A Vital Lesson from Political Polls in 1948
1948 Political Poll Data Had Little RelevanceTo Election Results
Nov 1, 2008
If one examines the relevance of opinion polling done in the 1948 presidential race, it is clear that it had limited value to the American public. The 1948 presidential election, which Harry Truman won, proved to be an embarrassment and dealt a serious blow to the survival of the young polling industry.
Poll results off target in 1948
Gallup, Archibald Crossley, and Elmo Roper, the major pollsters of the day, predicted Thomas Dewey would win the election by a margin of 5 to 15 points. However, history shows that Truman edged out Dewey by just over 4 points. Such results brought into question the relevance of opinion polls and threatened their very existence.
One major reason for this miscalculation of the developing polling industry was that the three top pollsters did not believe that opinions would change much, so they stopped interviewing weeks before election day. TheTruman campaign proved them wrong.
Prematurely halting business as usual, the pollsters could not have foreseen collapsing support for minor parties in the race. Only in the late days of the campaign, did people shift and solidify their support of the major parties, especially the Democrats.
In the wake of miscalculation in 1949
Newspaper publishers threatened to cancel the pollsters’ contracts and the pollsters were investigated by the Social Science Research Council (SSRC). The SSRC attempted to pinpoint specifc reasons for such a major miscalculation. At the time, concerns arose whether the pollsters were attempting to influence the outcome of the presidential election as opposed to simply providing an accurate reflection of public sentiment.
Considerable suspicion grew after this election that the polls themselves led to a late surge by the Democrats to increase voter turnout because the surveys showed Dewey well ahead of Truman. On the other hand, believing the poll results, Republicans became overconfident and felt no urgency to get more of their own people to the voting booths. The famous picture of Harry Truman holding up the Chicago Daily Tribuine, with the newspaper's headlines declaring a Dewey victory, reflected the irony of the situation.
Such results of this election brought into question the fundamental accuracy of the presidential opinion polls and questioned their motives in publishing poll results. Such a dramatic failure threatened the very existence of the polling industry.
Presidential campaign of 2008
After 60 years, U.S. political polls are very much alive and well. Pollsters survived the 1948 debacle and taking the pulse of the electorate has developed into a thriving industry. There has been a proliferation of polling enterprises in recent years with several people building successful careers centered around public opinion polling.
In this presidential campaign, voters have been kept apprised of the relative standings of the hopeful candidates on a daily basis since early 2007. The public has been fed poll data as if such results were major news as opposed to simple packaged opinion. However, even though polling procedures have improved and the accuracy of results has sharpened, the relevance of conducting ongoing polls several months prior to any election is a major question. What's the point?
It seems a bit odd when examined in light of the lessons learned by George Gallup and his compatriots in 1948. One of the vital lessons that should have been learned, was that poll results had little relevance several months before election day because public sentiment shifted in the final two weeks. And, if we took the polls in February of 2007 seriously, Hillary Clinton may have been running against Rudy Giuliani.
Moore, David W. The Superpollsters: How They Measure and Manipulate Public Opinion in America. New York: Four Walls Eight Windows, 1995
 |
 |
