[Doug64]

I thought of posting this in the Election 2020 thread, but it really deserves its own. Here’s Sean Trende’s two-part take on what the results might be even without extreme gerrymandering, state by state with a running tally.

Alabama is slated to lose a seat in reapportionment under the latest census estimates. Of course, given the deaths and relocations that COVID-19 has engendered, these estimates may be off; if New York loses two seats, then Alabama may be able to avoid its expected loss. Nevertheless, we will assume the most recent estimates are correct. The one Democratic seat remaining is probably Voting Rights Act protected, so it can’t easily be eliminated. We will also assume that potential bids to add an additional African American majority district in Alabama fail. Upshot: -1R

Arizona gains a seat through reapportionment, and accomplishes redistricting via an independent commission. Democrats currently occupy five of the nine seats in the state, but two of those seats are marginal in their baseline partisanship. With two heavily Hispanic districts arguably mandated by the VRA, it is difficult to add another Democratic-leaning seat without flipping some of the more marginal Democratic districts. Assume another Republican seat is added, though if the commission upsets the proverbial apple cart in the state, that number could grow. Upshot: No net change nationally.

Arkansas: Republicans might shore up the increasingly competitive 2nd District, but no net partisan change here.

California loses a seat in reapportionment, for the first time ever. The independent redistricting commission makes the state unpredictable. Republicans only hold eight seats (pending the outcome in a handful of races) so the party would have to be extremely unlucky to see a seat eliminated. On the other hand, a number of Democrats hold seats that are pretty marginal, which they have retained by virtue of their personal brands. A substantial redrawing could endanger three or four of them. Regardless, assume a Democratic seat is eliminated. Net change: -1D.

Colorado gains a seat in reapportionment. Democrats control the trifecta (both houses and the governorship) for the first time in decades, though the existing map is a Democratic drawn one that was accepted by the courts, and two of the Democratic-held seats would become competitive if they were weakened in the next redraw. Make it a Republican seat added, while the two Democratic-held seats are shored up: -1D, +1R.

Connecticut: Democrats lack the two-thirds majority needed to approve a map, but Republicans are inefficiently distributed in this state. No change. -1D, +1R.

Florida adds two seats in reapportionment. Republicans currently have the trifecta and have a 7-0 majority on the state Supreme Court. A lot will depend on how much leeway the court gives the majority party in light of a state constitutional amendment limiting gerrymandering. While an aggressive redistricting could probably knock off four Democrats and add two Republican seats, assume the redraw makes the two new seats Republican, and one Democrat finds his district altered substantially. -2D, +4R.

Georgia: With an aggressive redraw, Republicans could probably make the 6th and 7th districts Republican enough to reclaim them. Assume that they instead combine the Democratic portions of the districts into a fourth heavily Democratic district in the Atlanta area, and create a heavily Republican district farther out. -3D, +5R.

Hawaii: No major changes to the 2-0 Democratic delegation here. -3D, +5R.

Idaho: No major changes to the 2-0 Republican delegation here. -3D, +5R.

Illinois loses a seat. Republican strength is now almost entirely concentrated downstate, and Democrats control almost all of the metro Chicago seats. An aggressive gerrymander might be able to eliminate a downstate Republican district while preserving Cheri Bustos’ district in northwestern Illinois and without weakening Lauren Underwood in the Chicago suburbs. One would probably have to “baconmander” Chicago (the term for stretching districts from city cores into rural areas) or perhaps extend Bustos’ district into Springfield. We’re talking about modest maps, so assume Illinois Democrats opt to consolidate their already successful-beyond-intentions 2012 map. Net: -4D, +5R.

Indiana: It is difficult for Republicans to improve upon their 7-2 map. -4D, +5R.

Iowa: No one knows what the commission will do with this already balanced map. Republicans could go 4-0, or 1-3. Call it no net change: -4D, +5R.

