Exit Polls Influence Voting

3561 Words15 Pages
American Association for Public Opinion Research Do Exit Polls Influence Voting Behavior? Author(s): Seymour Sudman Reviewed work(s): Source: The Public Opinion Quarterly, Vol. 50, No. 3 (Autumn, 1986), pp. 331-339 Published by: Oxford University Press on behalf of the American Association for Public Opinion Research Stable URL: http://www.jstor.org/stable/2748722 . Accessed: 30/10/2012 10:41 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information…show more content…
Are 1 percent of voters affected, 10 percent, or more? The studies that have been done range from showing no effects to showing more than 10 percent of voters potentially affected. Obviously, these cannot all be right, or they may be looking at different elections with different characteristics. An importantissue is the perceived closeness of the election before the election occurredand the peceived closeness of the election based on exit poll results. Jackson (1983), whose paper we discuss in greater detail below, points out very sensibly that the only time the exit polls can have an effect is when they change voters' perceptions about the closeness of the race. This was the case in 1980,when a close race was expected based on preelection polls, but the exit polls showed a Reagan sweep. This was not the case in the 1984elections, where the EXIT POLLS AND VOTING BEHAVIOR 333 exit polls simplyconfirmedwhat the preelectionpolls had alreadyforecasted, a wide Reagan victory. Nor was it the case in the 1976 elections, which had been predictedas close and were also reportedby…show more content…
The Micro Level On a micro level, the paper by Jackson is the only one availablefor the 1980 election. There had been earlier uses of individualdata by Mendelsohn (1966) in an analysis of the 1964 election, but since this had been predicted as a very large victory for PresidentJohnson and the early results indicatedthe same thingthere was no reason to expect an effect, nor was there any. A recent paper by Adams (1985) about voting behavior in Oregon duringthe 1984presidentialelections also found no effects since the substantial victory for Reagan had been anticipated. Jackson used data from the University of Michigan Presidential Study in 1980. Respondents were interviewed before the election to determine if they were registered and their likelihood of voting. A sample of 1814 were reinterviewedtwo months after the election and were asked about their exposure to differentkinds of news coverage: The exit polls, Carter's concession speech, media projections of the winner, election results, and newspaperresults. The actual voting be- SEYMOUR
Open Document