause
in the bombing of North Vietnam. However, one-third of those who
thought resumption would strengthen North Vietnam's will to fight,
would bring China into the war, or would hurt the United States
among its allies nevertheless favored resumption, leading Free to con-
clude that there was no public mandate for or against.25


LBJ and the Polls

A second possibility would have been to use polls to provide a
basis for analyzing the drop in public support, in order to rethink
Vietnam policy and evaluate alternative policies. Such an analysis
would have been in line with the hopes of those who believe that
polls could be used to present bad news to a president, but Johnson
would have no part of such a procedure. Instead, he sought to tell
the public as little as possible, delaying public announcements, break-
ing requests to Congress into smaller parts, and not calling up the
reserves. His "middle course" rejected both full mobilization or
anything approaching it because of the danger to his domestic policy
initiatives, the risk of a wider war with either China or the Soviet
Union, and the possible loss of a face-saving way out through negotia-
tions. According to Berman, "President Johnson did not want a great
national debate on the war." Despite the pleas of some advisors to
prepare the public for a long war, Johnson adopted a "policy of mini-
mum candor" which would result in a "credibility gap" that caused
the press and much of the public to doubt official government pro-
Having eliminated the first two possibilities, Johnson was left with
the third, using the polls as a basis for a public relations effort to
persuade the press to cover Vietnam more favorably and to cultivate
support among both the general public and those with political influ-
ence. Quayle's polls provided considerable information demonstrat-
ing growing public disenchantment with Johnson's Vietnam policy.
An Iowa survey taken at the end of 1965 showed a majority disap-
proving of Vietnam policy despite an overall approval rating of 61
percent for the president. Johnson's popularity was even higher in
New Jersey yet only 40 percent in Oklahoma, but in all three states
Quayle suggested that domestic issues outweighed Vietnam. He also
found more hawks than doves.2
At about the same time, Hayes Redmon, in summarizing the public
polls for the president, warned Johnson that apparent public hawk-
ishness masked potential pitfalls. On the one hand, "the American
public has supported the military increases made during the past
year. In some respects they would be willing to go further," although
not so far as to invade North Vietnam, use nuclear weapons, or bomb
China. On the other, he pointed out that "if the public supports our
military increases to date, they are even more overwhelmingly in
favor of attempts at negotiated settlement," having supported every



peace proposal put before them. He concluded that public support
for escalation grew out of the hope that it would bring peace quickly.
If not, Johnson's support could evaporate.28
Quayle's 1966 surveys demonstrated that evaporation. A New York
poll in March showed, according to Moyers, "that he is in serious
trouble with New York voters" although his ratings on Vietnam re-
mained favorable. Three weeks later a West Virginia survey showed
62 percent disapproval on Vietnam, but strongly positive ratings on
domestic issues kept Johnson's overall ratings solidly favorable. Simi-
larly, a series of Kentucky polls showed positive ratings on all major
issues except Vietnam, while Tennessee and Wyoming contributed
negative ratings only on Vietnam and inflation.29
By the time of Quayle's New Jersey survey in late May, public disen-
chantment over the war had significantly eroded Johnson's overall
popularity. In that state his job rating had declined 19 percent in
two years. Even more significant, according to Quayle, were trial heat
losses to potential 1968 opponents Nixon and George Romney. The
public rated Vietnam the leading issue but disapproved of Johnson's
handling of it 65-35 percent, with most of the opposition coming
from hawks. To Quayle the situation was "alarming" enough to point
the need for a detailed national survey to discover more specific rea-
sons for Johnson's problems. If Johnson did not wish to spend the
$20,000 that it would cost, Quayle could tack on questions to another
survey with a short summary report for the White House for a frac-
tion of the cost. He repeated the plea in a New Hampshire survey
with similarly "gloomy" results, telling Johnson that he must find "a
way to rebuild hawk support without fighting an unlimited war."30
Clearly, the polls had provided plenty of evidence that Johnson
was in trouble. Although they caused considerable concern among
Johnson's aides, the result was not a reassessment of Vietnam policy
but a redoubled effort to do a better job of selling the current policy.
Quayle's requests for a thorough national survey to diagnose the
problem better went unheeded.
A clear example occurred when Moyers informed the president
that "conversations with Gallup, Harris, and other professionals in
the poll business confirm only one thing: that our standing is down
and likely to drop further," primarily because of frustration with Viet-
nam policy. Gallup and Harris suggested that Johnson "offer some
ray of hope" but that "a contrivance" for false hope followed by a


LBJ and the Polls

lack of real progress would only result in "a more disastrous position
with the people than now." Instead of such a real policy adjustment,
however, Moyers recommended selective leaks of the favorable por-
tions of Quayle's polls: "We would not tell of the slippage of the
Quayle reports but merely refer to some of the recent favorable rat-
ings." These leaks should also point out that most opposition was
from those who wanted stronger military action to show that the pres-
ident "is a man of restraint even in the face of growing public pressure
for increased military efforts.""3
These recommendations were promptly implemented. At a June
24 staff meeting Redmon reviewed the meaning of the polls, empha-
sizing the positive but pointing out some problems. He explained
that this information should be used only in background briefings
to correct public and press "misapprehensions." Robert Kintner, who
had been brought to the White House from NBC to help improve
Johnson's image, had already begun such briefings.2 A Quayle poll
of Wisconsin was leaked to Evans and Novak despite doubts about
its accuracy.33 The most revealing leak had appeared in the Christian
Science Monitor in March. Saville Davis reported that "Mr. Johnson
quoted at length, to at least one visitor, the results of the most recent
survey of American opinion by Louis Harris and Associates," which
concluded that the public was becoming more hawkish: "Since the
President considers himself the architect of restraint, he wants it
known that he regrets this trend." Davis pointed out, however, that
despite Johnson's argument that most of his- 14 percent decline was
because of hawks, that group had only increased by 4 percent.34
Johnson sought to meet hawkish criticism and the still deteriorat-
ing situation in Vietnam by increasing the number of troops and
stepping up the bombing. In a series of midwestern speeches he told
his audiences that the Vietcong was watching the upcoming congres-
sional elections, so serious talks would occur only if supporters of
policy won. Enthusiastic crowd reactions reassured him that, contrary
to all the evidence of the polls, the public backed his policies.35
Johnson and his aides appear to have succeeded in convincing
themselves more than they had the public. As Kearns put it, "The
lower the President's popularity fell, the more Johnson had to see
his decision to escalate as the only decision he could have made."
The result was a distortion of facts, overly optimistic public reports,
attacks on dissenters, and self-deception. To demonstrate restraint he



had to exaggerate opposition from the right, but, despite his negative
image he told Kearns after leaving office, "Deep down I knew-I
simply knew-that the American people loved me. After all that I'd
done for them and given to them, how could they help but love me?
And I knew that it was only a small percentage that had given up,
who had lost faith." As a result, Johnson listened more and more
to those few advisors he trusted, shutting out potential opposition.36
Such psychological defenses are a prime reason that polls alone can-
not break through presidential isolation on major issues, a theme that
will be more thoroughly explored in the concluding chapter.
The White House grasped at every improvement in the polls, no
matter how temporary, as evidence that the public really backed them
on Vietnam. When Johnson received an advance copy of a July Gal-
lup Poll showing a rise in popularity after increased bombing of
North Vietnam, he immediately sent copies to three House Demo-
cratic leaders. Panzer told Johnson in May 1967 that an upcoming
Harris Poll was "a dramatic turning point" that could "take the Viet-
Nam war out of the campaign." When Johnson learned at about the
same time that Gallup would also show an improvement in his ratings
on Vietnam, he immediately instructed his press secretary to call in
"AP, UP, and two of the networks" for a background briefing.37
Such favorable results were only brief interludes in a continuing
drop of public support for Johnson's Vietnam policy. Convinced that
the public really supported him, Johnson began to question both the
public polls and those taken for him by Quayle. In chapter 5, I will
examine how Johnson tried to counter his overall declines in the
public polls, but some important points relate specifically to Vietnam.
Johnson hoped to get higher Vietnam ratings by looking to additional
sources of polls. In December 1965, Harris was again secretly com-
missioned to take a Vietnam survey. It gave Johnson a favorable rat-
ing of 61 to 39 percent, which Redmon contrasted to lower ratings
in Quayle's state polls.38
The president and his aides continued to seek out such rays of
hope, no matter how dim their reality. When a poll of Dallas County
taken for Congressman Earle Cabell in August 1967 was sent to the
White House, Fred Panzer discounted the pollster's analysis that the
results were "potentially dangerous to Democratic candidates who
may be perceived by voters as being closely aligned with President
Johnson" by stating it was "full of gloom and doom. .. Yet his


