Reassessing public opinion polling in the Truman administration.
Presidential scholars (Brace and Hinckley 1992; Kemell 1986; Cornwell 1965), scholars of public opinion (Heith 1998; Geer 1996;Jacobs and Shapiro 1995; Herbst 1993;Jacobs 1993, 1992), and historians of President Truman (Hamby 1995; Heller 1980) all concede that the Truman administration did not have much interest in public opinion polling. Truman himself is also often quoted as negatively referring to polling and pollsters and disavowing the usefulness and accuracy of polls. As a revision to these claims, uncovered archival work shows that the Truman administration was concerned with public opinion polling and did consult polling on selected issues, both in the White House and during the 1948 election.
Truman's View of Public Opinion Polling
Truman's public position on opinion polls was very clear. He did not put much stock in the polls, the pollsters who developed them, or the politicians who obsessed over them. Truman was openly critical of these polls and pollsters for a variety of reasons. The fluctuation of Gallup's presidential approval polls (from a height of 80 percent to a depth of 23 percent) over the course of Truman's presidency combined with Gallup's famous failed prediction of the president's defeat to Thomas Dewey in the 1948 presidential election convinced Truman that polling was unreliable, ineffective, and capricious. Truman wrote in his memoirs,
I never paid any attention to the polls myself, because in my judgment they did not represent a true cross section of American opinion. I did not believe that the major components of our society, such as agriculture, management, and labor, were accurately sampled. I also know that the polls did not represent facts but mere speculation, and I have always placed my faith in the known facts. (Truman 1956)
In an unsent letter addressed to Elmo Roper responding to Roper's postelection poll results dated December 30, 1948, Truman writes, (1)
It [Roper's report] is interesting, but it still misses the main point. Candidates make election contests, not pole [sic] takers of press comments by paid column writers. Edited news columns and misleading headlines have some effect--not much. People in general have lost faith in the modern press and its policies. That is a good thing too. No one segment should be able to control public opinion. Leadership still counts.
Truman's opinion on polls is consistent with his definition of leadership. Truman is often quoted as saying,
I wonder how far Moses would have gone if he'd taken a poll in Egypt? What would Jesus Christ have preached if he'd taken a poll in Israel? Where would the Reformation have gone if Martin Luther had taken a poll? It isn't polls or public opinion of the moment that counts. It is right and wrong, and leadership--men with fortitude, honesty and a belief in the right that makes epochs in the history of the world. (Hamby 1995)
Based on these public and private assertions from Truman, it seems he was a president who led an administration that did not rely on public opinion polls on any level.
The Archival Data
To assess the Truman administration's poll usage for this article, an attempt was made to view all documents at the Truman Library relating to public opinion polling during Truman's tenure as president (from 1945 to 1953). These documents include poll results from a variety of sources, memos detailing the utilization of public opinion polling in any form, and handwritten notes documenting the flow of material to and from individuals in the White House concerning polling. Oral history interviews conducted with members of the White House staff and other individuals connected to Truman were also surveyed for references to public opinion and public opinion polling.
Papers of key individuals who served as administrative secretaries, press department officials, policy advisers, or department secretaries were systematically selected for review with the assumption that these individuals would be the most likely to use public opinion polling data and would have access to these data. General papers of the president were also surveyed for evidence of public opinion polling, including the White House Central Files, the White House Official File, the President's Personal File, Public Opinion Mail, and the White House Telegraph Office Files as well as the papers of the Democratic National Committee (DNC).
A counting technique created by Heith (1998) was employed to aggregate the total number of polling memos and private polls uncovered in the papers of the Truman administration and to demonstrate the depth and degree of poll consultation. To be counted as a "polling memo," a memorandum "could refer to a public poll, cite recent popularity figures, incorporate poll data into an argument, instruct others to use poll data for policy making, or discuss creating poll questions" (p. 166). The archival search turned up a total of eighteen polling memos (six in the White House, ten in the 1948 election, and two from the DNC) and three private polls (one from the White House and two from the 1948 election). These data demonstrate that the Truman administration was less concerned in general with polling than previous or subsequent administrations, but specific analyses of these memoranda illuminate the manner in which these few polls were used.
