As long time readers of this site know, back in 2011, we took a rather detailed look at the redistricting process. While we will once again be looking at redistricting, redistricting is going to be a little bit different in 2021 than it was in 2011.
We are nearing the end of the pre-redistricting period. This period roughly covers the four years preceding redistricting (Summer of the year ending in 7 through January of the year ending in 1). There are two basic things happening during this period.
On the political side, during these four years, we are, in most states, electing the state representatives, state senators, and governors who will pass the new redistricting plans. In other states, we are making decisions about the laws that will govern “independent” commissions that will draw up the new districts. We are now, with limited exceptions for filling vacancies, past this phase. It is too late to make changes to state constitutions to alter the rules for redistricting. In states that leave it to the legislative process, the elections for those positions are done and the winners are in office and have the power.
On the numbers side, the Census Bureau is spending this time period planning for and then conducting the decennial census. For the first part of the period, the Bureau is finalizing the form, planning for how to make sure everybody is counted, and hiring the additional staff needed to actually do the census. In the year ending in ‘0, the Census Bureau is first actually conducting the census — getting questionnaires to households (or as they did in 2020, sending mail to households directing them on how to complete the census on-line) with follow-up efforts to reach those who did not respond to the additional contacts. After the data collection part of the effort ends in the fall, the data analysis part begins with the Census Bureau tallying up the numbers and making sure that it’s data is valid.
Normally, this number process ends in late December of the year ending in ‘0 and we transition to the data release phase. This data release phase usually starts in January with the “national” numbers. Then, over a two month period, in February and March, the Census Bureau releases the local numbers on a rolling state-by-state basis (i.e. about 5-6 per week). The release of the local numbers is what then allows each state to draw its new districts.
Of course, 2021 is not a normal year. The Census Bureau missed its late December deadline for the release of the national numbers. Earlier, this year, the Census Bureau announced a target date of April 30 for the release of the national numbers. As of this point, we have heard nothing further from the Census Bureau. As such, we expect the national numbers within the next week or so. The Census Bureau has also announced that the local numbers will be released — in one batch in September.
So what does that mean for how we will discuss redistricting. In the past, we had very little time to look at the national numbers before the local numbers were released. Thus, we looked at key states as local numbers were released. And the staggered release schedule for the local numbers allowed us to do one or two posts per week doing data crunching for the most interesting states.
This year, we are going to have a long lag time between the national numbers and the local numbers. The national numbers (which will include the total population of each stat) that will be released (hopefully) over the next ten days will tell us how many representatives each state will get. Some states will gain representatives; some states will lose representatives; and some states will stay the same.
Even in the states that stay the same (other than the single member states), there will be some adjustment of the lines after the local numbers are released in September. And, some of these states might be interesting (Kansas and Utah for example). But population shift within the states will tell us what is possible. But, for now, the key focus of attention will be on the states with shifts in representation. We will not be able to give precise answers (again due to population shift in the states), but we can use the 2010 population figures to do a first rough guess at how the lines might look. In the states that lose population, we can look at which district is most likely to get carved up (and whether that means that Republicans or Democrats will lose a seat). In states that gain population, we can take a first guess as to what part of the state is most likely to gain that seat. And, so, over the next two to three months, we will have a series of posts looking at the states where redistricting is most likely to have an impact on the 2022 elections.
As in 2011, redistricting is not going to be a happy time for Democrats. The net result of redistricting in 2011 was that the median district was approximately 3% more Republican than the country as a whole. And that made it difficult for Democrats to regain the House (taking four elections) or to keep the House. This unhappiness is the product of three things. First, we lost several key elections in several states for legislative seats or for governor. While Donald Trump might be unhappy at Brian Kemp for not helping Trump cheat to win in Georgia, Kevin McCarthy is very happy that Brian Kemp is the one with the veto pen in Georgia rather that Stacey Abrams. Second, Democrats are at a geographic disadvantage. Democrats tend to be clustered in urban districts. That makes it easier for Republicans to draw lines that individually look defensible (i.e. a compact district in a geographically distinct area) that is 80+% Democrat) while keeping the one Republican leaning part of the urban area together for one Republican district. Meanwhile, it is almost impossible for the Democrats to draw favorable districts in the rural parts of the state. And, when doing a state-wide plan, if “natural” lines create districts that are 80% Democrat for the Democratic districts and 65% Republican for the Republican districts, the Republicans are going to get a disproportionate number of seats. Third, the states most likely to lose seats are Democratic states, and the states most likely to gain seats are Republican states. (Current estimates are a net swing of five seats from states that Biden won to states that Trump won.)
But, as the next several months of posts will probably show, raw numbers aren’t everything. In some states, rather than trying to gain new seats, it might make more sense for Republicans to shift the lines to preserve their current seats. For example, in Texas, in 2020, the winning candidate in ten of the current thirty-six seats got less than 55% of the vote with Republicans winning seven of the ten. Assuming that Texas gets two new seats, Republicans might decide that it is best to concede one of those two new seats as a Democratic district to move some Democrats out of those seven seats that Republicans barely won. (My first glance at the states that are losing seats does not show a situation in which the majority party is clearly better off conceding the loss of that seat to shore up swing seats as both parties currently have roughly the same number of vulnerable seats.)
Depending upon how things look in a state, we may revisit it after we get the local numbers. For example, if we were to do a post on Louisiana, it would probably note that — due to the Voting Rights Act — the Republicans have very little opportunity to change the current 5-1 split. The exact local numbers will change the exact location of the lines, but it will not make a big difference. On the other hand, which counties are gaining population in Texas and which are losing will make a very big difference as to the possible districts that could survive a court challenge. So some of the larger states that are having significant changes might get a second look in October and November.
For both rounds of posts, there is currently no planned set order. Because all of the numbers are coming at once, there will not be a logical this week is state X because it is the most significant of the five states that are in this week’s release. We’ll probably start with the smaller states and/or the states that are losing seats. Simply put, those states are less likely to have the type of big population changes from the 2010 numbers and, therefore, it will be easier to get a decent first guess using the 2010 numbers. The larger states (especially those which are growing fast enough to gain seats) are more likely to be very dependent on which parts of the state are growing fastest which will require having the local numbers. So the posts on those cases are more likely to be very wild guesses at this point making an early post less significant.
But, our coverage of redistricting will start with a summary post when we get the national numbers — which states are up and which states are down. After that summary post, we will get to playing with some redistricting software available on-line and start rolling out the state posts as we get some idea from that software of the realm of possibilities.