LBJ and the Polls

trial heat results strike me as good."3" This interpretation is typical
of Panzer's reports to the president, searching out the most optimistic
possibilities rather than telling the president just how serious the
problem was. This ploy served only to reinforce Johnson's mistaken
belief that published polls seriously underestimated his support. Ac-
cording to John Roche, "L.B.J. was perpetually OD-ing on polls.
Fred Panzer worked day and night finding items of interest-most
of which were in my opinion-worthless."40 While Roche is objectively
correct that Panzer's memos frequently provided material that did
not add accurate knowledge of public opinion, he misses the point.
Johnson was looking for information to support his own viewpoint,
and if Panzer could provide some with the objective veneer of scien-
tific polling, so much the better.
One hope for demonstrating the president's popularity was to show
that pollsters had been asking the wrong questions. In early 1966
Quayle offered respondents a choice among four policy options:
"(1) We should go even further and either win or force negotiations.
(2) We should do as we are. Keep on fighting but seek negotiations.
(3) We should stay in Vietnam but reduce military operations.
(4) We should get out of Vietnam now." Those who chose number
one were classified as hawks, number two as LBJ supporters, number
three as enclavists, and number four as doves. It is easy to see why
the White House disliked this formulation. Two dovish options and
a phrasing for the hawkish one that emphasized both escalation and
negotiation left little room to support the president. The White
House told Quayle to substitute these options: (1) I disagree with
present policy; we are not going far enough. We should go further
such as carrying the war more to North Vietnam. (2) I agree with
what we are doing, but we should increase our military effort to win
a clear military victory. (3) I agree with what we are doing, but we
should do more to bring about negotiations, such as a cease-fire re-
quest. (4) I disagree with present policy; we shouldn't be there. We
shouldn't be bombing North Vietnam and should pull our troops
out now.
This new phrasing seems designed to maximize support for the
president. Two of the four choices were counted as supporting ad-
ministration policy. Respondents could be critical of those policies
from either direction yet still be counted as supporting them. Appar-
ently the White House thought there were too few unconditional sup-



porters of the president's policy to be worth asking about. To be
counted as a dove, a respondent had to favor immediate and total
withdrawal. Thus, more doves would choose option three than hawks
would choose option two. Since Johnson believed there was a greater
danger from hawks than doves, this wording would likely lead to
results to confirm his preconceptions. Using this new wording, the
1966 Quayle polls discussed earlier found just that. Quayle himself
preferred his original wording because it separated moderate from
more extreme doves.41
It would take more than new wording to stem the decline in sup-
port for the Vietnam War. As that decline continued, the White
House found new reasons for doubting Quayle's results. When a Ver-
mont poll taken in June showed that Johnson's Vietnam rating had
dropped sharply even with the new questions, Redmon suggested
it could be of "little relevance" because published Gallup and Harris
surveys indicated that subsequent bombing raids had increased the
president's popularity. He also pointed to Quayle's prediction that
"those who wish to get out of Vietnam will have no place to go in
1968," since most were Democrats and the Republican was likely to
be at least as hawkish as Johnson. Events were to prove Quayle a
better pollster than political prognosticator. A South Dakota poll
taken at the end of June was not delivered to the White House until
mid-August, so a similar critique was made by Panzer and Tad Cant-
ril, the latter arguing that "these data are so dated that they are more
likely to be misleading." Both also pointed to the relatively small sam-
ple of 296 as making possible significant error. Quayle was having
trouble delivering polls on time, so such criticisms were frequently
possible, especially with the South Dakota survey which was prepared
so hastily that it included only tables without analysis.42
The decline continued. Yet another version of the Vietnam ques-
tion was tried without success. Still, the White House did all it could
to discount or reinterpret the message the polls were so clearly deliv-
ering. When a Gallup Poll of August 1967 indicated that Johnson's
lower ratings were largely due to Vietnam, Panzer told the president
that Harry Truman had similar declines in 1947 before rebounding
to win the 1948 election. However, Panzer failed to point out that
this was before the Korean War, an important factor in the Republi-
can victory in 1952. Even Panzer's optimism was damaged two weeks
later when he had to characterize the next Harris release as showing


LBJ and the Polls

"for the first time . the war appears to be hurting your chances
for reelection," although he countered by arguing that troubled times
made popularity ratings "unfair barometers." Panzer continued to
argue that the wording or analysis of Gallup and Harris surveys was
misleading, periodically leaking his views to the press. Johnson found
these activities important enough to authorize Panzer personally to
approach Gallup and to give background briefings to favorable re-
Hoping that still another pollster would find greater support for
the president, the White House hired John Kraft to take additional
state polls. A Missouri survey in October showed a 50 percent in-
crease in opposition to the war, but a few weeks later, Kraft was "sur-
prised" by better results in a Pennsylvania poll, which used new
phrasing suggested by Panzer. Even Undersecretary of State Nicholas
Katzenbach suggested a poll with new questions and a new pollster."
Nevertheless, Quayle was retained as the main White House poll-
The most bizarre study of all, in late 1966, was Quayle's study
of public opinion in Vietnam.45 Under the supervision of Robert Sulli-
van of the USIA, 974 interviews were conducted by trained Vietnam-
ese. It seems hard to believe that a survey of a war-torn country by
government-sponsored interviewers could yield valid results, but pre-
cautions were taken to try to elicit honest answers. Although the inter-
viewers needed a letter of identification from the national govern-
ment to local officials, they w-re told to keep their official sanction
secret. Each team's helicopter landed as far from the village as possi-
ble, then the team was driven the rest of the way in heavily armed
jeeps. Laughable as these precautions may appear today, Quayle be-
lieved the answers to be honest because many were critical of the
government. Honesty was not the only problem. The lack of respond-
ents aged 26 to 29 is suspicious, but there was no accurate census
by which to judge the sample's representativeness. Because some
areas were too dangerous to send interviewers to, their responses
could only be guessed at. Even in the safer areas one team was cap-
The key question for the White House was "If there were a free
election today would you vote for the present government in Saigon
or for the National Liberation Front." Eighty-eight percent supported
Saigon, 2 percent the NLF, 7 percent were not sure, and 3 percent



refused to answer. Adjusting for Viet Cong-controlled areas, the
results favored the government 72 to 20 percent. As was said in the
study, "The Viet Cong were not asked."
A three-volume report was presented to several leading public
opinion experts whose comments are included in a fourth volume.
According to the report, Paul Lazarsfeld asked for analysis of the
"don't know" answers that might indicate slightly greater anti-
Americanism than the tabulated results. Louis Harris thought the
answers showed "sensible patterns of discrimination," while Warren
Miller asked for several cross-tabulations.
Because more than twenty years have passed since this survey, most
of those involved have either died or simply have no recollection.
Miller's recent account casts doubt on the honesty of the fourth vol-
ume.46 He, Harris, and Lazarsfeld met twice, once in the Executive
Office where they were asked to evaluate the results. For Miller the
idea of obvious city people interviewing villagers made any results
unreliable. He questioned the entire operation, believing that "data
clearly had been cooked" since the poll showed that bombing villages
improved their morale. Because the cross-tabs he requested showed
the same problems, he refused to sign the report, believing the only
purpose of bringing in the reviewers was to provide a veneer of legiti-
macy. When the White House later called to ask what it would take
for him to sign, he asked to end his involvement. Thus, we cannot
be certain that the comments attributed to Harris and Lazarsfeld indi-
cate their approval. It would appear that the main purpose of the
study was to provide evidence of success of administration policy,
either to reassure Johnson or to be leaked to the press to gather public
support. The latter was not done. Since the results were given to
the South Vietnamese government, another possibility may have been
to convince them to hold truly free elections. Given the obvious flaws
of the study, the purpose could not have been actually to determine
what the South Vietnamese public thought.
For the most part, the main use of polls was to increase support
for administration policy. The selective leaking of private poll results
continued as did distribution of favorable published poll results. For
example, a Harris Poll of late January 1968 showing Johnson defeat-
ing potential Republican opponents in trial heats was circulated to
Democratic governors, state and county chairs, and members of the
Democratic National Committee. The White House also consistently


