Club Goods and Voter Turnout in Municipal Elections

• In Event: Participation and Policy Responsiveness in Cities

Abstract

Why are particular neighborhoods and districts within a city well-served by municipal government, while others are forsaken? Geographic inequality in services provision in U.S. urban areas has been well-documented, and previous research has identified politician incentives to target responsive constituencies as a source for this disparity. In this paper, I outline a simple theory of citizen engagement which accounts for this difference in responsiveness, often assumed to be ‘pre-existing’ and left for further investigation. I argue that participation in municipal elections is predicated on the information and stakes that an individual possesses in and about her local political environment, and that both of these are enhanced by the presence of club goods. Consumers of these goods (e.g., parks, libraries, community centers, high quality schooling) evaluate their quality and the role of the municipal government in delivery, and are then more well-positioned to engage in politician selection since these individuals will serve as the caretakers of such goods. In other words, politician supply of club goods creates inequity in engagement, which then leads to differences in demand between resident groups. To empirically evaluate this theory, I make use of a natural experiment on ‘participatory budgeting’ which was sponsored through council district offices in New York City from 2012-2018. Residents within a district vote on proposed projects, some of which are implemented while others are not, setting up counterfactual zones for those ‘treated’ with a project. I compare voters within buffer zones of a short distance around projects which are implemented with voters in buffer zones around projects which are not. Difference-in-difference estimates on participation in the 2013 and 2017 NYC municipal elections show a 1% increase in turnout for those within the buffer zones of implemented projects versus those in the counterfactual buffer zones, across a variety of specifications. Estimates are robust to the inclusion of district fixed effects and race-level covariates. This result suggests the presence of positive feedback loops in urban political geography whereby areas of a city which are well-served continue to receive club goods because of their importance in determining who wins office; once they receive these goods, they become even more capable of political engagement, and increase their turnout. Differences in engagement across a city which may begin as small are magnified as distributive conflict creates geographic winners and losers, and inequalities in provision emerge over time.

Author

• Carl Gustafson, Stanford University