Spatial Allocation of Government Investment: District-Based vs. At-Large Systems

• In Event: Electoral Systems, Coordination, and Congruence

Abstract

Since the 1970’s, a “quiet revolution” has occurred in local politics, wherein a large number of localities have shifted – voluntarily or not – from at-large to district-based elections to select their governing bodies (city councils, school boards, county commissions), especially in the South. More recently, driven by the California Voting Rights Act of 2001, the same pattern has occurred in that state. These reforms have analogous goals to their southern counterparts in the mid-20th century, namely, increasing the presence of Black and Latino elected officials in local councils.

While there is strong evidence that switching to district-based elections improves descriptive representation for racial minorities, in this project we shift our focus to spatial representation. We do so in the California context, assessing changes in cities that switched to district-based elections in the wake of the California Voting Rights Act. Namely, we first assess whether in the absence of district-based elections, elected officials primarily come from one area within the locality. That is, do district-based elections indeed diversify spatial representation on council? If so, when the local government shifts to district-based elections, what are the impacts on policy outcomes and other non-political outcomes, specifically in newly-represented neighborhoods? Evidence about the substantive policy impacts of these electoral reforms is limited in the existing literature and – in part due to scarcity of data on policy outcomes at a sub-city level – focuses on citywide outcomes or differential outcomes by race/ethnicity.

We use a difference-in-differences approach to studying the impacts on newly represented neighborhoods. In this case, we first identify which newly drawn districts have historically (in past several election cycles) not had a councilmember elected from that area. Areas for which that is true are the treatment group. Areas that have consistently had council representation under the at-large structure (by hypothesis: wealthier and whiter areas) are the control group, as they will still have some representation without districts. We then compare a variety of policy and neighborhood outcomes for treated and control areas before and after districting, including – but not limited to – housing prices, the placement of polluting facilities, etc. We hypothesize general improvements in outcomes in the newly represented areas, which we interpret as heightened policy attention to economic, infrastructure, and other policy areas for districts in the treatment group.

Authors

• Michaela Cushing-Daniels, University of Pittsburgh

• Brooke Nicole Shannon, University of Pittsburgh

• Daniel Jones, University of Pittsburgh