Polling in the Truman White House
The duties of amassing and interpreting the daily public opinion reports fell primarily to a few key members of Truman's press and domestic policy staff, whose members were among Truman's closest advisers. This is in keeping with other presidential administrations (Jacobs and Shapiro 1995; Heith 1998; and Eisinger 2000). Weekly clippings of public opinion reports (primarily Gallup polls ranging in issue from Truman's job approval ratings to domestic policy issues) were kept by these staff members. (2) Various unsolicited poll results taken by average Americans, daily newspapers, and university public opinion research groups were sent to the White House throughout Truman's tenure as president.
The most prevalent type of polling was unsolicited advance poll results from major daily newspapers and university polling organizations from across the country. Joseph Short, secretary to the president, in a memorandum dated May 3, 1951, thanks the editor of the Minnesota Poll for an advanced release of the poll's methodology and results. (3) Nebraska Congressman Eugene O'Sullivan reported poll results from the Omaha World-Herald, "a very reactionary newspaper," on issues ranging from Agriculture Secretary Brennan's farm plan to the North Atlantic Pact. (4) Charlie Ross, the president's press secretary, thanks James Milloy for an advance copy of the release of the "Sunday Tribune [Chicago] poll." (5)
University and private polling organizations were also generous with their advance results. George Gallup personally sent the president, despite the president's obvious dislike of polls, advance releases from the American Institute of Public Opinion (AIPO). (6) Gallup and W. J. Gaskill, associate editor of AIPO, were thanked on numerous occasions by staff members including Truman's press secretary and numerous special assistants. (7) Eminent social scientist Rensis Likert at the Institute for Social Research at the University of Michigan provided Truman with a report completed at the institute along with a handwritten note at the bottom explaining how the institute's methodology differed from the other polls. (8)
The communication of polling information was multidirectional between Truman's staff and the DNC. This practice is in contrast to that of other administrations (Heith 1998). In a memorandum from John Steelman, assistant to the president, to Frank McKinney, the chairman of the DNC, Steelman thanked McKinney for sending a report of the Research Division on the summary results of various opinion surveys. (9) Additionally, Joseph Short sent a memorandum and results from a Gallup poll to the new chairman of the DNC, Stephen Mitchell. Short wrote,
I am sending you an interesting analysis of the Gallup Poll. It shows that Gallup apparently is again crediting the GOP with more support than they really have, and the Democrats with less. The President has read it and hopes that the public's attention may be called to what is going on. (10)
The White House and the DNC were jointly interested in maintaining positive press and the accuracy of the information on key electoral and governing partners.
The president, despite well-documented entreaties, saw some of the polling results that were sent to the White House after they were funneled through his staff. Assistant Press Secretary Eben Ayers sent a memorandum to William Lydgate, editor of ALPO, thanking him for the Gallup poll he sent and assuring him that it "has been brought to the attention of the President." (11) Charles Ross sent a memorandum thanking the editorial staff at the Women's Home Companion for their advance proof on the question, "What five American men and women now living do you admire most?" A small typed note attached to the memorandum reads, "Note: Proof handed to the President by Mr. Ross." (12) A memorandum from C.G.R. (Charlie Ross) to the president, detailing a Washington Post Public Opinion News Service report written by George Gallup titled "One-Sided Opposition of Labor Leaders to Truman Proposal Not Reflected among Rank and File" states, "This is very reassuring. It is to be published tomorrow." (13)
Key political aides to the president consulted public opinion polling related to the president's political and policy agendas. The most telling of this evidence is in a memorandum dated May 29, 1951, addressed to "Mr. [Averell] Harriman, Mr. [Charles] Murphy, and Company" from George Elsey. (14) In the memorandum, Elsey made recommendations for a public crafting of Truman's rhetoric over the MacArthur firing. In it, he wrote,
I don't like to be a kill-joy but I wonder if we aren't a little too optimistic about the way things are going on the MacArthur row. I think we may be over-optimistic about the extent to which the public understands the President's position and sympathizes with the Administration in the firing of MacArthur. I don't see that we have any cause for joy because conservative newspapers like the Herald Tribune side with the President on this issue. The Herald Tribune's readers will always oppose the Fair Deal and all it stands for. The Gallup Poll published on May 16th contained what I thought were some very disturbing statistics. It shows that the vote of approval for the President's action in firing MacArthur goes down steadily with the amount of formal education. Gallup tested a cross-section in "Who's Who" and found that 51% approved the President's action and 46% disapproved. However, only 25% of the general public approved and 66% disapproved. It seems to me that we cannot afford to slack off in our constant emphasis and reiteration that Mac Arthur stood for war and the President stands for peace. This and this alone will sink in with the general public, while technical arguments about "civilian control" won't mean a thing to the people at large.