LBJ and the Polls

sought to counter critical press reports. In a memo to Johnson, Pan-
zer cited polling data disagreeing with Joseph Kraft's February 15
column, in which Kraft claimed that public polls "show only a feeble
percentage for a win strategy based on all-out use of force." Johnson
approved Panzer's recommendation to give the counterarguments to
politicians and other columnists.47
Because Quayle's polls were taken in individual states and even
specific cities or congressional districts, good results could be used
to demonstrate to legislators the political wisdom of supporting the
president. These numbers were especially useful for individual meet-
ings with legislators. Johnson was legendary for his skill in such meet-
ings, but he sometimes had to delegate them to others. Even George
Ball, the prime opponent of escalation in the administration, was
used. In a meeting with Rhode Island Senator Claiborne Pell, Ball
showed him a section of Quayle's most recent poll, which indicated,
"On balance, three out of five Rhode Islanders agreed with the Ad-
ministration's stance in Vietnam. ... Of the remainder, almost three
in ten are harshly 'hawkish,' a tenth are extreme doves, and only
one in a hundred agrees with [Senator] Aiken's moderate dove posi-
tion."48 Since the highlights of nine previous state surveys were also
included, it was likely that the administration had polled Rhode Is-
land more frequently than its own senator.
Polls were also frequently used to measure the impact of important
events, not so much to gather information to help make policy or
even to make the best arguments but to supply a positive picture
to the press, the public, and even to the president. Interpretations
always emphasized the positive. For example, after the 1966 State
of the Union address, the first of Johnson's to be dominated by for-
eign policy, a telephone poll "utilizing the facilities of the Gallup
organization" was taken. Although Mrs. Johnson described the con-
gressional audience's reaction as "cold and lethargic," Redmon empha-
sized public approval, comparing the 28 percent who made favor-
able references to Vietnam policy to the 8 percent unfavorable.
However, even he could not overlook the 41 percent who believed
that the president was withholding "important information."49
A few months later the White House was sent a Mutual Broadcast-
ing phone survey taken after Senate Foreign Relations Committee
hearings on the war. Fifty-two percent agreed with Johnson's policy,
16 percent disagreed, and 32 percent had no opinion. Redmon sug-



gested an optimistic public interpretation, that "of those indicating
an opinion 76 percent agree with the President's policy," but warned
privately that the bare majority coupled with a recent Harris survey
"shows the trouble we are in.""50
A final example shows just how far the White House would go
to force these polls into telling what it wanted to hear. Quayle took
a telephone poll the night of November 20, 1967, to measure reaction
to a series of events-the president's Veterans' Day trip, a press con-
ference, and Romney's declaration of his presidential candidacy. The
results showed little impact from these events, with Johnson's job rat-
ing 1 percent lower than in the previous week's Gallup Poll. Panzer,
however, decided that the trip had a 3 percent positive impact. He
then proceeded to discount the poll's accuracy because its 507 inter-
views in 20 large cities did not include New York, Chicago, Los Ange-
les, Detroit, or Boston. He was sure Johnson's rating was improving,
"but this poll may be too crude to register it. Gallup people tell me they are
picking up comments from people that are highly favorable" (Panzer's em-
phasis). Domestically, he argued, Vietnam was not as bad as people
think, since the strongest Republican in the trial heats was Nelson
Rockefeller, who supported the administration's policy.51 Like most
political experts, Panzer had overlooked Eugene McCarthy, who was
about to enter the presidential race.

De-escalation and Withdrawal from the Race

Although Johnson kept up his travels and speeches, the most surpris-
ing event of his presidency was in the works, his March 31, 1968,
speech announcing four major decisions: his withdrawal from the
presidential contest; a halt to most bombing of the North; more atten-
tion to improving the South Vietnamese military; and a relatively
small increase in American forces in Vietnam.52 The first two led
the North Vietnamese finally to agree to negotiations. Why, after re-
fusing to reconsider his Vietnam policy despite the evidence of grow-
ing opposition, did Johnson suddenly change his mind?
Johnson had been considering withdrawal for some time, largely
because of his and his family's history of stroke and heart disease.
As early as 1965 he revealed these thoughts to friends. Horace Busby
had written a withdrawal statement in December 1967, but Johnson
kept putting its delivery off. He claims that, although his Vietnam


LBJ and the Polls

initiative was the main reason for selecting March 31, Truman's simi-
lar timing in 1952 to give other Democratic candidates enough time
was also a factor. Johnson had not formally declared his candidacy
despite the urging of advisors who were warning him of trouble in
upcoming primaries, something that will be discussed later in this
chapter as well as in chapter 6.
As for the de-escalation, the Tet offensive at the end of January
was the galvanizing event. To counter adverse news accounts, the
White House told Ambassador Ellsworth Bunker and General West-
moreland to brief the press daily. At a February 2 news conference,
Johnson declared the Tet attacks, about which "we have known for
several months," a failure. The military, however, requested an addi-
tional 108,000 troops by May 1 and 97,000 more by the end of the
year. New Secretary of Defense Clark Clifford was asked to study
the request. He recommended deferring all but 22,000 of the re-
quested increase pending a reassessment of the entire Vietnam strat-
egy. Secretary of State Rusk proposed an unconditional halt to most
bombing of North Vietnam, accompanied by negotiation proposals.
He expected North Vietnam to refuse but believed that this move
would increase domestic support for eventual troop increases. Trea-
sury Secretary Henry Fowler pointed out the heavy economic costs
of future escalation-a need for more tax increases and substantial
budget reductions.
Outside pressures began to become more obvious. Although John-
son defeated McCarthy in the New Hampshire primary, his narrow
margin was viewed as a virtual defeat. With the Wisconsin primary
coming up and Robert Kennedy now in the race, the possibility of
primary losses had to be considered.
To gain an additional perspective, the "wise men," a group of for-
mer high-level officials who had endorsed the president's policy only
five months earlier, were called in. After several briefings, they told
the president on March 26 that most favored some form of de-
Johnson's memoirs give four reasons for his decision to deescalate:
improvements in the South Vietnamese military; the unlikelihood of
another Tet-style offensive; serious domestic economic problems; and
increased public opposition to the war.53 The first two have the air
of rationalizations, for they can hardly be the dramatic events neces-
sary to cause such a major change in Johnson's policy. The third was



an important consideration. Johnson's proposal for a tax increase was
bogged down in Congress, but his de-escalation policy and his with-
drawal could significantly reduce partisan debate resulting in quicker
passage of the bill. This sequence of events did occur and will be
discussed in the next chapter. The fourth factor is the most problem-
atic. For nearly two years, polls had been showing growing opposi-
tion to Johnson's Vietnam policy. What made that opposition sud-
denly so much more obvious?
Clearly public reaction to the Tet offensive was more dramatic than
slowly eroding poll results. Curiously, however, Quayle was so occu-
pied polling the primary states that no polls were taken specifically
about the offensive. Reports from New Hampshire proved grossly
overoptimistic, starting with a Quayle survey in mid-December show-
ing Johnson with 65 percent to 7 percent for McCarthy and 6 percent
for Kennedy. Apparent confirmation was found in a Joseph Napoli-
tan poll with a much larger sample taken at the end of the month,
which gave Johnson an even bigger lead. Panzer concluded that "New
Hampshire is a hawk's nest." Quayle's final poll gave Johnson a 42
percent lead although it was critical of the campaign's strategy. As
late as March 7, state campaign manager Bernard Boutin told John-
son, "McCarthy's campaign has already reached its peak and is slip-
ping," predicting a 64-26 percent victory. Panzer had earlier told
the president that the public polls showed that "McCarthy is failing
miserably in tapping dove sentiment."54 McCarthy's excellent showing
sent shock waves through the White House. When Kennedy declared
his candidacy on March 16, Johnson's political aides were seriously
Even before Kennedy's declaration, pessimistic reports were com-
ing in from Wisconsin. The situation was deemed so serious that
James Rowe, one of Johnson's top campaign advisors, concluded that
"somebody ought to be blunt to the president," so he sent a memo
telling him that Wisconsin would be lost without a dramatic move
for peace.55
A few days before Johnson's speech Panzer forwarded a
forthcoming Harris column to the president. The basic conclusion
was that "the public wants concrete results, and that means getting
the Communists to the negotiating table. The assurances of political
and military leadership are not going to be swallowed whole any


LBJ and the Polls

James Jones, Johnson's chief of staff in 1968, wrote, "I can state
categorically that fear of losing the 1968 election was not the reason
he retired. Several days before the speech, Mr. Johnson commissioned
a poll, which indicated that he would be re-elected over all possible
candidates. I always have felt that he took that poll to satisfy himself
that he wasn't being run out of office."57 Although the reasons given
by Jones for Johnson's withdrawal are plausible-concerns about
health and family and worry over Vietnam-a single national trial
heat poll showing an incumbent ahead seven months before the elec-
tion is not enough to support his argument that worry about reelec-
tion was not a factor. At best, Johnson was facing a difficult campaign.
Quayle's final state polls, also received only days before Johnson's
withdrawal, indicated a tough general election race in many states.58
Before the election there would be a battle for the nomination. John-
son was facing a probable defeat in Wisconsin and, if he did not
quickly declare his candidacy and begin an active campaign, more
defeats in upcoming primaries. Thus, he had to decide immediately
whether to run. Had he chosen to stay in the race, his control over
the party machinery (only about one-third of the delegates were cho-
sen in the primaries) would likely have given him the nomination
even with additional primary losses, but such a divisive campaign
would surely have cost him public support. Nor would a deteriorating
economy, his decision to press for a tax increase, and a lack of prog-
ress in Vietnam have given him many good issues to campaign on.
Trial heats may have given him a temporary edge, but the dynamic
of the campaign made it likely that such an edge would not last.
He had only to look at the New Hampshire primary to see that.
Health and family concerns may have been factors in his withdrawal,
but they do not explain the de-escalation. And if the de-escalation
was motivated by a belief that, as Jones wrote, Vietnam was "a blot
on his Administration he wanted to remove," why did it take him
so long?
Johnson's dramatic shift occurred only because of a series of
shocks-the Tet offensive, the report of the "wise men," McCarthy's
surprising showing in New Hampshire, Kennedy's declaration of can-
didacy, and the likelihood of defeats in at least some of the upcoming
primaries. When economic problems and Johnson's concern for his
health are added, we can see just how serious a jolt it took to cause
him to change his Vietnam policy and decide not to run. Evidence



of popular disenchantment provided by polls will rarely be as clear
or as sustained as it was on Vietnam, yet it proved insufficient to
prompt Johnson to do more than make tactical adjustments that were
more public relations than substance. On issues central to a president,
the hope that polls can break through his isolation, at least based
on the evidence in this chapter, appears a faint one. I will try to
explain why in the concluding chapter.