Kansas: An aggressive redraw might dislodge Sharice Davids from her suburban Kansas City district, but assume that Republicans instead try to shore up their existing districts. No net change: -4D, +5R.

Kentucky: With their veto-proof majorities, Republicans control redistricting for the first cycle in modern times. They could conceivably split Louisville in an attempt to create a 6-0 Republican delegation. Assume instead that they concentrate on consolidating. -4D, +5R.

Louisiana: Gov. John Bel Edwards might try to force the drawing of an additional minority-majority district, but assume that he’s unsuccessful. -4D, +5R.

Maine: Jared Golden may lose as the 2nd District continues to shift leftward, but either way, no net change from redistricting or reapportionment. -4D, +5R.

Maryland: Democrats have supermajorities in the legislature, so Republican Gov. Larry Hogan is irrelevant to the process. An aggressive map could result in an 8-0 Democratic delegation (from the current 7-1), but again, we’re assuming modest map draws with no changes. -4D, +5R.

Massachusetts: Democrats can’t improve on this 9-0 Democratic map. -4D, +5R.

Michigan had 19 districts in the 1970s, but will have just 13 after reapportionment. This map will be drawn by a newly empowered commission. The Democrats’ problem is that they now control all of the seats around Detroit, where a Republican-leaning seat currently held by a Democrat probably has to be sacrificed. Unwinding Dan Kildee’s district in Flint (originally drawn as a serpentine Democratic vote sink) could make it very competitive as well. -1 Democratic seat, for -5D, +5R net.

Minnesota loses a seat. Assuming Republicans hold the state Senate, which seems likely as of this writing, this probably winds up as a court-drawn map. The Democrats control the state Supreme Court, but the problem they have is this: The two districts centered on Minneapolis and St. Paul will have to gain a lot of voters, and those voters are likely to come at the expense of the suburban 2nd and 3rd districts. Those districts, in turn, will have to expand into Republican exurbs and rural areas. -1 Democratic seat for -6D, +5R net.

Mississippi: Assuming a court doesn’t order the creation of a second minority-majority district, no change. -6D, +5R.

Missouri: The Republican legislature is unlikely to try to improve upon it 6-2 majority. -6D, +5R.

Montana gains a seat in redistricting. Republicans control the trifecta of governorship and both houses of the legislature, but the state’s geography is such that the Democratic vote is concentrated in the mountainous western portion of the state, with its roots in mining communities. Republicans will likely be in contention in whatever new district they draw, but in assuming modest map drawing, we will say it is Democratic. -5D, +5R.

Nebraska: The legislature will probably attempt to shore up Republican voting strength in the 2nd District, although there isn’t too much they can do here. Regardless, it can’t improve upon a 3-0 Republican delegation. -5D +5R

Nevada: Democrats hold the trifecta, but eliminating the one remaining R seat is tricky. NV3 and 4 will likely be shored up, perhaps turning the second into more of a Republican vote sink. -5D +5R.

New Hampshire: Republicans control the trifecta and may try to weaken Chris Pappas in the 1st District. But the basic shapes of the state’s congressional districts date back to the 1800s and have survived previous Republican trifectas, and the Democratic vote is efficiently spread enough that it is difficult to gerrymander too much. For now, assume no change. -5D, +5R.

New Jersey: This map will be decided by a commission, which traditionally accepts either a Republican-drawn version or a Democratic-drawn one (though in playing for the commission’s votes, the parties probably draw more modest maps than they would with a trifecta). Things can’t really get worse for the GOP here, and the map might create a more favorable outcome for R’s either by drawing more Republican-friendly districts or by changing things up enough that Democrats are effectively running in new districts. Regardless, for now assume no change. -5D, +5R.

New Mexico: Democrats control the trifecta, but risk spreading themselves thin if they try to weaken the Republican-held 2nd District too much. They may try, but as with New Hampshire, assume no change for now. -5D, +5R.

