Why Is City Living “Liberal”? Political Psychology and Residential Preferences

• In Event: Urban Partisanship, Housing and Elections

Abstract

It is a well-established fact of life in American politics that Democratic voters tend to be concentrated in urban areas, whereas the Republican Party is generally stronger in suburban and rural areas. In recent years, political scientists have given increasing attention to the nature of the relationship between people’s ideology and the characteristics of their place of residence (Cramer 2016; Rodden 2010, 2019; Brown, Mettler, & Puzzi 2021; Jones 2020; Wilkinson 2019; Jacobs & Munis 2019; Williamson 2008). Even after accounting for demographic differences, Gimpel et al (2020) find, significant attitudinal differences persist between residents of urban and rural areas, suggesting that residence within dense (vs. sprawling) areas may itself shape individuals’ political attitudes. An alternative mechanism – residential self-selection, with liberals sorting into denser locales – has been shown to be too limited to explain patterns of geographic polarization (Martin & Webster 2018; Mummolo & Nall 2017).

Even if residential self-sorting is limited, however, individual-level evidence strongly suggests that liberals are more favorable to living in dense and urban environments than conservatives are (Pew Research Center 2021; Martin & Webster 2018). Indeed, political ideology is among the strongest predictors of individual preferences regarding land-use patterns. Whereas liberals express more favorability to living in compact, walkable, transit-oriented living environments, conservatives prefer low-density, driving-oriented suburbs, where residences are separated from businesses (Lewis and Baldassare 2010). How can we account for this ideological split, given that preferences for urban or suburban living are more akin to consumption preferences than to attitudes on public policy?

One possibility is that certain individual characteristics associated with liberalism (conservatism) also are associated with favorability to urban (nonurban) living. In this manner, the association of ideology with urbanism attitudes may be confounded by unobserved individual characteristics that are correlated with ideology. This paper will examine this hypothesis, investigating whether psychological characteristics known to be correlated with ideology–e.g., need for cognitive closure, disgust sensitivity, or personality traits (e.g., liberals’ higher level of Openness) – confound the observed relationship between ideology and neighborhood preferences. An alternative explanation derives from racial politics. Urbanized environments in the U.S. may be perceived as minoritized space. Thus, conservatives’ tendency to dislike the idea of city living may derive from their higher levels of racial resentment (Nelson & Petsko 2021).

The empirical analysis draws from three studies conducted thus far, with a fourth to be completed prior to the conference: (1) An exploratory survey of a relatively diverse sample of 333 undergraduates, which included several validated psychological scales and political characteristics; (2) A brief module included on the nationally representative CCES in 2018 (n=1000), enabling an exploration of the role of racial resentment and need for closure on urbanism attitudes; and (3) An experimental study of 176 undergraduates in which subjects in the treatment condition completed a writing exercise designed to temporarily induce a sense of uncertainty (to heighten need for closure) before indicating their city/suburb preferences. I plan to conduct an additional experimental study during Spring 2023, in which, prior to asking about city/suburb preferences, subjects in the treatment condition will complete an induction exercise intended to bring to thoughts of race (specifically regarding Black Americans) to top of mind.

My dependent variables are of two types: First, a series of three survey questions ask respondents to consider tradeoffs regarding the advantages and disadvantages of living in compact vs. sprawl-oriented neighborhoods (e.g., transit access and proximity to stores and services, balanced against high density and smaller homes) and then state their residential preferences. The second method involves asking subjects to view a series of photographs of neighborhoods – some urban, compact, and mixed-use; others low-density and single-use – and rate the desirability of living and/or working in each. From these photo ratings I have constructed a reliable scale of pro-urban preference, which also correlates at moderate to strong levels with the index of responses to the tradeoff questions.

Author

• Paul G. Lewis, Arizona State University