Neighborhoods, Perceived Inequality, and Preferences for Redistribution: Evidence from Barcelona

Abstract

I study the effects of neighborhoods on perceived inequality and preferences for redistribution. Using administrative data on the universe of dwellings and real estate transactions in Barcelona (Spain), I construct a novel measure of local inequality — the Local Neighborhood Gini (LNG). The LNG is based on the spatial distribution of housing within a city, independent of administrative boundaries, and building-specific. I elicit inequality perceptions and preferences for redistribution from an original large-scale survey conducted in Barcelona. I link those to respondents' specific local environments using exact addresses. I find that one standard deviation increase in LNG is associated with 4% higher perceived inequality, but with (if anything) lower demand for redistribution. Residential sorting explains the counter-intuitive pattern. To causally identify the effects of local environments, I exploit within-neighborhood variation from the rise of new apartment buildings as a quasi-experiment. Exposure to new buildings shifts perceived inequality by 7% and demand for redistribution by 2.5%. I finally exploit cross-neighborhood variation in information about local inequality from a survey experiment to explore whether more distant neighborhoods also play a role. The information treatment does not yield significant effects, consistent with the idea that what is close is more relevant. Neighborhoods shape inequality perceptions and (to a lesser extent) demand for redistribution.