Picking the Winners: Politician vs. Voter Predictions of Two Controversial Ballot Measures

Both politicians and voters were asked to predict outcomes of two Oregon ballot measures in 1982. As expected, politicians' predictions always were closer to the mark than voters' were. Further, voters showed stronger signs of wishful thinking (the "Looking-Glass effect") in their predictions than did politicians. Using published preelection polls apparently improved politicians' accuracy in 1982, as well as voters' accuracy in a separate 1984 survey. No other sources of data improved predictive accuracy. Findings have implications for theories of representative government and are consistent with a new theory of public opinion. James Lemert is Professor and Director of the Division of Communication Research, School of Journalism, University of Oregon. Public Opinion Quarterly Vol. 50:208-221 ? 1986 by the Amerncan Association for Public Opinion Research Published by The University of Chicago Press 0033-362X/86/0050-208/$2.50 This content downloaded from 157.55.39.104 on Mon, 20 Jun 2016 06:07:15 UTC All use subject to http://about.jstor.org/terms POLITICIAN VS. VOTER PREDICTIONS 209 elected politicians to consider public opinion in decisions assumes that these decision makers' perceptions of public opinion are accurate "enough." But what is "enough"? Though Hedlund and Friesema found that, in each case, Iowa legislators were right much more often than wrong in predicting how their constituents would vote on four ballot measures, these authors interpreted their results as raising serious questions about representative democracy (1972: 4 10-1 1). A major problem with Hedlund and Friesema's study design is that it considered only predictions made by politicians, merely comparing various subsets of them against other subsets. This seems a tenuous basis on which to conclude that the entire theory of representative democracy is threatened. In the absence of a baseline that is external to the sample of elected politicians, it is difficult to say what percentage of accuracy is sufficient. Fortunately, however, the pluralistic ignorance literature suggests-almost by definition-that politicians should at least do better than the mass public. Lemert (1981: 12) has based a key part of his public opinion theory on the as-yet-untested proposition that perceptions of public opinion held by politicians and other decision makers will be more accurate than perceptions held by ordinary citizens, as a result of a series of constraints experienced by politicians. Constraints may include such things as (1) questions from journalists about decision makers' public opinion perceptions and their evidence for them, (2) contrary perceptions offered by rival decision makers, (3) uncertainty as to whether polls or other independent sources of public opinion data might surface at some future time, and (4) at least for elected politicians and their advisors, the continuing opportunity (or threat) presented by the need to face the voters. No such pressures are felt by ordinary members of the mass public (Lemert 1981: 13-15). Without these constraints, the result for ordinary members of the public can be what Fields and Schuman (1976: 445) call the Looking-Glass perception: "In the absence of strong counter-forces, a large proportion of people feel that the world they live in agrees with their own opinion on public issues." A kind of pressure upon ordinary citizens has been proposed by Noelle-Neumann in her writings about the "Spiral of Silence." In her book she tends to assume a kind of "quasi-statistical" accuracy on the part of ordinary citizens, but even there, she notes, inaccuracy can be created through the Looking-Glass effect and through inaccurate media portrayals of dominant opinion (1984: 124-27; 157-64). And in an earlier journal article (1974: 45), she explicitly asserts that the spiral of silence itself can mislead both the vocal, active people and the silent ones about the true distribution of opinion. This content downloaded from 157.55.39.104 on Mon, 20 Jun 2016 06:07:15 UTC All use subject to http://about.jstor.org/terms