Forecasting presidential elections: When to change the model

Here, we address the issue of forecasting from statistical models, and how they might be improved. Our real-world example is the forecasting of US presidential elections. First, we ask whether a model should be changed. To illustrate problems and opportunities, we examine the forecasting history of different models, in particular our own, which has tried to foresee presidential selection since 1984. We apply what we learn to the question of whether our Jobs model, which offered an accurate ex ante point estimate for 2004, should be changed for 2008. We conclude there is room for judicious, theory-driven adjustment, but also raise a caution about inadvertent curve-fitting. Some evidence is offered that simple core models, based on strong theory, may perform almost as well as more stretched models.

[1]  R. Fair.  The Effect of Economic Events on Votes for President: 1980 Results , 1982 .

[2]  Michael S. Lewis-Beck,et al.  Voters as forecasters: a micromodel of election prediction , 1999 .

[3]  L. Sigelman,et al.  Presidential Popularity and Presidential Elections: An Update and Extension , 1983 .

[4]  Michael S. Lewis-Beck.  Comparaison de Prévision des Élections Présidentielles en France et aux États-Unis , 1995 .

[5]  Alan I. Abramowitz.  An Improved Model for Predicting Presidential Election Outcomes , 1988 .

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

[7]  M. Lewis-Beck.  Election Forecasts in 1984: How Accurate Were They? , 1985 .

[8]  H. Klingemann,et al.  The Oxford handbook of political behavior , 2007 .

[9]  H. Norpoth.  Of Time and Candidates , 1996 .

[10]  Robert S. Erikson,et al.  Temporal Horizons and Presidential Election Forecasts , 1996 .

[11]  D. Hibbs.  PRESIDENT REAGAN'S MANDATE FROM THE 1980 ELECTIONS , 1982 .

[12]  Michael S. Lewis-Beck.  Modelers v. Pollsters: The Election Forecasts Debate , 2001 .

[13]  Josep M. Colomer,et al.  what other sciences look like , 2007 .

[14]  Michael S. Lewis-Beck,et al.  Forecasting presidential elections: A comparison of naive models , 1984 .

[15]  Michael S. Lewis-Beck,et al.  The Future in Forecasting , 1996 .

[16]  Robert Forsythe,et al.  Anatomy of an Experimental Political Stock Market , 1992 .

[17]  Michael S. Lewis-Beck,et al.  Presidential Popularity and Presidential Vote , 1982 .

[18]  M. Lewis-Beck,et al.  The Jobs Model Forecast: Well Done in 2004 , 2005, PS: Political Science & Politics.

[19]  M. Lewis-Beck,et al.  Presidential Election Forecasting: The Bush-Gore Draw , 2002 .

[20]  M. Paldam.  The distribution of election results and the two explanations of the cost of ruling , 1986 .

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

[22]  M. Lewis-Beck,et al.  Economic Models of Voting , 2007 .

[23]  A. Abramowitz,et al.  Bill and Al's Excellent Adventure , 1996 .

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

[25]  Dean Lacy,et al.  Economic Performance, Job Insecurity and Electoral Choice , 2002 .

[26]  M. Lewis-Beck,et al.  National Economic Voting in U.S. Presidential Elections , 2001, The Journal of Politics.

[27]  Thomas M. Holbrook.  Do Campaigns Matter , 1996 .

[28]  Ray C. Fair,et al.  The effect of economic events on votes for president: 1984 update , 1987 .

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

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

[31]  Michael S. Lewis-Beck,et al.  Election Forecasting: Principles and Practice , 2005 .

[32]  M. Lewis-Beck,et al.  Jobs and the Job of President: A Forecast for 2004 , 2004, PS: Political Science & Politics.

[33]  C. Eijk.  Election Forecasting: A Sceptical View: , 2005 .