Polarized America: From Political Partisanship to Preference Partisanship

Abstract

In light of the widely discussed political divide and increasing polarization post the 2016 U.S. presidential election, we investigate in this paper whether the polarization of political ideology extends further to the preferences, intentions, and consumption of individuals for commercial brands, media sources, and nonprofit organizations. Using publicly available social media data of over 200 million Twitter users’ brand followerships, we first establish measures of political ideology of brands, which we term preference partisanship. Applying this measure to a longitudinal dataset of Twitter followership we assess the evolution of brand preference polarization. Our results show that the claimed polarization in political ideologies post the 2016 election stretches further to the daily lives of consumers. We observe increased polarization in preferences, behavioral intentions and actual purchase decisions of consumer brands. Consistent with compensatory consumption theory, we find that the increase in polarization post-election is stronger for liberals relative to conservatives. From a brand perspective, we show that brands can affect the degree of their political polarization by taking a political stand. We provide a publicly available API that allows access to our data and results.

Keywords: Political Marketing, Social Media, Data Mining, Political Polarization, Branding

JEL Classification: M31