Andy Koch, Associate Professor of Finance, and Leming Lin, Associate Professor of Finance
Koch: BS, Georgia Tech; PhD, University of Texas
Lin: PhD, University of Florida
For decades, politicians and pundits have repeated the line “it’s the economy, stupid,” when it comes to elections. But does the stock market actually matter for elections? The market is usually seen as forward-looking—it anticipates expected political changes. But can it also shape politics in return?
Politicians and the media often assume that the stock market plays a role in elections. However, establishing this link has been surprisingly difficult. In our recent study titled “Real Effects of Markets on Politics: Evidence from US Presidential Elections” (co-authored with Alan D. Crane at Rice University), we take on the challenge and reveal a novel way stock market fluctuations can impact election outcomes.
The main hurdle in studying this relationship is that the stock market is “forward-looking.” It often anticipates future events, including election results and potential policy changes, making it hard to tell if the market causes voting patterns or simply anticipates them. To overcome this, we focus on differences across counties in how much their residents participate in the stock market. Their key idea is simple: if stock market performance truly sways votes, then the effect of market returns should be stronger in areas where more people are invested in the market.
We use the ratio of dividend income to total adjusted gross income as our measure of stock market participation and interact this with cumulative real stock market returns over the four years leading up to each election from 1992 to 2020. We find that counties with higher stock market participation are more likely to vote for the incumbent party when the market has performed well compared to counties with lower participation.
To put this in perspective, we estimate that if market returns leading up to the 2008 election had been as strong as those from 1997-2000, John McCain, the Republican incumbent party candidate, may have won 242 more counties and carried key states like Florida, Indiana, North Carolina, Ohio, and Virginia, potentially translating to 86 additional electoral votes. The estimated effect is distinct from broader macroeconomic factors like GDP growth, wage growth, or unemployment rates. Even when these macro factors are accounted for, the stock market’s unique influence remains.
Their research is published in the American Economic Review.