An analysis of the Act 43 Wisconsin Assembly district map using the $\sqrt{\varepsilon}$ test

TL;DR

This paper applies a previously developed statistical test to Wisconsin's Act 43 assembly district map, finding it to be a significant outlier indicative of potential gerrymandering.

Contribution

It demonstrates the application of a rigorous $\sqrt{ ext{ε}}$ test for outlier detection in district maps, extending prior work to Wisconsin's districting.

Findings

The Wisconsin Act 43 district map is an outlier among valid districtings.

Outlier significance ranges from p=0.0002 to p=0.0008.

The test confirms potential gerrymandering in the district map.

Abstract

In previous work [arXiv:1608.02014], we developed a rigorous statistical test for outlier status in a reversible Markov Chain, and demonstrated its utilization with an application to detecting gerrymandering in Pennsylvania's Congressional districting. In this note, we apply our test to the current (Act 43) assembly districting of the state of Wisconsin, and find that the districting is indeed an outlier among the the landscape of valid districtings of Wisconsin. Outlier status is significant at betwee p=.0002 and p=.0008, depending on assumptions.