What can we conclude about the role played by the polls for Vietnam
policy? In the first stage they did provide information warning of
the dangers of a large-scale commitment, but that evidence required
a careful reading that is more obvious in hindsight than it was at
the time. The tendency of the public to support a president, especially
on foreign policy issues which they lack information about, coupled
with Johnson's high overall popularity, gave the appearance of sup-
port. The approach that worked so well in passing Johnson's early
legislative initiatives, blurring the lines between the different types
of consensus by simply emphasizing the approval figures while ignor-
ing the details, proved disastrous for Vietnam. Returning to Key's
formulation (see chapter 2), we can see that at the beginning of 1965,
Johnson faced at best a supportive consensus on Vietnam. The public
supported what they saw as current policy, but as the costs rose and
success seemed no nearer, Johnson passed the boundary and wound
up in a conflict situation with widespread disagreement about what
policy to follow.
In the second stage the polls provided plenty of evidence of trou-
ble. Rather than using that evidence to reconsider Vietnam policy,
Johnson used polls for public relations-gaining better press cover-
age, lobbying the politically influential, and making modest adjust-
ments in policy for tactical reasons. Polls were used to counter other
polls. Unfavorable poll results were ignored, reinterpreted, or dis-
counted. The primary use of polls was to attempt to manipulate opin-
ion rather than to gather useful information.
Only in the third stage was public opinion a major factor, but not
because of the information provided by polls. Instead, it took a series
of major shocks to cause a rethinking of Vietnam policy.
Once again, the potential of polls was not fulfilled. Polls did pro-


60 LBJ and the Polls

vide information that could have been useful in policy formulation,
but, as before, Johnson used only that information which was consis-
tent with his preconceptions. For the 1964 election and his early do-
mestic policy initiatives, the failure to utilize polls properly mattered
little. When it came to Vietnam, the result was fatal to the Johnson


Economic Troubles

SINCE THE GREAT DEPRESSION, presidents have been, as James W.
Davis wrote, "expected by the public not only to maintain the peace
and security of the country, but also to maintain high employment
and control inflation. . Though the president's ability to manage
the economy is minimal at best, the public nevertheless holds him
personally accountable for the state of the economy."'
When Johnson became president, inflation was low, but the 5.7
percent unemployment rate was considered high. Thus, Kennedy's
proposal for a tax cut to stimulate the economy was, along with civil
rights, Johnson's first priority. The tax cut was passed in time to pick
up the economy well before the 1964 election, but this stimulus set
off some inflationary pressures. By today's standards inflation rates
of 1-2 percent hardly look threatening, but at that time they were
considered potentially dangerous.
Johnson had inherited Kennedy's wage-price guidelines, which at-
tempted to tie prices and wages to productivity without providing
specific figures.2 Kennedy had developed the guidelines during his
successful 1962 effort to roll back steel price increases. Johnson's first
test came when United Auto Workers President Walter Reuther told
him in December 1963 that the union expected a large wage increase
because of high industry profits. The president's Council of Economic
Advisors suggested retaining the voluntary nature of the guidelines
while beefing them up by providing specific numbers based on a 3.2

LBJ and the Polls

percent productivity increase. In his 1964 economic report the presi-
dent endorsed the guidelines, but the UAW challenged him to per-
suade the auto companies to cut prices before the union would lower
its wage demands.
CEA Chair Walter Heller urged Johnson to take a strong stand,
citing a two-year-old Gallup Poll showing how strongly the public
had favored such action after Kennedy's widely publicized success.
He underlined the poll's conclusion that, by a margin of 68 to 19
percent, union members "would favor Presidential action against a
labor union that was seeking an inflationary wage increase." Heller's
suggestion was that this conclusion "may have a bearing on what you
say to the UAW."3 In his March 23 speech to the UAW convention,
Johnson paid heed to this advice. After urging support of his legisla-
tive program, especially the War on Poverty, he asked the union not
to "choke off our needed and our speeded economic expansion by
a revival of the price-wage spiral. Avoiding that spiral is the responsi-
bility of business. And it is the responsibility of labor." He then quoted
Kennedy's 1962 speech to the same convention urging voluntary wage
and price restraint, concluding that he would continue the same pol-
icy. Despite this plea, the auto companies did not lower their prices,
and the eventual wage settlement exceeded the guidelines. Because
of the election, Johnson did not wish to alienate both business and
labor by pressing the issue.4
In 1965 the focus returned to steel. When a price increase was
announced in December 1964, Johnson ordered a CEA study of the
industry. Meanwhile the union, which had not had a wage increase
since 1961, set a strike deadline of September 1. With the annual
inflation rate now up to nearly 2 percent, Johnson decided to take
action. He brought union officers and management to the White
House where he pressed them to settle within the guidelines. Among
his available weapons were a possible injunction under the Taft-
Hartley Act and proposals in Congress for a strike moratorium ac-
companied by attempts at fact-finding. To measure public reaction,
Johnson asked Quayle to add questions to his Kentucky poll for late
April. Quayle found the threatened strike to be of little importance
to most voters, but 90 percent did approve of Johnson's handling
of it.5
In September a contract at the 3.2 percent guideline was agreed
to, but Bethlehem Steel soon announced 4 percent increases in the


Economic Troubles

prices of some products. Johnson quickly wired the heads of the lead-
ing steel firms while members of the CEA met with Bethlehem, result-
ing in a rollback of half the increase.
However, the costs of the Vietnam War were beginning to start
a more serious inflationary spiral that such jawboning could not sup-
press. In October, Joseph Califano urged Johnson to take action on
food price increases, citing both an alarming report from the CEA
and a Harris poll showing a decrease from 44 to 28 percent in the
public's approval of the president's "keeping the cost of living down"
since March. Johnson agreed, calling a meeting of Califano, CEA
Chair Gardner Ackley, and Attorney General Katzenbach, who had
been investigating possible collusion among food distributors. By this
time, Quayle's polls were beginning to show problems. An early No-
vember survey, for example, showed negative ratings on only two
issues, inflation, and taxes and spending. On inflation, which was
ranked second in importance behind Vietnam, 64 percent disap-
proved of Johnson's performance. Analyzing this poll, Hayes Red-
mon disagreed with Quayle's inconclusive conclusions that "we are
not really sure what, if anything, can be done about it." Instead, he
argued, the voters, as indicated by Harris's polls as well as Quayle's,
wanted action. He suggested Johnson dramatize his concern by visit-
ing a supermarket, but inflation proved too serious a problem to be
eliminated by theatrics.6
By December both Ackley and Johnson's Budget Director Charles
Schultze were telling him that the only solution was a tax increase.
Wilbur Mills, in whose committee a tax increase would have to origi-
nate, opposed one. Seeking to minimize the public's perception of
the cost of the Vietnam War, Johnson declared in his 1966 State of
the Union address, "I believe that we can continue the Great Society
while we fight in Vietnam." At the same time he deliberately underes-
timated spending in his budget so as to require only a modest revenue
package reinstating previously reduced excise taxes and speeding up
tax collections. So anxious was the president to keep the economic
dangers from the public that he wrote on one memo, "Caution them
not to go into detail with staff and keep away from all reporters."7
Such half measures did little to stop the steady rise in prices. At
least seventeen 1966 Quayle polls showed negative ratings on the
issue, leading Robert Kintner to tell the president that inflation had
now overtaken Vietnam as the public's leading concern.8 Unfortu-


