The impact of turnout on partisan bias in US House elections, 1972-2018

Objective Partisan bias occurs when votes are distributed across districts in such a way that even if the vote between two parties were equal, one party would win more seats than the other. Gerrymandering is a well-established cause of partisan bias, but it is not the only one. In this article, we ask whether the decline of voter turnout can also influence partisan bias. Methods We modified the Gelman–King partisan symmetry measure to make it sensitive to turnout differences across U.S. House elections from 1972 to 2018. Results We found that turnout variation has caused partisan bias in U.S. House elections in the Democratic Party's favor since at least 1972, though turnout bias has gotten weaker in recent elections. Conclusion While turnout bias can buffer the impact of turnout reductions, it has the potential to dramatically increase the number of seats a party loses when its supporters fail to vote.