Research
Job Market Paper
A growing number of states have implemented commissions in order to design political districts, in large part as a response to concerns about partisan gerrymandering. While a significant amount of work endorses the use of independent redistricting commissions in theory, very little research has analyzed the causal effects of implementing redistricting commissions. In this paper, I contribute to our understanding of the role redistricting institutions play in gerrymandering outcomes by evaluating how Arizona's independent redistricting commission affected gerrymandering outcomes in congressional elections. To this end, I examine election outcomes in Arizona between the years of 1982 and 2016; two full redistricting cycles before the commission was implemented, and over one and a half redistricting cycles afterward. I use a novel variant of the synthetic control method, a recently popularized empirical tool for generating plausible control groups when none naturally exist, to facilitate this analysis. I find some suggestive evidence that commission-based redistricting in Arizona may have reduced partisan gerrymandering. While my baseline results fall short of full statistical significance, there is also no evidence that Arizona’s redistricting commission made partisan gerrymandering outcomes worse; at a minimum, it seems to have done no harm where gerrymandering is concerned.
Working Papers
First popularized by Stephanopoulos and McGhee (2015), the efficiency gap prescribes a mathematical formula to indicate whether voters have been strategically concentrated or spread across districts in order to bias legislative representation along partisan lines. The efficiency gap was quickly adopted to supplement existing gerrymandering measures, and was even utilized to overturn Wisconsin’s district maps as an unconstitutional partisan gerrymander. Nonetheless, the efficiency gap has also managed to spark a significant amount of controversy, as its mathematical properties have been increasingly scrutinized by scholars. In particular, the range of potential values the efficiency gap can take is attenuated based on partisan vote shares, which undermines researchers’ ability to make comparisons across states or election years. This paper introduces a modified version of the efficiency gap that addresses concerns about its mathematical properties, while maintaining its key features.
Existing measures of partisan gerrymandering that rely on ex post vote counts are predicated on the notion that parties bias elections in their favor through two strategies: packing and cracking. However, they do not attempt to measure the extent to which parties pursue either strategy, individually. I address this gap in the literature by creating separate metrics for cracking and packing. Intuitively, the extent to which a party pursues either strategy entails a trade off between risk and efficiency. Moreover, I combine these metrics to form a novel measure of gerrymandering, the cracking differential, that is robust to concerns regarding granularity and consistency. I show that the cracking differential indicates gerrymandering trends that are qualitatively similar to an existing metric, the efficiency gap, but has superior performance when there is an imbalance between the two parties’ vote shares.