ZIP Codes as Geographic Bases of Representation

Abstract

We consider the desirability of ZIP codes as the basis for configuring congressional districts. There are several issues. First, it is likely that voters lack knowledge of ZIP code boundaries. Second, ZIP codes may not coincide with existing political subunit boundaries, such as cities or counties whose non-fragmentation is legally important for redistricting. Thus, using ZIP codes would add yet another layer of complexity to districting. Third, ZIP codes do not perfectly coincide with larger census units, such as tracts, or even with smaller units such as census block-groups. Moreover, while districts can be drawn using ZIP codes that seem better in many respects than existing maps, the same is at least as true for line drawing using existing census geography and maintaining city and county boundaries to the greatest extent feasible. Fourth, we find the empirical evidence offered by Curiel and Steelman (2018) that ZIP code splits are a major aspect of the ability of voters to identify their legislator and engage in political communications with them to be less than fully convincing, though we applaud their efforts at data gathering and the care with which they have sought to control for potential confounding factors. We further commend them for their emphasis on the need to put operational content into the abstract and multivocal concept of community of interest, a concept which still is undertheorized, especially in the legal context.