Research

The Voting Rights Act of 1965 (VRA) caused dramatic changes in the political economy of the United States South, leading to greater equality between white and Black citizens across a range of political and economic outcomes. In the 2013 case Shelby County v. Holder, the Supreme Court of the United States struck down a core provision of the VRA that had enabled federal oversight of elections in states with a history of discriminatory voting practices, leading to a wave of voter suppression laws that decreased the political participation and political power of racial minorities across the South. Using data from the US Census Bureau, I test whether this decline in political power for minority citizens has increased racial inequality in state-level public spending. I find that municipal governments that were previously subject to preclearance and have larger Black populations saw decreases in state transfers since 2013, compared to similar municipalities with smaller Black populations.

This paper explores how access to in-person early voting impacts turnout, overall and by race. We study ease of access in two senses: spatial proximity to early voting sites and length of the early voting period. We draw on data from North Carolina from 2010-2014. First, using a spatial regression discontinuity design, we find that spatial proximity is associated with increased use of early voting. This increases overall turnout of Black voters, but not white voters, as white voters close to an early voting site substitute away from election day voting. Second, we combine our regression discontinuity design with a difference-in-differences approach to assess the impacts of a 2013 North Carolina law reducing the early voting period by a week. We find that the law specifically reduced turnout of Black voters.  Early voting declined for both impacted Black and white voters, but white voters simply shifted to election day voting. 

The proliferation of voter ID laws across the US has had an ambiguous impact on turnout rates among registered voters, despite fear from civil-rights advocates that these laws would suppress turnout. I advance the literature on voter ID laws by studying how Virginia’s 2014 law impacted both turnout and registration rates, finding significant and durable declines in both measures. To do this, I gather data on registered voters who lack DMV records and track changes in voting precincts over time to identify areas of the state where more people are likely to be impacted by the voter ID law. My findings suggest that the decline in overall turnout rates in voting precincts with higher shares of voters likely to lack valid ID is driven almost entirely by declines in registration rates. Identifying this suggests that voter ID laws have an important deterrent effect that prevents new voters from participating in the electoral process. I also consider the role of countermobilization against new voting restrictions and find evidence that more Democratic parts of the state saw smaller effects from the voter ID law. 

This paper examines the role of local politics in residential sorting and its implications for political polarization. Using mayoral elections between Democratic candidates and non-Democratic candidates, coupled with individual-level migration data in North Carolina, we explore the impact of local official elections and subsequent policy changes on the migration decisions of political partisans. Preliminary results show that the election of a Democratic mayor boosts the overall population of municipalities in North Carolina. This increase is attributed to a decrease in out-migration and an uptick in in-migration among Democrats. 

This paper examines how the direct election of U.S. senators reshaped political representation and the geographic distribution of resources in the early twentieth century. We assemble a new dataset that links state legislative districts and officeholders (ca. 1900-1925) to U.S. Senate selection mechanisms, county characteristics, and multiple measures of federal and state spending. The design exploits three sources of timing variation: (i) states that implemented de facto popular-election procedures prior to 1913, (ii) staggered Senate class elections that forced states to reform at different moments, and (iii) within-state comparisons of counties differentially advantaged by pre-1913 legislative malapportionment. We use event-study and difference-in-differences estimates to show whether the 17th Amendment affected senators' ideology to align with state voters and weaken the old rural bias. We study whether resources shifted toward larger, growing counties and away from overrepresented rural areas.

Local governments are often constrained in their ability to set taxes and spending levels, creating a challenge for policy makers to raise adequate funds for local public goods. I propose and test a model for how housing values within a jurisdiction determine preferences for different property tax structures and public goods. Differences in the distribution of housing wealth help determine policy makers’ decisions in whether to raise revenue through either a flat or a progressive tax on property. To test the model, I first demonstrate that California school districts with higher shares of white, wealthy, and highly educated voters are disproportionately likely to propose and raise revenue via flat taxes. Differing levels of support across groups of voters causes the introduction of a flat tax option to distort previous spending levels sustained under a progressive system. I then turn to precinct-level turnout data on individual tax elections and find that flat taxes receive significant opposition from areas with lower home values. My findings suggest that switching from a progressive tax to a flat tax shifts families with low home values to oppose increases in public spending and exacerbates gaps in public goods spending across low- and high-wealth jurisdictions.