LBJ and the Polls

nately, the pollsters did little probing into details. Even if they had,
the public's lack of economic knowledge and the uncertainty even
of experts about the costs and benefits of any particular action cre-
ated a conflict rather than a consensus situation. Johnson, seeing the
lack of political support for the tax increases favored by his economic
advisors, chose not to play a leadership role, according to Ronald
King: "Instead of playing the advocate, he chose to stand among the
unconvinced, publicly expressing doubts that immediate action was
required, and awaiting incontrovertible evidence capable of indepen-
dently mobilizing consensus" while he also "suppressed debate, misled
the public about the severity of the situation, and put excessive strains
upon monetary policy, with distorting effects."9 In other words, John-
son chose to maximize his support in the short run by taking as little
action as possible, exacerbating his long-run problems. By waiting
for a consensus to form behind a single solution he was freed to
make this choice, for in a conflict situation the public, faced only
with unpalatable alternatives-tax increases, inflation, wage and
price controls; withdrawal from Vietnam; or severe domestic spend-
ing cuts-was not going to rush to line up behind any of them. Polls
would help Johnson continue this course, for, although they showed
considerable public discontent, they also provided evidence for the
lack of support for proposals requiring sacrifices, especially when
those sacrifices provided no certainty of solving the problem.
In essence the polls had made it clear that the public wanted the
president to take whatever action was needed to lower the inflation
rate as long as it did not involve too many sacrifices. The White House
again opted for public relations. Kintner suggested that Johnson as-
semble information showing that despite price increases, incomes
were up by more. In what today reads like an ironic echo of Ronald
Reagan's 1980 campaign theme, Kintner told Johnson to ask the pub-
lic, "Are you better off now than you were on December 1, 1963?"
or perhaps January 1, 1961? 0 Johnson continued to urge voluntary
restraint, telling the public, "Prices are moving up too fast to be com-
fortable." Speaking to meetings of business executives, unions, gover-
nors, and mayors, he asked them to cut spending to avoid a tax in-
crease. In May, the secretary of commerce sent a letter to 26,000
corporations asking for price restraints."
Despite such exhortations, the wage-price guidelines were in a
shambles by late 1966. At the beginning of the year, Johnson's public


Economic Troubles

criticism of a 6.3 percent New York City transit settlement had no
effect. The most telling blow came from the airline machinists. When
they struck, Califano, using the threat of the government's power
to regulate fares, pressured the airlines to keep their offer within
the guidelines. Even with such pressure the 4.3 percent settlement
exceeded the guidelines, but Johnson, in what one report called "a
triumphant TV appearance," announced that it was not inflationary
because of large productivity increases in the industry. Nevertheless,
the union's members voted the settlement down by an overwhelming
margin. On August 15, a new settlement of 4.9 percent was reached
which union leaders told their membership "destroys all wage and
price guidelines." Similar or larger settlements followed, as 5 percent
became the de facto guideline."
Public polls had encouraged Johnson to take on labor. In an analy-
sis prepared for Califano, Redmon argued that Gallup Polls showed
a trend against organized labor with a majority favoring binding gov-
ernment arbitration to settle strikes while opposing the 35-hour work-
week and increases in overtime pay. A plurality, even among union
families, thought that laws regulating unions were not strict enough.13
Califano passed this knowledge along to the president, but whatever
Johnson's inclinations might have been, organized labor was too im-
portant a constituency, especially considering his need for its support
of the Vietnam War, for him to alienate. Even had he decided to
do so, his powers as president were limited. This skirmish is worth
careful attention, for, as noted in the introduction, both proponents
and critics of polls argued that they would weaken the power of inter-
est group leaders. In the concluding chapter, I will take up this argu-
ment in detail.
With the effective collapse of the guidelines and Johnson's rejec-
tion of mandatory controls, the only choice left was to "bite the bullet"
of an income tax increase to pay for the Vietnam War." The available
polls did not contain much analysis of what remedies the public might
accept. Preoccupation with the war was so great that it dominated
Quayle's polling analysis. A Massachusetts report of October 1966,
for example, covered the president's popularity ratings and Vietnam
but made no mention of inflation despite tables showing that the pub-
lic ranked it second to Vietnam in importance. An overwhelming 86
percent rated Johnson unfavorably on inflation. Perhaps Quayle's
lack of analysis is due to his lack of ideas about what to do about


LBJ and the Polls

inflation, as mentioned in an earlier discussion. If so, surely someone
in the White House should have discussed the ideas that they were
considering so that questions about them could have been asked and
analyzed. But no one did. However, Walter Heller, by then out of
government, sent Johnson a published Minnesota Poll showing that
even though 63 percent of the public opposed a tax increase, 67 per-
cent expected one. He also reported that many business leaders who
publicly opposed a tax increase supported one in private conversa-
tions with him. Johnson even took his own personal survey at meet-
ings with business and labor leaders by asking those in favor of a
tax increase to raise their hands. None did. Of course, this ploy was
used by Johnson to make a point; it was not an attempt to gather
information about public opinion. Johnson decided to wait until after
the 1966 elections to recommend the tax increase that most of his
advisors now favored.15
In his 1967 State of the Union address, Johnson asked for an in-
come tax surcharge but delayed sending a formal proposal to Con-
gress. In late February, Panzer sent the latest Harris release to the
president, pointing out that the public, by a margin of 65 to 24 per-
cent, opposed the surcharge, a large rise from the 49 percent against
a year earlier. At the same time, an overwhelming 75 percent favored
spending cuts. Since the economic statistics improved slightly until
June, the proposal was not sent to Congress until August.16
In Washington there were three opposing points of view, none
commanding a majority.17 Led by Mills, congressional conservatives
were demanding large domestic spending cuts in exchange for their
support of a surtax. Most in the administration favored maintaining
both the Vietnam War and domestic programs, preferably without
a tax increase, if not, then with one. Liberals believed that it was
more important to maintain or even increase domestic programs than
to fight in Vietnam. Thus evolved a typical conflict situation-no ma-
jority, no clear solutions, and a public confused about the economic
details but worried that present policy was failing.
For the rest of the year Johnson tried to work out an agreement
in Congress, but Mills stood firm. Mills suggested that the presi-
dent make a televised speech explaining that guns had to be chosen
over butter and that the Vietnam War required both a tax increase
and cuts in domestic spending, but one of Johnson's aides countered
by telling the president, "We have just got to have it both ways


Economic Troubles

and Mills has to be convinced of this."'8 Neither was convinced.
During this period, Quayle's polls showed opposition to Johnson's
proposals but concentrated more on other issues such as Vietnam
and urban riots as well as trial heats against possible Republic oppo-
nents.19 Other surveys provided a little information about opposition
to the surtax. An August Gallup Poll showing an 8 percent decrease
in Johnson's job rating occurred, explained the Gallup analyst, be-
cause of Vietnam, urban riots, and the tax proposal. A survey taken
for Texas Representative Earle Cabell, showing public dissatisfaction
with Johnson on the same three issues, was sent to the president.
The White House commissioned Political Surveys and Analysis to
take a national survey, which produced similar results.20 These and
earlier surveys hammered home the trouble the president was in but
provided no useful information about solutions. The public was un-
happy with the economic situation but against his proposals. Having
downplayed the costs of the Vietnam War, he was now presenting
the bill.
By the end of the year Johnson had given in to Mills's insistence
on budget cuts, but debate over the details held up the passage of
legislation. Johnson continued to make speeches urging enactment,
including his 1968 State of the Union, but Quayle's phone poll con-
cluded that the public reacted unfavorably to that speech largely be-
cause of the surtax, an argument dismissed by Panzer as "only a con-
jecture.""21 Once again, it took Johnson's withdrawal speech, a
substantial part of which was devoted to the surtax, to break the legis-
lative logjam, although it took nearly three months more, as well
as an international gold crisis and intensive administration lobbying,
to gain final passage.
Ronald King has described Johnson's role during this period of
economic difficulty as that of "the myopic democratic calculator, in-
sisting that necessary tax correctives be readily acceptable to the pop-
ulace and enactable by a Congress wary of approaching Novembers."
This view, he argues was a failure of leadership, preventing short-
term political damage at the expense of the long-term health of the
economy.22 Such a role is facilitated by polls that made clear the pub-
lic's dissatisfaction with Johnson's economic policies but provided no
information about what solutions or sacrifices the public might be
willing to accept to solve them. More detailed questions would have
been unlikely to help much, for such hypothetical questions in the


LBJ and the Polls

absence of presidential leadership would surely show opposition to
any program with a realistic chance of success. Every proposal in-
volved serious costs, yet none promised comparable benefits. By com-
parison, earlier surveys had shown that the public was willing to pay
higher Social Security taxes to pay for Medicare. That trade-off was
a relatively clear one in which the public was able overwhelmingly
to conclude that the benefits outweighed the costs. These benefits
had been made clear by strong presidential leadership. On the econ-
omy the costs were clear, the benefits uncertain. Nor was the president
willing to provide a candid public explanation of the costs of the
Vietnam War early in his administration, when his popularity and
credibility might have been high enough for successful leadership.
Instead, the White House tried to solve the problem with relatively
low-cost proposals, starting with public relations, then voluntary
guidelines, then minor tax increases. As support in the polls declined
and inflation worsened, only the more costly alternatives were left.
Johnson's delay not only hurt the economy; it added to doubts about
both his candor and the Vietnam War.



