On the Relationship Among Partisan Effects, Electoral Probability, and Economic Uncertainty

Abstract:     
The literature on the political economy of finance suggests that government partisanship affects financial risk. Systematic risk should increase (decrease) as the chances of a Democrat (Republican) winning an election improve. At the same time, however, economic voting theories highlight that financial risk should affect vote intentions, and thus parties’ electoral prospects. We address this endogeneity problem, which has bedeviled previous work on partisanship and financial markets, by using positive Democratic campaign events as an instrument for the electoral probabilities of the Democratic presidential candidate, Barack Obama, during the 2008 presidential election. We find that a 1-percentage point increase in the probability of the Democratic candidate winning can raise systematic risk on the U.S. stock market by 0.096 percentage points (according to IV estimates). An unadjusted OLS estimation finds a 0.061-percentage points increase on average. The corrected estimates are considerably higher than estimates that do not account for reverse causality. In monetary terms, the difference approximates an average market value of about $4 billion. Thus, there exists a strong reversal effect in which economic uncertainty reduces the probability that the Democratic candidate will win an upcoming election.

Number of Pages in PDF File: 53

Keywords: Partisan effects, U.S. presidential election, prediction markets, systematic risk, conditional and implied volatility, instrumental variable estimation

JEL Classification: P16, O16, G38

working papers series