Widespread partisan gerrymandering mostly cancels nationally, but reduces electoral competition

What is it about?

Congressional district lines in most US states are drawn by politicians or partisan actors. Many people have raised concerns about “gerrymandering” — that these actors may draw district lines to favor their own party and incumbents. Detecting gerrymandering is difficult due to the number and complexity of factors that influence districts, like the geographic distribution of population and redistricting laws in each state. This makes it statistically challenging to separate the partisan effects of redistricting from the effects of these other factors. In our paper, we address this problem by comparing recently drawn 2020 congressional districts to large sets of alternative simulated non-partisan plans. These plans are drawn using fine-grained data on population, election outcomes, and incorporate algorithmic constraints to implement state-specific laws. For example, we respect Iowa’s law that districts be constructed only from counties, Ohio’s uniquely complex rules about splitting cities and counties, and make sure our plans reasonably comply with the Voting Rights Act by ensuring electoral representation for Black voters when their preferences diverge from White voters (as maintained by the Supreme Court in Allen v. Milligan). We find that the majority of states show evidence of partisan gerrymandering in the 2020 redistricting cycle. However, these biases mostly cancel at the national level, giving Republicans two additional seats on average. That is, seat gains made by Republicans in some states are canceled out by seat gains in other states for Democrats. We show that Republicans made large gains in states like Texas, Florida, and Ohio by packing urban Democrats. In contrast, Democrats made many smaller gains in states like Illinois and several maps drawn by courts in states like North Carolina and Pennsylvania had a Democratic bias.Consistent with other past work, we also find that geography and redistricting rules separately contribute a moderate pro-Republican bias. Although these biases cancel out to only a handful of seats at the national level, we find strong evidence that gerrymandering also reduces electoral competition at the district level. First, we show that gerrymandered districts make elections less competitive by making more safe seats that are usually won by a member of the same party with a comfortable margin. Second, we also show that gerrymandered districts make Congress less responsive, by decreasing the rate at which changes in voter preferences translate into changes in election outcomes.

Featured Image

Photo by Darren Halstead on Unsplash

Photo by Darren Halstead on Unsplash

Why is it important?

Legislative redistricting plans determine how voters’ political preferences are translated into representatives’ seats. If political parties manipulate the redistricting process to gain additional seats and insulate incumbents from electoral competition, future politicians may be less likely to act in ways their constituents want.