Voters can be aggregated into a large, multi-member district, with each citizen voting for several candidates (‘at-large elections’), or disaggregated into several smaller, single-member districts, with each citizen voting for only one candidate (‘district elections’). How voters are aggregated not only affects who legislators are accountable to, but also which coalitions achieve representation. Political scientists Michael Hankinson and Asya Magazinnik will use the permitting of new housing to measure the effects of voter aggregation on policy. They will leverage the California Voting Rights Act of 2001 (CVRA) as a conditionally-exogenous treatment that led some city councils to switch from at-large to single-member district elections. The principal investigators will use city-level panel data to measure the effects of this conversion on the amount, structural composition, and affordability of new housing units permitted annually. Then, they will use an original panel dataset of geocoded housing approvals in six cities—three treated (Santa Barbara, Escondido, and Anaheim) and three controls (Santa Cruz, San Buenaventura (Ventura), and Glendale)—to examine the effects of district elections on the spatial distribution of new housing within each city.