Poll Wars:

Fighting the Popularity Gap

THE PRECEDING CHAPTERS HAVE made clear just how important
public standing was for Lyndon Johnson or, for that matter, any presi-
dent. The leading indicator of that public standing has become pub-
lished political polls. They are so universally accepted today that we
may forget that the first such syndicated survey, the Gallup Poll,
began only a little over fifty years ago. George Gallup's chief competi-
tor, Louis Harris, gave up polling for political candidates in favor
of public polling in 1963. Although numerous others publish poll
results, these two have been so much the best known that I will con-
centrate primarily on them.
The tendency of approval ratings of presidents to decline during
their terms creates a nearly automatic conflict between presidents and
those who take the public polls.' In order for polltakers to maintain
their credibility as detached observers, they must keep their distance
from the president, especially if the latter wishes to influence them
to upgrade his standing. Gallup wrote that "the Gallup Poll is a fact-
finding organization, or looked at in another way, a kind of score-
keeper in the political world."2 He believed that a private organization
would be more objective than a government polling agency that
"would be placed under constant pressure to produce the 'right' fig-
ures" lest it lose its funding. Thus, the measurement and analysis
of public opinion "is probably best left with persons and organizations


LBJ and the Polls

that have no connection with parties or candidates, that obtain their
financial support from groups or sources that are not committed to
any single political or ideological viewpoint, that have neither a con-
servative nor a liberal orientation, and whose findings can be re-
garded as reliable but not 'official.' "3 In order to maintain his standing
for objectivity, Gallup stopped voting in presidential elections (and
other elections about which he took polls) in 1936. Neither he (nor
his son George Gallup, Jr., who now runs the business) nor Harris
took polls for political candidates after going into public polling.4
Although Gallup's statement appears to indicate that anyone
studying the relationship between presidents and the takers of public
polls would have little to write about, I will argue that there was
a significant relationship between the Johnson administration and
public pollsters. We have already seen that analysis of the public polls
was regularly forwarded directly to the president by Hayes Redmon
and Fred Panzer. Because the major polls are generally made avail-
able to the president several days prior to publication,5 publicity for
favorable results and rebuttals to unfavorable ones could be prepared
to appear at the same time as those polls.

Using Favorable Polls

The most obvious use of favorable public polls for a president is to
publicize them as widely as possible to demonstrate his popularity
to those he needs to influence. Because these polls are taken by well-
known independent researchers with considerable experience, the re-
sults have far greater credibility than surveys commissioned directly
by the president. The quantitative nature of polls and their scientific
sampling methods add to this credibility. Between 1963 and 1966,
when Johnson's approval ratings were consistently high, he carried
poll results in his pocket, showing them to visitors at the slightest
provocation.6 Even when the polls were less favorable, positive results
were widely disseminated. At staff meetings, Redmon or Panzer
would point out which polls should be publicized and how.7 Johnson
sometimes directly involved himself in these details, on one occasion
telling Press Secretary George Christian to invite "AP, UP, and two
of the networks" to a background briefing to tell them that the next
Gallup Poll would show his highest ratings in five months.8


Poll Wars

Fighting Negative Poll Results

Countering negative polls is a more difficult matter, given the credi-
bility of the polls and the obvious interest of the president in disprov-
ing bad results. Johnson used several methods-debating the inter-
pretation of the polls, publicizing results, attacking the methods used
by the polls, trying to influence the results, and, paradoxically, court-
ing the pollsters. In order to enhance the credibility of the rebuttals,
the role played by the White House was frequently hidden.
The interpretations sent to Johnson by his aides usually sought
to emphasize the positive, even when a more neutral observer might
have found little positive to emphasize. Referring to a Harris release
of May 1967 indicating a nearly two-to-one negative rating for the
president, Panzer noted some "important unpublished Harris poll
data ... note that confidence in you is outstandingly positive among Negroes
and Jews" (Panzer's emphasis).9 Panzer failed to point out that these
two groups were the only ones with majorities rating Johnson posi-
tively. Even majorities of Democrats and 1964 Johnson voters gave
him negative ratings. A decline in the president's ratings on Vietnam
"will probably be smothered" by favorable reaction to the president's
broadcast speech, wrote Panzer in July 1967. A later drop in the
polls was accompanied by a memo comparing it to Truman's rating
before the 1948 election, while a late August Harris release showing
that the Vietnam War was hurting the president's reelection prospects
was coupled with an attack on the unfairness of presidential job rat-
ings which are stacked against an incumbent, especially "in stormy
passages of history."'0
On other occasions Panzer told Johnson that Gallup's analysis
underemphasized the positive. In January 1968 he wrote that good
results in presidential trial heats against Richard Nixon were "buried
as Gallup tries to generate headlines" by emphasizing gains by
Eugene McCarthy. However, Panzer claimed that even this tactic was
good, as a favorable Johnson showing "would have been very damag-
ing to Nixon. And we wouldn't want that."" Two weeks later he wrote
that dramatic Johnson gains were "buried" under a lead about Nelson
Rockefeller. Similarly, in mid-February he wrote that Gallup's empha-
sis on a 7 percent decline in Johnson's job rating "accentuates the
negative" but noted that the column ended with a discussion of the


LBJ and the Polls

temporary nature of such shifts, which should be distributed to
counter the headlines.12
Such reinterpretation of the polls was a major method of counter-
ing unfavorable results, but, to be believed, Johnson had to get others
to do the rebutting. His prime targets were sympathetic officeholders
and reporters. In June 1966, Johnson spoke to Special Assistant Rob-
ert Kintner "about his desire to disseminate more affirmative polls."
Kintner responded by calling a staff meeting to discuss how to do
it. Redmon and Panzer drew up statements for members of Congress
"on the good portions of the polls," and Redmon reported on conver-
sations with reporters especially interested in polls that he believed
would lead to favorable stories. According to Redmon, several of the
reporters had indicated their hope that the president's ratings would
soon rise.13 The drafting of letters or speeches was a common tactic.
At Johnson's request, Panzer drafted five letters to be sent to Gallup,
criticizing a poll showing late 1967 gains by a potential Republican
ticket of Rockefeller and Ronald Reagan. Of course, none mentioned
their White House origins.'4 Johnson approved a way to counter Jo-
seph Kraft's column of February 15, 1968, in which Kraft stated that
Gallup and Harris polls showed little support for the hawkish posi-
tion on Vietnam, by citing other Harris surveys in congressional
speeches and trying to "plant material for other columnists."15 When
Gallup and Harris polls in late January indicated a temporary up-
surge in Johnson's popularity, Johnson ignored Panzer's caution not
to oversell, asking to have speeches written for insertion into the Con-
gressional Record and sending reprints of these polls to all Democratic
governors, state and party chairs, and members of the Democratic
National Committee.16
Attempts to generate favorable stories in the press are common
to all presidents. Johnson's aides often called reporters in for back-
ground briefings at which information given out could not be specifi-
cally attributed to its source. When mid-1966 polls showed Johnson's
standing sinking, Bill Moyers not only spoke to Gallup and Harris
but consulted polls taken by Oliver Quayle for the president. Even
though all agreed that "our standing is down and likely to drop fur-
ther" because of Vietnam and inflation, Moyers recommended
backgrounding the press only with the positive: "We would not tell
of the slippage of the Quayle reports but merely refer to some of
the recent favorable ratings.""17 Specific reporters, notably Robert


Poll Wars

Spivack and Roscoe Drummond, could be counted on to write favor-
able stories. Spivack even sent ideas for speeches and campaign tactics
to the White House.18
Johnson himself told Panzer to contact Louis Bean, hoping that
he would provide a more optimistic view of the Gallup figures.19 Bean
had made his reputation in 1948 by predicting that, contrary to the
results of the published polls, President Truman would be reelected.
Although most scholars of public opinion have little faith in Bean's
methods,20 Johnson hoped for a repeat of his 1948 performance.
Bean told Panzer that Johnson's popularity had touched bottom in
October 1966, then fluctuated between 44 and 48 percent until June
1967, when it rose to 52 percent. He claimed that "this may be evi-
dence of a turnaround," but, fortunately for Johnson, this prediction
was not widely publicized as it turned out to be completely wrong.2'
The next Gallup Poll (August 12, 1967) showed a decline in Johnson's
rating to 39 percent.

