That’s one of the bottom-line findings that will surprise many in today’s Washington Post article that Nick Stephanopoulos co-authored and blogs about below. The five states most biased toward one party or the other are CA (+4.5 D seats); IL (+3.5 D seats); NC (+2.4 R seats); NY (+2.2 D seats); and NJ (+2.0 D seats).

Using the same metrics, in the first elections after the 2010 Census and round of redistricting, the House had a 27 net seat bias toward Republicans. All four subsequent elections in that decade maintained that R bias, though by lesser amounts: 2014: +7.4R; 2016 +17.6R; 2018 +15 R; 2020 +5 R.

These conclusions are all based on using one particular metric, known as the efficiency gap, for measuring the partisan bias in plans. There are a variety of other approaches to defining a “fair map” in partisan terms, and I have serious reservations about the efficiency gap. Nonetheless, it’s a commonly used metric and journalistic coverage often relies on this metric. So these are important findings and it’s good to see that Nick et. al. have taken their academic work and presented its findings for a general public audience.