Anyone who takes this redistricting news and extrapolates that Democrats are likely to keep the House in November is really missing most of the trees that make up the forest. People hastily reading Wasserman’s Donkey-rific take on redistricting might have missed this proviso: “A 42 percent President Biden approval rating could still equate to several dozen losses in November, and Republicans remain overall favorites for the House majority.” “Several dozen” is a lot more than the “two to three seats” Democrats are expected to net from redistricting.

Now it’s true that Biden’s job-approval rating could go up by November, with some presidential skill and luck. But there are only a few occasions (the 1998 and 2002 elections) when the president’s party actually gained House seats in the midterms, or even lost fewer than a dozen House seats (as happened in 1954, 1962, 1986). And Biden is an impossibly long way from the over-60 approval ratings (and positive right-track assessments for the direction of the country) that prevailed on those occasions.

Aside from the downward pressure on Democrats exerted by Biden’s relative unpopularity, the consistency of midterm national House popular-vote swings we’ve seen in both directions recently can’t be wished away.