The Art of the Leak

Because Johnson had so many private polls available, it was a simple
matter to leak any that were more favorable than the public polls.
Quayle regularly took polls, usually of individual states or parts of
states from 1964 to 1968. Because Gallup and Harris used national
samples, there was considerable material that could be made available
to the press as "contradicting" unfavorable results. While such leaks
were useful when the public polls were unfavorable, they were not
necessary during the early stages of the administration when ratings
were high. For example, a Quayle survey of July 1964 gave Johnson
such a big lead that it was suppressed, no doubt to avoid generating
overconfidence and expectations too high to be met.22 On the other
hand, as Johnson's popularity declined, leaks became more frequent,
peaking during the 1967-68 preelection period ending with John-
son's withdrawal.23
Quayle's polls were not the only ones available to be leaked. Por-
tions of John Kraft's surveys showing Johnson with high ratings in
Pennsylvania and ahead of Robert Kennedy in trial heats in New
York were also circulated widely.24 Although these leaks generated
some favorable publicity, not all attempts were successful. Erie
County (N.Y.) Democratic Chair Joseph Crangle leaked a Kraft poll


LBJ and the Polls

showing that Johnson's approval rating had increased from 58 per-
cent in August 1967 to 63 percent in October, but the New York Times
omitted the August rating and stressed the decline from 75 percent
in May to October's 63 percent. According to Panzer, the lesson was
that county leaders "could use some guidance on leaking the findings
in the best possible light."'5
Two polls taken by Archibald Crossley generated the most contro-
versy. Crossley, a respected pollster, was called out of retirement in
mid-1967 by Arthur Krim, a friend of Johnson. Krim hired him to
take two surveys measuring the president against potential Republi-
can opponents-the first of New York; the second of California,
Pennsylvania, and Strafford County, New Hampshire, a supposed
bellwether. Krim, who asked to remain anonymous, insisted on Straf-
ford County and also ordered Nelson Rockefeller omitted from the
New York trial heats. Crossley agreed, subsequently claiming that it
was with the condition that his permission would be required before
any results would be released.26 Since the supposed bellwether was
significantly more Democratic than the rest of the state and Nelson
Rockefeller was not included in his home state, it appears that the
poll was designed not to learn about public opinion but to put John-
son in the best light. The results largely did this. In New York, John-
son led Richard Nixon, Charles Percy, Ronald Reagan, and George
Romney by more than 20 percent. Results in California and Pennsyl-
vania were only slightly less positive (except that the now-included
Rockefeller ran approximately even with Johnson). In New Hamp-
shire only Nixon led Johnson (Rockefeller was only 2 percent behind)
while Reagan, Romney, and Percy trailed by margins ranging from
17 to 21 percent. The results were sent to Johnson by Panzer as "one
of the most encouraging polls I have seen."
The White House began to plan how best to release these results.
Krim informed Watson that he had been told by Crossley that the
polls could be released but only in their entirety. Jack Valenti sug-
gested to Johnson that the polls should be leaked to individual report-
ers rather than being revealed at a press conference, because "this
way, we can get the fullest circulation in key papers without subjecting
any of us to hostile questions from questioning reporters at a confer-
ence."27 At first, this strategy worked well. Poll results were sent to
state and national Democratic party leaders as well as reporters, re-
sulting in favorable mid-October articles by the wire services, by Wil-


Poll Wars

liam Eaton of the Chicago Daily News, and by columnists Roscoe
Drummond, William S. White, and Drew Pearson, among others.
Eaton told Panzer that he was especially pleased to clear up his confu-
sion over "hearing so much about your low popularity," while
Drummond made certain that the White House was aware of his work
by sending them a copy.8
Unfortunately, no one had realized that Crossley, believing himself
used, would go public. Without revealing his client's name, he told
the press that the leaks had neglected to point out that "Strafford
County is not typical of New Hampshire, that there were few inter-
views there and that it is a Democratic county." Columnists Rowland
Evans and Robert Novak made the source of the leaks clear, writing
that presidential aide Marvin Watson showed these polls to all Demo-
cratic visitors to the White House, then made it "perfectly clear that
he wouldn't mind one little bit if these polls found their way into
the newspapers." The White House was aware of the methodological
defects of the Strafford County poll, since one of Panzer's memos
to Johnson had a penciled notation "very high" next to the 8 percent
margin of error.29 Despite efforts to rebut Crossley, the disclosures
caused more harm than good.30

Other Attempts at Manipulation

Another way of reducing the impact of unfavorable polls is to argue
that they are incorrectly done. Believing that including George Wal-
lace as a third-party candidate in trial heat polls would help Johnson's
standing, Panzer suggested in an August 1967 memo that Harris be
urged to do this. In January, Krim made the suggestion to George
Gallup, Jr., who told him that such a poll had already been scheduled,
then wired his father for permission to begin it. When the results
appeared, however, Johnson and Nixon finished in a dead heat both
with Wallace and without him.31
Panzer also believed that Gallup favored the presidential candi-
dacy of Nelson Rockefeller. When an October 1967 poll showed a
Rockefeller-Reagan ticket well ahead of the Democrats, Panzer wrote,
"It is unusual for Gallup to distribute the undecided vote as he has
in this release. It looks like he wants to show the Rockefeller-Reagan
ticket in the best light. In fact, his story is almost a plea for Rockefeller
to run."32 This accusation led to drafting the letters to Gallup referred


LBJ and the Polls

to earlier. Similarly, Panzer complained that a January 1968 survey
buried Johnson's gains under those of Rockefeller.33 Earlier, in dis-
cussing why he believed the polls had not reflected what he thought
was an improvement in Johnson's image, he explained by criticizing
polling as "still a crude tool that is not as fool proof as it is reputed
to be by the pollsters," while also arguing that both Gallup ("a Repub-
lican") and Harris ("very close to the Kennedy camp in 1960") had
political biases that could have influenced their analyses.4"
Panzer even went so far as to circulate a letter from a Gallup inter-
viewer who was bothered by the lack of an answer for those who
approved of some presidential actions but not others.35 Such method-
ological criticisms had little impact; on several occasions White House
aides would suggest congressional investigations, and Johnson once
even proposed that Lawrence O'Brien "get a bill drafted and get
some Senator to do it," but nothing significant resulted.36
Even less effective were efforts to influence the results of polls
which should be differentiated from efforts by presidents to influence
public opinion, about which much has been written.7 Instead, these
are attempts to discover when and on what subjects interviews are
being conducted, followed by scheduling events specifically designed
to produce a brief upsurge in presidential support to coincide with
the taking of the poll. In order to prevent such manipulation, poll-
sters treat their interview schedules like military secrets. Panzer tried
to cultivate sources who would reveal these secrets. When he thought
he had learned the dates and subjects, he would suggest that Johnson
schedule trips or speeches a day or two before the interviews were
to take place.38 Nevertheless, there is no evidence that these efforts
had any significant effect.
Despite trying so hard to discredit the public polls, Johnson simul-
taneously sought to cultivate the pollsters. Given their declarations
of objectivity and their efforts to remain nonpartisan (Gallup's non-
voting and Gallup's and Harris's refusing to work for political candi-
dates), one would expect strong resistance to any presidential over-
tures, but the opposite took place. There appear to be two reasons
for this occurrence-a desire to be influential and a belief that help-
ing the president would be good for the country.
The simplest method was to pay personal attention to the pollsters,
a technique used on Harris. Favorable polls at the start of the Johnson
administration led to a call from Walter Jenkins to say how pleased


Poll Wars

the president was. When Harris injected a note of caution, Jenkins
replied, "Well I'm sure you are absolutely right, but you know the
boss. He won't forget this big show of support for a long time to
come.""39 Although Harris notes that such calls stopped as Johnson's
ratings sank, White House attention continued. At the beginning of
1968, Panzer informed Johnson that he was trying to cultivate Harris
and had sent him a copy of a book of the president's speeches. Harris
called Panzer to tell him that he "was very impressed and apprecia-
tive" as well as to give him information about favorable poll results
that had not yet been published.4 Within a month, Harris had met
with Johnson's friend, Texas Governor John Connally, for three
hours to discuss poll results and had been invited to a White House
Surprisingly, the pollsters were more eager to see Johnson than
he was to see them. When Richard Nelson called Gallup in early
1964 to thank him for an advance copy of a survey, Gallup enthusias-
tically volunteered that "the President is doing a fantastic job. ...
We just polled Republican county chairmen and they nearly all se-
cretly feel that the GOP doesn't have much of a chance in November."
After William Crook, the director of VISTA, told George Christian
in late 1967 that Gallup wished to meet privately with the president
because he felt bad about low popularity ratings, Johnson replied
that he would see Gallup only if asked directly.42
On several occasions, Harris gave Johnson detailed advice concern-
ing political strategy but asked that his involvement be kept secret.
In September 1967 he met with Panzer to give his "strictly confiden-
tial" formula for Johnson's reelection, immodestly claiming, "If I
know anything, I know how elections are won." Panzer noted that
Harris "pleaded that there be no Harris memos leaked from here,"
to which he replied that he would claim Harris's ideas as his own
but did not want to hear any complaints about plagiarism.4 In a later
conversation with Panzer, Harris suggested, based on data that would
not be published for another two weeks, that Johnson strongly attack
Robert Kennedy in his speech of March 31, again requesting confi-
dentiality.44 Since Johnson announced his withdrawal from the presi-
dential race in that speech, none of the advice was used. Harris met
with Johnson or his aides on other occasions as well to offer additional
Gallup and Harris were not the only prominent public pollsters


