Spatial Predictions of State Legislative Candidate Emergence

• In Event: Candidacies, Campaigns, and Behavior in the American Context

Abstract

How much do spatial demographic factors as opposed to electoral factors contribute to candidate emergence running for a state legislative seat?

As more states move toward the use of independent citizen redistricting, studies of candidate emergence also need to adapt. Studies of political party controlled redistricting focus on gerrymandering and representation attributes like partisan bias and responsiveness (Gelman and King 1994) or incumbent protection (e.g., Cox and Katz 2002). These types of studies stress redistricting’s choice of the political features of a district and its impact on candidate emergence.

In comparison, much less is known about redistricting choice of a district’s geographic composition and its impact on the number and types of candidates that emerge (although see Gimpel, Lee, and Thorpe 2011; Myers 2015). Yet race of candidacy studies make clear that the racial/ethnic composition of a district is the primary factor promoting the emergence and success of co-minority candidates (e.g., Fraga, Juenke, and Shah 2020). How might then other general demographic factors also play a role in determining the nature and attributes of a district’s candidate pool?

This question is very important to address for independent citizen redistricting commissions. Behind the scenes stories of partisan redistricting often describe political parties as keenly aware what areas include a strong latent pool of candidates that help (or hurt) a party’s cause. This seems to be an area of knowledge sorely lacking within independent redistricting commissions that explicitly exclude politicians.

In preparation for Michigan’s first independent redistricting effort, this study compares the role that geographic and political factors play in determining the number and type of political candidates for state legislative office in the State of Michigan over the previous district cycle. A key contribution of this study is the new step of combining the State of Michigan registered voter file (7.5 million observations) along with census data from each election cycle to identify the “supply” attributes of a district and how it impacts emergence in comparison to the political “demand” features of a district. By matching observed candidates for state legislative office with registered voter records of each candidate’s address, we will be able to identify what types of neighborhoods are more likely to produce a sufficient number of candidates for office and whether those types of neighborhoods differ for Republican or Democratic candidates.

The findings from this study will be used to generate a set of recommendations for Michigan’s new Independent Citizen Redistricting Commission.

Author

• Corwin D. Smidt, Michigan State University