Gerrymandering and fair districting in parallel voting systems

TL;DR

This paper introduces a new method to evaluate if a districting map favors a particular party in parallel voting systems, applied to Moldova's 2019 elections, challenging previous claims of unfairness.

Contribution

We develop a novel methodology for assessing gerrymandering in parallel voting systems and demonstrate its application to real election data, providing an objective analysis.

Findings

No clear evidence of unfair districting in Moldova 2019 elections

Our method effectively identifies partisan advantages in district maps

Challenging prior assumptions about districting fairness in the case study

Abstract

Switching from one electoral system to another one is frequently criticized by the opposition and is viewed as a means for the ruling party to stay in power. In particular, when the new electoral system is a parallel voting (or a single-member district) system, the ruling party is usually suspected of a biased way of partitioning the state into electoral districts such that based on a priori knowledge it has more chances to win in a maximum possible number of districts. In this paper, we propose a new methodology for deciding whether a particular party benefits from a given districting map under a parallel voting system. As a part of our methodology, we formulate and solve several gerrymandering problems. We showcased the application of our approach to the Moldovan parliamentary elections of 2019. Our results suggest that contrary to the arguments of previous studies, there is no clear evidence to consider that the districting map used in those elections was unfair.