LBJ and the Polls

anxious to see the president. In March 1966, Elmo Roper asked to
meet with Johnson because "some of the so-called polls were not
doing a very good job of reflecting a rather rapidly growing feeling
of concern and apprehension." He had a plan "to shift much of the
spotlight and concern away from Vietnam" but had to settle for a
meeting with Valenti and Redmon several weeks later. Redmon, how-
ever, found unconvincing Roper's claim that Johnson was losing sup-
port among moderate to liberal Democrats but gaining among Re-
publicans, which led him to review Gallup Polls and speak to Harris.
None agreed with Roper, so his suggestions were ignored.46 Albert
Sindlinger dropped in shortly before the 1968 election to discuss a
telephone survey about the bombing halt in Vietnam which was to
be taken the next day.47
Although neither Gallup nor Harris takes polls for candidates,
the line they draw is fuzzier than they imply. Both took polls for
the Johnson administration. During November 1964 and January and
February 1965, Harris did three polls about Vietnam on condition
that the White House "not let anyone know he did it for us." As
discussed in chapter 3, his analysis stressed public support for admin-
istration policy. The next December he took another Vietnam poll
which showed more support for Johnson than Quayle's had. In addi-
tion, Redmon wrote to Johnson in late 1965, "Lou Harris has agreed
to do the poll you asked for on the Great Society," although no other
references to this poll appear in the available documents.48
Gallup's polls are taken by a network of firms, all headquartered
in Princeton, New Jersey, and sharing facilities and staff. The Ameri-
can Institute of Public Opinion handles the syndicated Gallup Poll,
while the Gallup Organization is in charge of private surveys. Thus,
there is some but not much separation between public and private
polls, although often both will be run as part of a single survey.
The Gallup Organization or its officers took several polls for the
Johnson administration. As early as August 1964, Moyers wrote John-
son that "through a friend of ours at a private foundation, the Gallup
Organization is willing to cooperate with us-on a confidential basis"
by providing unpublished survey data and taking "flash" telephone
polls to gauge reaction to events within twenty-four hours. A fee
of $3,500 was spent for a survey done by Irving Crespi, vice president
of the Gallup Organization, to check reaction to Johnson's speech
accepting the presidential nomination. According to Crespi, he took


Poll Wars

this survey "NOT as a Gallup employee" but as coordinator of "an
ad hoc, voluntary group" that neither used Gallup facilities nor re-
ceived any payment. The available documents are not completely con-
sistent with these recollections. Several mention payment, citing the
specific amount already pointed out. The correspondence and polling
reports identified both Crespi and Robert Bower by their titles and
affiliation with the Gallup Organization so that any distinctions be-
tween their acting as individuals and in their official capacities were
lost on the White House aides involved.49
Several months later, Lloyd Free took a similar poll "utilizing the
facilities of the Gallup Organization" to measure reaction to Johnson's
"peace offensive," but when Free's name was later suggested for addi-
tional polling, Panzer recommended against using him because he
had done considerable work for Rockefeller.50 A third Gallup Organi-
zation reaction poll was taken after the 1966 State of the Union ad-
At least two other polls were taken. According to the documents,
including a memo from Crespi to Redmon, Crespi took a survey on
immigration for which Redmon suggested questions. Crespi, how-
ever, now says that "I do not recall doing any further polling," other
than the survey mentioned.52 A poll of the attitudes of college students
toward VISTA was taken in 1967 for the Office of Economic Oppor-
tunity. In fact, OEO spent a large sum of money to hire a veritable
who's who of pollsters-$4,000 as part payment for Gallup, $44,600
for a Harris survey of Job Corps dropouts, and $147,000 for a Daniel
Yankelovich study of Community Action programs. According to
Panzer, Gallup was "very interested in applying public opinion poll-
ing" to the study of violence in America, but it is not clear if such
a study was ever undertaken.53
What were the results of all of these efforts to influence the percep-
tion of Johnson's popularity? Despite considerable effort by Johnson
and his aides, they were unable to provide any long-term help. Manip-
ulation and public relations can be effective in cases of temporary
decline, but Johnson's loss of popularity was far more sustained.
Johnson generated some favorable stories in the media and some sym-
pathy from pollsters, but in the long run his efforts made no signifi-
cant difference. Why he acted this way will be discussed in the con-
cluding chapter.
Pollsters also gained little from these efforts. Although some did


LBJ and the Polls

get a chance to speak to the president or his aides, their suggestions
had no significant impact on policy. On the other hand, their attempts
to gain influence endangered their objectivity. If they are to be, as
Gallup put it, simply scorekeepers, they need to maintain their dis-
tance from the players. Polling is as much an art as a science. Lipset
and Schneider wrote, "There are enormous variations in poll results
which can be attributed to question wording, to timing, and to the
context in which a question is asked in the interview."" In addition,
public polls are presented along with brief interpretations, adding
to the possible variations. That Gallup and Harris were aware of such
dangers is apparent from the contrast between their words and deeds.
Each stressed neutrality in his words-no polling for candidates, no
participation in election campaigns, in Gallup's case not even a vote.
When they gave advice to or took polls for the administration, they
always insisted that their participation be kept secret. While no one
can be absolutely neutral, claims to objectivity require less involve-
ment with the president. A scorekeeper's relevant actions must be
open to public scrutiny lest that public begin to lose faith in the integ-
rity of the game.



Politics and Elections

"POLITICAL PARTIES MAY WELL be more integral to the operation
of the legislative branch than the executive branch, but it is the presi-
dential party that captures the public imagination and shapes the
electorate's opinion of the two parties," writes Larry Sabato. "Along
with the inevitable headaches party leadership brings," he argues,
"there are clear and compelling advantages that accompany it." For
Sabato the ability of the party to mobilize support for the president's
program far outweighs its potential for divisiveness.' Similarly, Frank
Sorauf describes American parties as executive-centered coalitions,
combining "the executive-dominated party in government with the
party in the electorate."2
Maintaining this coalition requires presidents to take part in elec-
tions other than their own. Because his time and resources are lim-
ited, polls can be extremely helpful in deciding which elections to
concentrate on. Possible activities include recruiting candidates, cam-
paigning, making public statements for or against candidates, sending
money or other resources, and even deliberately withholding support
from a fellow partisan whose support of the president has been inade-
quate.3 The president has two related goals-maximizing congression-
al support and demonstrating popular support for his policies.
Johnson's limited involvement in other elections in 1964 was dis-
cussed in chapter 1. The two-year period before the next congression-
al election did not mean, however, that there were no campaigns

LBJ and the Polls

requiring presidential attention. The 1965 New York City mayoral
contest was particularly important, not only because New York is the
nation's largest city as well as its financial and media center but also
because a Democratic president needs strong city support to counter
the Republican majority in the rest of the state. Ordinarily the over-
whelming Democratic majority in registration assures the easy elec-
tion of the party's candidate, but in 1965 there was a strong challenge
from reformer John Lindsay who had been nominated by both the
Republican and Liberal parties. In contrast, the Democrats were
faced with a multicandidate primary. Nor was it self-evident that
Johnson was anxious for his party to win. Robert Kennedy, viewed
by Johnson as a potential rival, had been elected a New York senator
in 1964. Thus, a Lindsay victory could weaken Kennedy by providing
a New York counterweight. In addition, Johnson did not wish to an-
tagonize the Liberal party whose support could be important in the
1968 presidential election.4
Such considerations made Johnson seek more information before
deciding what to do. Before committing himself, he wanted to know
who the Democratic candidate would be and what his chances of vic-
tory were. Oliver Quayle's first report, sent to the White House in
June, well before the primary, consisted of a series of trial heats pre-
sented in tabular form without analysis. First, potential Democratic
primary voters were given a list of all possible party candidates.
Franklin Roosevelt, Jr., placed first with 30 percent, next was Paul
Screvane with 19 percent, while Abraham Beame, the eventual pri-
mary winner, placed fifth with only 9 percent. Beame was not even
included in the two four-person primaries then put to those ques-
tioned, with Roosevelt edging Screvane 37 to 36 percent in one and
Screvane easily winning with 52 percent when Roosevelt was ex-
cluded. Lindsay defeated each Democrat in a general election. Al-
though such a poll was more a measure of name recognition than
a useful predictor, it did indicate how serious the Lindsay challenge
would be. It also provides some confirmation for the widely held view
that Johnson's preferred candidate was Screvane.5
When Beame won the primary, Johnson resisted pressure to make
an endorsement that would ordinarily be a matter of routine from
a president who was a fellow partisan. Vice President Humphrey en-
dorsed Beame during a mid-October visit to the city, but given the
closeness of the race, Beame was desperate for direct public word