I see that Justin already beat me to the punch, but I wanted to add some more thoughts about the implications of yesterday’s elections for redistricting:

1. First off, yesterday was the most successful day in U.S. history in terms of voter initiatives to prevent gerrymandering. Colorado, Michigan, and Missouri all passed anti-gerrymandering measures, and a fourth initiative, in Utah, is narrowly leading but still too close to call. By the standard of past efforts, this is astonishing. In a 2007 paper, I found that about two-thirds of redistricting initiatives to that point had failed, typically because of the intense opposition of the majority party. But now reformers seem to have cracked the code; they have repeatedly managed to turn anti-gerrymandering measures into valence issues instead of partisan contests.
2. Reformers have done so, moreover, while advancing more ambitious visions of what fair redistricting should look like. None of the commissions in place in this cycle affirmatively took election results into account in order to try to achieve partisan symmetry. The Michigan initiative, however, requires that district maps not “provide disproportionate advantage to political parties.” The Missouri measure, similarly, states that “districts shall be designed in a manner that achieves . . . partisan fairness,” in that “parties shall be able to translate their popular support into legislative representation with approximately equal efficiency.” These criteria mean that the Michigan and Missouri commissions, at least, will have to do more than comply with traditional redistricting criteria. They will also have to consider partisan data to make sure that neither party is significantly—even if unintentionally—helped or harmed.
3. Also encouragingly, it’s now highly likely that several states that had severe gerrymanders in this cycle will not in the 2020s. In Michigan, Ohio, Pennsylvania, Virginia, and Wisconsin, there will probably be either divided government (thanks to Democratic gubernatorial wins) or a redistricting commission responsible for drawing the lines. The odds are therefore good that gerrymandering in the next decade will not be as prevalent, or as one-sided, as it was in this cycle.
4. On the other hand, barring judicial intervention, we’re still likely to see some extreme gerrymanders. Democrats will probably have unified control in major states such as Illinois, Massachusetts (since they can override a gubernatorial veto), and New York (where the new commission’s maps are merely advisory). For their part, Republicans will probably have unified control in large states including Florida (where constitutionally prescribed criteria impose some check on gerrymandering), Georgia, North Carolina (since the governor can’t veto district maps), and Texas. These are the places that will likely be the flashpoints of redistricting litigation in the 2020s.
5. Also on the bad news front, several of the existing gerrymanders that are currently being litigated had little difficulty weathering the blue wave. In North Carolina, for example, Republicans seem again to have won ten of the state’s thirteen congressional districts—exactly as in 2016, and exactly as intended by the map’s architects. In Wisconsin, similarly, Republicans appear to have retained their 64-35 advantage in the state house—even as Democrats swept every statewide office. This is further evidence of the durability of contemporary gerrymandering, even in the face of dramatic political developments.
6. Lastly, with respect to Congress as a whole, Democrats look poised to win the national House vote by roughly 7 percentage points, and to translate this advantage into about 230 seats. While a majority is a majority, this outcome is actually quite biased in favor of Republicans. Both historically and in order to achieve a zero efficiency gap, a seven-point win nationwide should yield about 250 House seats for the majority party. The cost exacted by intentional gerrymandering—as well as unintentional factors like the geographic distribution of the parties’ supporters—is thus on the order of twenty seats. This cost did not deprive Democrats of the majority this time, but it did in 2012, and it may well again in 2020.