Chapter 5 - Welfare I

This chapter uses a simulation model to generate predictions about the effects of gerrymandering institutions on four different measures of voter welfare, with interactions for partisan tides and party polarization. Four representation norms are defined: policy median, compositional, personal, and discursive. The simulation generates three major predictions with respect to these norms: (a) bipartisan maps generate higher average personal and discursive utility than nonpartisan maps, and sometimes generate higher policy median utility; (b) partisan maps occasionally generate higher policy median utility than bipartisan or partisan maps, and compromise between the two on the other three measures; and (c) while extreme polarization is always deleterious to all measures of voter welfare, moderate polarization can sometimes improve discursive utility for many maps under all tides, and can improve other utility measures under strong partisan tides. In an extension, the simulations find that majority-minority districting mandated by the 1982 VRA amendments benefits ideologically extreme minority voters more under the electoral conditions in which the amendments were passed (frequent Democratic majorities, low polarization) than under those in which they were implemented. I. Competing Norms in Redistricting Research Since the advent of the Baker v. Carr line of Supreme Court cases mandating equal population in legislative districts, scholars, politicians, and interest groups have supported numerous different measures and standards for what constitutes fair districting. Implicit in these different standards lie different conceptions for how voters should best feel represented by their government. In advocating for a particular standard, politicians and scholars often broach their own conception of representation norms, speaking past the advocates of other norms, typically leaving it to courts to decide between competing fundamental and non-cumulative values. This chapter uses a simulation model to predict the effects of districting institutions on these competing fundamental values, engaging them each equally and in parallel. Some scholars argue that the most important goal in “fair” districting is to maximize the number of unbiased, competitive seats (e.g. Niemi & Deegan 1978). This goal has been favored by “good government” groups such as Common Cause, and codified into several successful