COMBINING METHODS TO FORECAST THE 2004 PRESIDENTIAL ELECTION THE POLLYVOTE

We present an evaluation of a project to forecast the 2004 presidential election by applying the combination principle, a procedure which in other contexts has been shown to reduce error. This involved averaging within and across four categories of methods (polls, Iowa Electronic Markets quotes, quantitative models, and a Delphi survey of experts on American politics) to compute a combined forecast of the incumbent’s share of the two-party vote. We called it the Pollyvote, signifying “many (methods).” Both approaches reduced error. With the Pollyvote, the mean absolute error was reduced by one third relative to the next most accurate method, the Iowa Electronic Markets, when tested across the 163 days preceding the election. Gains were achieved at all forecast horizons that we tested. On the morning of November 2, the Pollyvote had Bush winning 51.5 percent of the two-party vote, which came within 0.2 percent of the outcome (51.3%).

[1]  James E. Campbell Introduction—The 2004 Presidential Election Forecasts , 2004, PS: Political Science & Politics.

[2]  R. Fair The effect of economic events on votes for president: 1992 update , 1996 .

[3]  Allan J. Lichtman The keys to the White House, 1996 : a surefire guide to predicting the next president , 1997 .

[4]  Ray C. Fair,et al.  The Effect of Economic Events on Votes for President , 1978 .

[5]  Larry M. Bartels,et al.  Presidential Vote Models: A Recount , 2001, PS: Political Science & Politics.

[6]  James E. Campbell Polls and Votes , 1996 .

[7]  J Scottarmstrong,et al.  Forecasting presidential elections , 1986 .

[8]  A. Abramowitz,et al.  When Good Forecasts Go Bad: The Time-for-Change Model and the 2004 Presidential Election , 2004, PS: Political Science & Politics.

[9]  Allan J. Lichtman,et al.  The thirteen keys to the presidency , 1988 .

[10]  V I Keilis-Borok,et al.  Pattern recognition applied to presidential elections in the United States, 1860-1980: Role of integral social, economic, and political traits. , 1981, Proceedings of the National Academy of Sciences of the United States of America.

[11]  Allan J. Lichtman The Keys to the White House , 1996 .

[12]  George Wright,et al.  The Delphi technique as a forecasting tool: issues and analysis , 1999 .

[13]  Larry M. Bartels,et al.  Why Are American Presidential Election Campaign Polls So Variable When Votes Are So Predictable ? , 2004 .

[14]  James B. Lemert Picking the Winners: Politician vs. Voter Predictions of Two Controversial Ballot Measures , 1986 .

[15]  James C. Garand,et al.  Before the vote : forecasting American national elections , 2000 .

[16]  Forecasting the Presidential Vote in 2004: Placing Preference Polls in Context , 2004, PS: Political Science & Politics.

[17]  Herbert B. Asher Polling and the Public: What Every Citizen Should Know , 1987 .

[18]  H. Eyre Introduction I , 1979, New Surveys in the Classics.

[19]  R. L. Winkler,et al.  Averages of Forecasts: Some Empirical Results , 1983 .

[20]  Andrew Leigh,et al.  Three Tools for Forecasting Federal Elections: Lessons from 2001 , 2001 .

[21]  James E. Campbell,et al.  Trial-Heat Forecasts of the Presidential Vote , 1990 .

[22]  J. Wolfers,et al.  Prediction Markets , 2003 .

[23]  S. Bikhchandani,et al.  Learning from the behavior of others : conformity, fads, and informational cascades , 1998 .

[24]  J. Armstrong,et al.  PRINCIPLES OF FORECASTING 1 Principles of Forecasting : A Handbook for Researchers and Practitioners , 2006 .

[25]  De Vera,et al.  U. S. News and World Report シ ニ ミラレル ニホン ノ イメージ , 1976 .

[26]  P. Squire,et al.  WHY THE 1936 LITERARY DIGEST POLL FAILED , 1988 .

[27]  Deconstructing the 2004 Presidential Election Forecasts: The Fiscal Model and the Campbell Collection Compared , 2005, PS: Political Science & Politics.

[28]  Paul Perry,et al.  Certain Problems in Election Survey Methodology , 1979 .

[29]  Koleman Strumpf,et al.  Historical presidential betting markets , 2004 .