Forecasting the Vote: An Analytical Comparison of Election Markets and Public Opinion Polls

The dominant scienti c methodology for short term forecasting in important elections uses trial-heat polls, in which a sample of respondents report their current electoral preferences. Election markets, where self-selected participants trade shares of candidates at prices predictive of election day results, provide an alternative forecasting method. Because this new forecasting methodology often produces forecasts more accurate than those obtained using sample surveys, increasing attention is being paid to it as an alternative (and/or supplement) to polling. However, the methodology is not well understood and lacks statistical justi cation. Nor, for that matter, is the rationale for forecasting using sample surveys completely developed. We develop the justi cation for using election markets and public opinion polls to forecast electoral outcomes, giving, respectively, conditions under which each method performs ideally. For this case, we prove (under the reasonable assumption that market participants have access to poll results) that the mean square prediction error for the market forecast is smaller than that for the poll forecast. The case where the assumptions supporting each method fail is also considered. Often it is reasonable to expect that the best case results hold approximately, in which case the market forecast should also beat the poll forecast. We also discuss the potential use of markets as a general forecasting tool.