Similarly, in reaction to the single private poll (from the Schwerin Research Corporation) found among the White House papers, Ken Hechler wrote to George Elsey on August 7, 1950, (15)
The questions asked were in some cases too broad to make the answers of real significance. For example, on page 13, "The audience members were also asked after the speech, what they thought about the powers sought by the President to regulate industry." The answers are given in terms of asked for too much power, asked for enough power, or didn't ask for enough power. These answers are too generalized, and the questions should be broken down to cover the various steps which the President discussed and the specific powers that he requested. I am a little puzzled by the audience reaction to the question on page 9, and the fact that 49 percent of the audience felt the President's statement on communism should have been stronger. It would be useful to ascertain why this percentage of the audience felt the statement was not strong enough, and the kind of thing they had in mind in order to make it stronger.
These memoranda illustrate that Truman's staff was aware of public opinion polling as it related to rhetorical techniques attempting to identify public opinion on important political matters on which the public might be persuaded. (16) This growing level of political sophistication slowly took root in the staff's approach to the maintenance and governing of the country and demonstrates the permeation of public opinion polling into the White House.
Polling in the 1948 Election
As early as September 1947, the Truman administration was planning its campaign offensive for the 1948 election. A particularly prescient confidential planning memorandum written by Clark Clifford, who served formally as special council to the president but was informally the top strategist in the 1948 campaign, titled "The Politics of 1948" charted the strategic course for the reelection of the president. (17) In this memo, Clifford outlines the probabilities concerning the Republican nominee (as reported by the Fortune Poll), the geographic targets of the campaign, the likelihood of Henry Wallace as a third-party candidate, and the "mechanics" of the campaign structure including the necessity of a "brain trust." The most interesting suggestion in this memo is Clifford's interpretation of the necessity of obtaining and using polling results to assist the campaign. In the memo, he wrote,
If the private Princeton poll can be made available to the Truman Administration (as it was to the Roosevelt Administration) tests would be made of certain questions and submitted to various groups or in certain geographical areas; these would show where hard political work is necessary. For example, the attitude toward President Truman of the Negroes in Harlem [sic] or the farmers in Iowa, or the Italians in Detroit would be scientifically checked at regular intervals. This poll was useful in the 1940 campaign; it can be applied to general political issues as well as to groups.
Clifford's directive did occur, and Clifford's papers include a number of these polls from a variety of sources on a variety of subjects. These items ranged in scope from newspaper clippings of Gallup polls to extensive survey methodology and analysis from major polling organizations at universities. (18) The Dunn Survey from Greenwich, Connecticut, titled 1948 Pre-Election Analysis" reported that Truman would win the 1948 election and the Democrats would control both the House and Senate. (19) The National Opinion Research Center at the University of Chicago conducted a survey sample that covered a range of questions concerning perceptions about the stands of the parties and the most important issues of the day. (20)
The most telling poll analysis (likely commissioned by the DNC for Truman's campaign) in Clifford's papers reports a sophisticated statistical analysis integrated with political campaign strategy pertaining to the election. (21) The introduction to this poll indicates the sample was a national sample of 1,287 individuals and was conducted to "determine baseline measures of attitude by which the changes in political leaning during the next important year of campaigning could be watched." The major findings of this poll were that, contrary to Gallup reports, the election was an open issue in the public mind, independent voters were persuadable with Truman's record, and there were no solid partisan blocks of voters. Specific voting intentions of the public's tendency to support one party or another and the satisfaction with specific administration policies were surveyed. The language and the conclusions suggest that this poll was conducted on behalf of the Truman campaign with a strategic directive as the conclusion. The existence of this type of poll result demonstrates that the administration, especially the staff during the campaign, was aware of public opinion polling and likely consulted this poll to direct campaign strategy.
