Polarization and the Partisan Purse: Bias in Economic Sentiment, Evidence from Recent U.S. Elections

Abstract

Increasing political polarization in the United States has changed the way in which individuals receive and interpret information. While historically, a multitude of factors inform the choice of an individual’s party affiliation, the relationship between choice input and existing political bias has been complicated in recent elections. Using data from Gallup’s US Daily Poll, this study analyzes how political polarization biases one type of non-political opinion: economic sentiment. Drawing from concepts in behavioral economics (e.g. limited attention and imperfect recall), this paper tests the hypothesis that individuals are biased by their pre-existing party affiliations when making an assessment on the performance of the economy, both sociotropically and egocentrically. This study uses the framework of a difference-in-difference model to estimate the causal effects of an individual’s preferred party being elected, and vice versa, on self-reported economic sentiment and spending. Observed results suggest that political bias is present in sociotropic assessments and weakly present in egocentric assessments. This bias is especially large in magnitude, in both directions (i.e. for both Democrats and Republicans) after the 2016 election. In election periods in which the party in offi ce was reversed after an election, individuals whose preferred party was voted out after an election were more economically optimistic throughout the 3 months before the election. The converse was true for individuals whose preferred party was voted into o ffice. This study also found that the optimism in sentiment for individuals whose party was elected, and vice versa, did not translate to a change in the same direction for reported spending.