Michael Li and Laura Royden NYT oped:

We conducted an analysis to measure how hard it would be for Democrats in each state to win additional seats under these gerrymandered maps. The results are sobering. In 2006, a roughly five-and-a-half-point lead in the national popular vote was enough for Democrats to pick up 31 seats and win back the House majority they had lost to Newt Gingrich and his Contract With America 12 years before.

But our research shows that a similar margin of victory in 2018 would most likely net Democrats only 13 seats, leaving the Republicans firmly in charge. Just to get the thinnest of majorities in the House, Democrats would need around an 11-point win in the national popular vote. They haven’t come close to winning by that much in a midterm election since 1982.

Of course, every election is shaped by local circumstances. There can be upsets. But because we know that the results in any given district tend to move in tandem with a party’s statewide share of the vote, we can reliably measure how likely it is that Democrats will win a district — and how “responsive” a map is to electoral shifts. The differences among the states are striking.