The DNC and the Truman campaign also shared polling information during the campaign. William Batt, director of the DNC Research Division, and Clark Clifford exchanged polling information on a range of subjects from the candidacy of Henry Wallace to the Gallup and Roper horse race polls. (22) The most important issue tracked was the decision to politically exploit the failings of the Republican-run Eightieth Congress, the main theme of Truman's rhetoric during the campaign. In a memo from Batt to Clifford titled "How the President Can Reach the People" dated July 22, 1948, Batt identifies the communication of the failure of the Eightieth Congress to act on Truman's proposals as the most important strategy the president should employ. (23) Batt also reports to Clifford in a July 22, 1948, memo that Syracuse, New York; New Orleans, Louisiana; and Pittsburgh, Pennsylvania all participated in a telephone poll "Opinionaire" by Mutual Broadcasting System on the question, "Should there be a special session of Congress?" (24) Batt reports that the average of the yes response was 61 percent.
Although causality cannot be ultimately determined, both the Clifford memos and the DNC polls helped to shape Truman's 1948 electoral strategy, a strategy that centered around placing the blame directly at the doorstep of the Eightieth, "do nothing," Congress and focusing on key constituency groups. Public polls indicated that Truman could not win the election, but the private polls conducted by the DNC on Truman's behalf predicted otherwise. Truman's strategies of courting independent voters, making a "whistle stop" campaign tour of key battleground states in the Midwest, and exploiting the Republican-led Congress likely resulted from the administration's private polling.
The archival evidence illustrates that despite Truman's distaste for polls and scholarly accounts arguing that the Truman administration was not as substantively interested in polling as previous or subsequent administrations, Truman's staff was often working behind the scenes with the nascent science of opinion polling to further the objectives of the administration while in the White House and during the 1948 campaign. This evidence is critical in evaluating and explaining the development and institutionalization of public opinion polling in the administrative apparatus of the White House because it fills an important empirical gap in the literature. The consultation of polling by the Truman administration, although dwarfed by the poll use of Franklin Roosevelt before Truman and by presidents subsequent to Truman, is nonetheless an important point in the development of the public opinion apparatus in the White House.
Brace, Paul, and Barbara Hinckley. 1992. Follow the leader: Opinion polls and the modern presidency. New York: Basic Books.
Cornwell, Elmer E., Jr. 1965. Presidential leadership of public opinion. Bloomington: Indiana University Press.
Eisinger, Robert M. 2000. Gauging public opinion in the Hoover White House: Understanding the roots of presidential polling. Presidential Studies Quarterly 30:643-61.
Eisinger, Robert M., and Jeremy Brown. 1998. Polling as a means toward presidential autonomy: Emil Hurja, Hadley Cantril and the Roosevelt administration. International Journal of Public Opinion Research 10:237-56.
Geer, John G. 1996. From tea leaves to opinion polls: A theory of democratic leadership. New York: Columbia University Press.
Hamby, Alonzo L. 1995. Man of the people. New York: Oxford University Press.
Heith, Diane J. 1998. Staffing the White House public opinion apparatus: 1969-1988. Public Opinion Quarterly 62:165-89.
Heller, Francis H. 1980. The Truman White House: The administration of the presidency, 1945-1953. Lawrence: The Regents Press of Kansas.
Herbst, Susan. 1993. Numbered voices: How opinion polling has shaped American politics. Chicago: University of Chicago Press.
Jacobs, Lawrence R. 1992. The recoil effect: Public opinion and policymaking in the United States and Britain. Comparative Politics 24:199-217.
--. 1993. The health of nations: Public opinion and the making of American and British health care policy. Ithaca, NY: Cornell University Press.
Jacobs, Lawrence R, and Robert Y. Shapiro. 1995. The rise of presidential polling: The Nixon White House in historical perspective. Public Opinion Quarterly 59:163-95.
Kernell, Samuel. 1986. Going public: New strategies of presidential leadership. Washington, DC: Congressional Quarterly.
Truman, Harry S. 1956. Memoirs of Harry S. Truman, 1946-52: Years of trial and hope. New York: Da Capo.
(1.) Harry S. Truman Library (hereafter, HST Library), Papers of HST, President's Secretary's File, Longhand Personal Memos, 1945 to 1952, Box 333.
(2.) HST Library, Selected Public Opinion (Gallup poll), 1947 to 1948, Box 1.
(3.) HST Library, Papers of Harry Truman, Official File, Short to Goldish, May 3, 1951, Box 1373.
(4.) HST Library, Papers of Harry Truman, Official File, O'Sullivan to Mr. President, May 8, 1949, Box 1373.
