What economic/financial variables are most useful in predicting re-election prospects for incumbent U.S. presidents? In the November 2012 revision of their paper entitled “Social Mood, Stock Market Performance and U.S. Presidential Elections: A Socionomic Perspective on Voting Results”, Robert Prechter, Deepak Goel, Wayne Parker and Matt Lampert analyze relationships between prior U.S. economic and equity market performance and incumbent performance in relevant U.S. presidential elections. They focus on incumbent national popular vote margin, but also consider for validation: percentage of total popular vote, percentage of total electoral vote, electoral vote margin and election wins/losses. They estimate annual U.S. stock market returns from November through October based on a modeled Dow Jones Industrial Average (DJIA) starting 1789 dovetailed with actual DJIA returns starting 1897. They look at Gross Domestic Product (GDP), inflation (producer price index, PPI) and unemployment as key economic performance measures. Using specified DJIA, economic data as available and election results during 1824 (first availability of popular vote results) through 2008, they find that:
- There is a significantly positive relationship between incumbent vote margin and prior 3-year DJIA return, with R-squared statistic 0.33 (DJIA return explains 33% of variation in vote margin).
- 1-year, 2-year and 4-year DJIA prior returns are also significant predictors of incumbent vote margin, but with weaker R-squared statistics in the range 0.13 to 0.26.
- Findings are generally robust across subperiods, for other measures of stock market performance and election results, for alternative statistical methods and after adjustments for economic data.
- Findings do not extend to the political party of the incumbent when he does not run for re-election.
- Isolation of incumbent landslide wins and losses in the electoral college confirms a strong relationship between past DJIA performance and election outcome.
- GDP is a significant predictor of incumbent popular vote margin in simple regression, but not when adjusted for DJIA performance.
- There are no significant relationships between incumbent vote margin and prior inflation or unemployment.
In summary, evidence indicates that U.S. stock market performance is a better indicator of U.S. presidential re-election outcomes than economic variables such as GDP, inflation and unemployment.
Cautions regarding findings include:
- Testing multiple lookback intervals for DJIA returns introduces data snooping bias, such that the best-performing interval overstates predictive power.
- Though retaining significance, R-squared statistics for different subperiods are materially different (0.44 for 1824-1900 versus 0.30 for 1901-2004; and, 0.19 for 1824-1924 versus 0.46 for 1925-2004).
- Exploitation of findings requires betting on election outcome or having useful beliefs about how the stock market will react to this outcome (see “Ziemba Party Holding Presidency Strategy Update” and “Party in Power and Stock Returns”).
