Partisan Gerrymandering as Differential Geographic Representational Harm

• In Event: Geography and Representation

Abstract

In Rucho, the Supreme Court rejected as insufficiently “discernable and manageable” all the leading tests for partisan gerrymandering, including those based on partisan asymmetry, the efficiency gap, the mean-median difference, and algorithmic outlier methods. According to the Rucho majority, all these measures suffered from the same fatal flaw, a fundamental conceptual mismatch with America’s historic tradition of geographic electoral districting: whereas each measure essentially asked the federal courts to reallocate power between political parties according to some prescribed relationship between votes earned and seats won, the very practice of electoral districting reflects a prioritization of geographic representational interests over any such allocative votes-seats relationship. As redistricting reform efforts pivot to Congress, ballot initiatives, independent redistricting commissions, and state courts, opponents advance many of the same critiques offered by the Rucho majority. In this article, we address these critiques by offering a new way to conceptualize and measure partisan gerrymandering that is “geographic” rather than “allocative.” Our approach does not rely on any votes-seats relationship. Instead, we consider the extent to which a map differentially burdens the legally cognizable representational interests of disfavored voters, where those interests are defined in terms of the very geographic relationships that the Rucho majority and opponents of redistricting reform cite as justifying electoral districting in the first place. We identify four objective, quantitative measures that capture geographic representational interests well established in state law, judicial decisions, and academic scholarship: (1) proximity to a Congressperson’s district office; and the alignment of district boundaries with those of (2) census tracts; (3) counties; and (4) other electoral boundaries. Any map will necessarily burden these interests to some extent - voters must travel some distance to visit the district office of their congressional representative, and district boundaries must sometimes split census tracts, counties, or other electoral districts to meet legitimate requirements like population equality and VRA compliance. But mapmakers should eschew - and courts should scrutinize - district lines that differentially and unnecessarily impose these burdens on groups of voters defined by partisan affiliation or race. Many current maps do precisely that, to a statistically significant degree, especially when enacted through a legislative process dominated by one party. Specifically, we find evidence across a sample of eight states that the party in power drew lines that disproportionately subjected out-party members to unnecessary divisions of their home counties into multiple congressional districts and imposed incongruities between their congressional district and state legislative district lines. With regard to race, we find that White voters were advantaged as they remained considerably less likely to be subject to divisions of their counties and census tracts or to incongruities with state legislative districts. We also find a pervasive, gerrymandering-fed mismatch between constituents and their closest congressional district offices. These findings strongly suggest, without any appeal to votes-seats relationships, that many maps now in effect constitute illegitimate partisan gerrymanders, failing electoral districting on its own terms.

Authors

• John S. Cover

• David Niven, University of Cincinnati