Relax: US elections don't drive consumer spending

Credit: Chance Bone

Erik Kobayashi-Solomon | Sep 03, 2015

Sections Economics

It’s not uncommon for opponents of a recently elected candidate to suggest that the mere specter of his policies can spook the economy into a slump. On November 6, 2008, two days after Barack Obama was elected president of the United States, radio host Rush Limbaugh took note of a falling stock market and noted, “[Obama] hasn’t done anything yet, but his ideas are killing the economy.” Although some American voters may take a pessimistic view of a president-elect’s expected policies, research suggests the sentiment doesn’t make them spend less.

Although election results can affect consumer sentiment, particularly among those who oppose the victor, the sentiment doesn’t necessarily affect consumer activity, according to a study by Atif Mian of Princeton University, Chicago Booth’s Amir Sufi, and Nasim Khoshkhou of Argus Information & Advisory Services. The authors acknowledge the strong correlation between the results of the widely followed University of Michigan consumer-sentiment survey and consumption growth, but they point out that correlation does not prove causality. And in analyzing county-by-county US presidential election voting results, University of Michigan sentiment data, and records of planned and actual consumer purchases, they find that ideological opposition to an election outcome may not be a driver of consumer behavior.

The study spanned the four presidential elections between 2000 and 2012 and focused on voters who were ideologically opposed to the winning candidate. The researchers find that the stronger a voter’s opposition to a successful candidate (as measured by a voter-ideology score developed by political scientists Chris Tausanovitch and Christopher Warshaw), the lower the voter’s score on the Michigan sentiment survey.

But from there, the link between sentiment and spending breaks down. For example, when President Bush was elected in 2000, negative sentiment had no effect on consumers’ self-reported spending plans or on their subsequent automobile purchases and credit-card use. Consumers ideologically opposed to President Bush were not happy about the election outcome, but this slump in sentiment had no effect on what they planned to spend, or actually spent.

The election of President Obama in 2008 presented a twist: self-reported planned purchases were significantly lower in counties that demonstrated the greatest ideological opposition to the incoming president, but subsequent actual spending in those counties didn’t change.

The study is notable for its use of granular data to isolate the effects of real changes in economic conditions (e.g., a local housing boom) from the effects of sentiment. However, the increased granularity also limits the extent to which the findings can be generalized—and because the research focuses specifically on sentiment changes surrounding presidential elections, it can’t be extrapolated to other sentiment-affecting events, such as announced changes in central-bank policy.