Bringing Retrospective Voting Back In: Economic Voting and Latino Electoral Volatility

Latino electoral behavior has proven remarkably volatile over the past two decades, shifting from substantial support for George W. Bush in 2004 to overwhelming Democratic margins under Obama, before moving consistently toward Republicans through 2024. Existing theories rooted in group-identity or ideological alignment struggle to explain this pattern. We argue that retrospective economic voting is an underappreciated driver of Latino electoral volatility. Latinos are structurally positioned for heightened economic responsiveness because they are disproportionately likely to identify as political independents, who lack the partisan motivated reasoning that attenuates the economy–vote link among strong partisans, and are disproportionately concentrated in working-class occupations exposed to macroeconomic fluctuations. We test this argument using three complementary empirical strategies. First, extending a race-disaggregated macropartisanship time series through 2024, we show that Latino aggregate partisan identification tracks economic conditions more closely than White or Black macropartisanship. Second, using Cooperative Election Study (CES) data from 2008–2024, we demonstrate that the Latino independent voting is strongly associated with economic perceptions and that Latinos switch votes and are differentially mobilized in response to economic perceptions at higher rates than Whites. Finally, using objective county-level unemployment data, we show that Latinos in counties experiencing rising unemployment are more likely to vote against the incumbent than Whites in similarly affected counties. Together, our results reframe Latino electoral volatility as partially a predictable consequence of an electorate in which the attachments that attenuate economic voting are structurally weaker.

Recommended citation: Reny, T., Wakefield, D., Dyck, J., & Johnson, G. "Bringing Retrospective Voting Back In: Economic Voting and Latino Electoral Volatility."
Download Paper