(5.) HST Library, Papers of Harry Truman, Official File, Ross to Milloy, May 31, 1945, Box 1373.
(6.) HST Library, Papers of Harry Truman, Official File, Gallup to Truman, December 28, 1949, Box 1373.
(7.) HST Library, Papers of Harry Truman, Official File, Steelman to Gallup, July 30, 1947, Box 1373; HST Library, Papers of Harry Truman, Official File, Ross to Gallup, January 5, 1950, Box 1373; and HST Library, Papers of Harry Truman, Official File, Ross to Gaskill, February 5, 1947, Box 1373.
(8.) HST Library, Papers of Harry Truman, Official File, Likert to Steelman, May 16, 1949, Box 1373.
(9.) HST Library, Papers of Harry Truman, Official File, Steelman to McKinney, March 3, 1952, Box 1373.
(10.) HST Library, Papers of Harry Truman, Official File, Short to Mitchell, August 28, 1952, Box 1373.
(11.) HST Library, Papers of Harry Truman, Official File, Ayers to Lydgate, July 9, 1945, Box 1373.
(12.) HST Library, Papers of Harry Truman, Official File, Ross to Lane, October 19, 1945, Box 1373.
(13.) HST Library, Papers of Harry Truman, Official File, C.G.R. to President, January 3, 1946, Box 1373.
(14.) HST Library, Papers of Harry Truman, Official File, Elsey to Harriman, Murphy and Company, May 29, 1951, Box 1373.
(15.) HST Library, Papers of George Elsey, Hechler to Elsey, August 7, 1950, Box 1.
(16.) Precise claims concerning how this polling was used or whether it had any impact on shaping decisions made is not entirely provable; however, the presence of this polling represents credible evidence of the staff's interest (which demonstrates the institutionalization of polling) and as a factor in decision making.
(17.) HST Library, Papers of Clark M. Clifford, Confidential "The Politics of 1948" Memo, 1947, Box 22.
(18.) For an example of the Gallup clippings, see HST Library, Papers of Clark M. Clifford, Sulzberger to Clifford, November 29, 1948, Box 22; HST Library, Papers of Clark M. Clifford, Clynes to Houck, July 29,1947, Box 22.
(19.) HST Library, Papers of Clark M. Clifford, The Dunn Survey "1948 Pre-Election Analysis," Box 22.
(20.) HST Library, Papers of Clark M. Clifford, National Opinion Research Center, University of Chicago, October 1947, Box 22.
(21.) HST Library, Papers of Clark M. Clifford, "The Major Findings" Poll Analysis, no date, Box 22. The archival and research staff at the HST Library presumed that this poll was conducted by the Democratic National Committee on behalf of President Truman despite the fact that no date or source exists on the actual document.
(22.) HST Library, Papers of Clark M. Clifford, Batt to Clifford, July 31, 1948, Box 22; HST Library, Papers of Clark M. Clifford, Batt to Clifford, August 17, 1948, Box 22.
(23.) HST Library, Papers of Clark M. Clifford, Batt to Clifford, "How the President Can Reach People," July 22, 1948, Box 22.
(24.) HST Library, Papers of Clark M. Clifford, Batt to Clifford, July 22, 1948, Box 22; HST Library, Papers of Clark M. Clifford, Bart to Clifford, July 23, 1948, Box 22.
Brandon Rottinghaus is a Ph.D. candidate in political science at Northwestern University. He is currently completing his dissertation on presidential responsiveness to public opinion from Franklin Roosevelt to George H. W. Bush.
AUTHOR'S NOTE: The author would like to thank Benjamin I. Page and Patricia Conley for their continuing advice and support. He would also like to thank two anonymous reviewers for greatly enhancing the final product. The Harry S. Truman Library Institute provided a generous research grant, and archivists Liz Safly and Dennis Bilger offered helpful and knowledgeable assistance. A version of this article was presented at the annual meeting of the Midwest Political Science Association in 2002.
|Printer friendly Cite/link Email Feedback|
|Publication:||Presidential Studies Quarterly|
|Date:||Jun 1, 2003|
|Previous Article:||Our chief magistrate and his powers: a reconsideration of William Howard Taft's "Whig" theory of presidential leadership.|
|Next Article:||Evolution of the modern rhetorical presidency: presidential presentation and development of the state of the union address.|