The Redistricting Commission May Have Failed, But Redistricting Reform in Virginia Was a Success
The final maps are a credit to the state; even if they could have done a better job protecting communities of interest
Virginia’s Supreme Court on Tuesday approved final district maps for congressional and state legislative elections, and with that, Virginia’s long, messy, novel redistricting process has drawn to a close. In the end, the redistricting commision failed, but the redistricting process was a success. Virginia got fair maps that properly reflect the partisanship of the state. Unfortunately, likely because the maps were drawn by two technocrats with limited public involvement, the map splits up some communities of interest. These maps, in the end, are not A+ maps (very few maps are). But Virginia could have done much, much worse than these solid B / B+ maps.
The final maps fairly reflect the partisanship of the state
The maps this process produced do a great job in mirroring the overall partisanship of the state. 538 and the Princeton Gerrymandering Project both conducted mathematical analyses of the maps and gave them high scores for partisan fairness. 538 found the congressional map to have a +1.7% Democratic efficiency gap, which means there are slightly more “wasted” Republican votes than Democratic votes, and found that the median congressional seat was 2.2% more Republican than the state as a whole. No map can produce a perfect efficiency gap or have a median seat exactly in line with the state as a whole; such small deviations from the ideal for these two metrics (with each pointing in opposite directions) are indications of a very fair map. The Princeton Gerrymandering Project, meanwhile, gave the draft congressional and state House maps an “A” rating for partisan fairness, and found the draft state Senate did not advantage either political party–although it gave the state Senate maps a B for not drawing enough competitive districts.1
The New York Times’ Nate Cohn expects these maps to allow Democrats to gain a congressional seat in 2022 as compared to the previous maps,2 which were Republican gerrymanders that mostly (but perhaps not entirely) broke down. And Michael Li from the Brennan Center for Democracy also found the maps to be fair and unbiased. Every nonpartisan analysis and every nonpartisan expert that has examined these maps has concluded they are fair. If we project these maps out over the next decade (and not just the next election, as many surface-level analyses tend to do), the seat share for each party should about match the parties’ vote share. This is the hallmark of a map that fairly reflects the partisanship of the state.3
The maps were drawn by two technocrats; and it shows in how they split up some communities of interest
While the final maps overall are very solid, the redistricting commission that was the heart of the new system did not create these maps. First, a bit of background: Virginia voters last November passed a constitutional amendment that took authority over redistricting away from the General Assembly and gave it to a bipartisan commission equally divided between citizens and politicians and Republicans and Democrats. That commission tried very hard to reach a compromise and pass maps that were acceptable to all sides, but ended up failing spectacularly. The redistricting constitutional amendment specified that if the commission failed, the task of drawing the maps fell to the Supreme Court of Virginia (SCOVA). SCOVA, in turn, was required to appoint two “Special Masters”–technocrats who are well-versed in redistricting–to assist them in drawing the maps.
SCOVA ended up passing the job entirely to the Special Masters. While the Special Masters did a good job ensuring partisan fairness, they did less well in protecting communities of interest,4 particularly at the state legislative level (the congressional maps do a relatively good job protecting communities of interest).5 The draft maps that the Special Masters released in early December, and particularly the draft state legislative maps, kept communities of interest together at a broad level, but when looking at the neighborhood level, they divided neighborhoods in some areas unnecessarily and with no clear rationale. They looked, frankly, like maps drawn by people who did not have a strong knowledge of the entire state, which is in fact the case.6 The public left numerous comments pointing the issues out on the draft maps, and in some areas, their comments approached something resembling a consensus.
The Special Masters took these public comments regarding communities of interest into account and did a great job addressing them…. in northern Virginia.7 As far as I can tell, the Special Masters rectified every concern that a large number of commenters raised with the draft maps in northern Virginia. They wrote in their memo describing the final state Senate maps, “We made a number of slight changes to the Senate of Virginia map, some of which were significant, particularly in Northern Virginia.”8 Outside of northern Virginia, the Special Masters wrote, “[O]ur changes were minimal.”9 Numerous serious issues outside of northern Virginia on the state legislative maps were not rectified.
The Special Masters at least explained why they did not listen when public comments overwhelmingly asked for a change. But some of these explanations fall short. With regard to concerns regarding the Roanoke / New River Valley area in the state House map, the Special Masters wrote that they did not make any changes because “most of the proposed changes had significant political consequences on the districts.”10 I assume here the Special Masters are pointing out that uniting Blacksburg and Christiansburg, as most of the public comments seemed to ask for, would turn HOD District 41 from a swing district into a strong Democratic district. But the Special Masters themselves admit that the House of Delegates maps are ever so slightly tilted in favor of the GOP.11 Making this district more Democratic would thus seem to help the Special Masters improve on the partisan fairness of the maps. The Special Masters simply should have listened to more of the public comments outside of northern Virginia. All of this is to say that the final maps, while not throwing communities of interest to the wind, could have done a better job protecting them outside of northern Virginia.
Two technocrats drawing maps behind closed doors can only do so much
Redistricting is hard. The Special Masters did not discuss in public meetings the thorny dilemmas they were confronting as they drew their maps, as the (failed) Virginia redistricting commission did. The only input the public got into the process was one round of public comments. Not being able to get public feedback throughout the process undoubtedly hamstrung the Special Masters as they were drawing the maps. At one point, they acknowledged they may have made a change, but that public comments asking for the change came too “late in the game” for them to be able to incorporate it.12
This structure that discouraged deep citizen participation in the mapdrawing is likely what led to the final maps splitting up several communities of interest outside of northern Virginia. And this splitting of communities of interest is what keep these maps in the B or B+ range rather than the A or A+ range. I will discuss in my next blog post on Monday what we learned about redistricting reform from the Virginia experiment; and what we can do to reform the redistricting process to ensure Virginia gets A+ maps next cycle. But for now, we should celebrate, because as politicians around the country shut out the public will and distort districts to advantage themselves and their friends, Virginia got congressional and state legislative maps that mostly do a good job in ensuring the public’s values are reflected in their legislature. Redistricting reform in Virginia succeeded.
The Princeton Gerrymandering Project has not yet analyzed the final maps but said they expect the grades to be the same since the partisanship did not change from draft to final maps.
Democrats would be favored to win five out of the state’s 11 districts in 2022 under the old maps given the Republican-leaning environment, but are favored to win six out of 11 districts in 2022 under the new maps.
The Democrats’ main election lawyer, Marc Elias, also blessed the maps as fair in a tweet he later deleted for unknown reasons.
“Communities of interest” are areas that share similar characteristics whose residents deserve to be represented by one person to ensure their unique needs are properly addressed.
I will not opine on how well these maps allow communities of color to elect candidates of their choice because that is a complicated question which I am probably not qualified to answer.
Special Master Bernie Grofman does not live in Virginia; Special Master Sean Trende lives in northern Virginia and it is not clear he had a strong understanding of the geography of the rest of the state going into this process.
To be fair to the Special Masters, they also did a good job “fixing” Congressional District 10, which no longer takes in suburbs of Charlottesville which have no connection to this northern Virginia exurban district; and they also were able to split the Richmond suburbs only two ways instead of three ways in the congressional maps.
Special Masters’ Memo pg. 21.
Id.
Id. at pg. 39.
The Special Masters write that they created “a small Republican advantage in the House of Delegates.” Id. at pg. 55. Emphasis should be placed on the word “small,” as nonpartisan analyses indicate the House of Delegates map is reasonably fair.
Id. at pg. 13.