Sidak Yntiso (URochester): Data & Democracy Research Initiative Invited Speaker
Please join us for a lecture hosted by DSI’s Data & Democracy Research Initiative.
Data Science Institute
5460 S University Ave
Room 105
Title: Party All the Time? Evaluating the Role of Partisan Redistricting and Measurement Strategies for Racially Polarized Voting
Abstract: Academic and legal debates over minority vote dilution hinge on a quantity we fundamentally cannot observe: individual vote choice. To date, most approaches to estimating racial vote polarization work down from aggregate vote totals and estimates of voting age population by group within each district. Using a formal model, we examine the reliability of these ecological methods when districts can be strategically manipulated for partisan gain. The key feature of our theory is that a district’s racial composition serves as a signal of the decisive voter’s partisan attachments. If a mapmaker is not aligned with the minority-preferred party, she has an incentive to place her party’s strongest supporters in high–minority-share districts—where outcomes are less certain—rather than in similar low–minority-share districts. We show how strategic redistricting affects the correlation between precinct racial composition and voting behavior by group. Next, we simulate electorates under different levels of geographic segregation between voters and different mapmaker objectives. While underlying political geography has a far larger influence on overall model performance, we find that partisan gerrymandering can induce ecological correlations that underestimate ecological estimates of racial vote polarization in maps most likely to dilute minority votes.
Bio: Sidak Yntiso conducts research on the influence of political institutions on racial inequities in the administration of criminal justice and electoral representation, with two overarching goals: (1) offering a theoretically rigorous, empirically grounded account of disparities in criminal justice and (2) evaluating the quality, and tracing the consequences, of racial fairness in electoral representation. His ongoing work focuses on the impact of partisanship on local law enforcement decision-making and the development of tools to diagnose partisan and racial gerrymandering.