New York is tough to game out. It loses a seat and the map will be drawn by a commission. Because of the state’s geography, the NYC districts will have to expand to take on more residents, creating a chain reaction on Long Island and upstate, as Democratic-leaning suburban districts get pushed further into exurban and rural upstate New York.

If the commission feels it has to maintain an African American majority district in Queens by connecting the current 5th District with non-white populations around Hempstead, Kathleen Rice’s district would probably be de facto eliminated; if not, Sean Maloney’s and Antonio Delgado’s upstate districts would be reconfigured.

At the same time, some of the Democratic “island” districts in upstate New York will have to expand. How that is accomplished could threaten longstanding Democratic holds on districts surrounding Rochester and Albany. This analysis is even further complicated by the question of how much Republican support in upstate New York is Trump support versus traditional Republican support (and how much daylight exists between the two these days).

On top of all this, the state may lose two districts, in which case one of the Democratic-held districts in New York City will probably get the proverbial axe. For now, out of an abundance of caution, assume an upstate Republican gets eliminated. -5D, +4R.

North Carolina: Republicans drew two aggressive maps during the 2010 redistricting. The first was rejected as a racial gerrymander by federal courts, while the second was rejected as a political gerrymander by state courts. Republicans did not appeal the latter ruling, however, so a statewide precedent was not set.

The state Supreme Court would probably not allow an 11-3 map in the state, but may be inclined to accept a 9-5 map (achieved by making a new seat Republican-leaning), which would probably not look like an extreme outlier under various partisan fairness metrics. Assume it will accept this map, for -5D, +5R.

Ohio: The Buckeye State had 24 congressional districts in the 1960s. After this reapportionment, it will likely be reduced to just 15. While it wouldn’t be impossible to eliminate another Democratic seat – in fact, Republicans could probably eliminate two Democratic districts given the realignment of working-class voters and the need for Toledo- and Youngstown-based districts to expand into rural areas – we are assuming very modest redistricting. So, we will eliminate a Republican seat. It is worth noting that if Democrats do not accept this map, under the terms of a recent constitutional amendment the map would only survive for four years. -5D, +4R.

Oklahoma: Republicans will likely alter the 5th District to ensure that there is not another upset win by Democrats. This remains a 5-0 Republican map. -5D, +4R.

Oregon gains a seat. Democrats hold the trifecta, but a number of their seats are shaky as it stands; the 4th and 5th districts have even Cook PVIs. Squeezing another Democratic seat out of the state is possible, but dangerous. Assume instead that the state adds a Republican-leaning seat. -5D, +5R.

Pennsylvania loses a seat. Gov. Tom Wolf will veto an aggressive GOP gerrymander, which will leave the outcome up to the state Supreme Court. Under the terms of its 2018 ruling breaking up the previous GOP map, which minimizes the number of times that localities can be split, it is difficult to maintain the six Democratic districts in the southeastern portion of the state.

On top of this, Mike Doyle’s district in Allegheny County will have to take on new constituents, which will push Conor Lamb’s district further out into Republican-leaning territory. Matt Cartwright’s district will likely have to take on more Republican-leaning areas, especially if the Allentown-based 7th keeps its Democratic lean. Both Lamb and Cartwright won narrowly in 2020, so assume that one way or another a Democratic seat is lost (either by being outright eliminated or as a second-order effect).

This, incidentally, isn’t that different from the current map, which produced a 9-9 split in a very good Democratic year, but which would probably result in a 12-6 Republican majority in a GOP wave election. -6D, +5R.

Rhode Island loses a seat, which by definition must be a Democratic seat. -7D, +5R.

South Carolina: Republicans are unlikely to alter their 6-1 GOP map, except perhaps to shore up Nancy Mace in her Democratic-trending district. -7D, +5R.

Tennessee: In 2010, Republicans redrew a map that shored up newly elected Republicans in historically Democratic districts. Today, those Republicans are entrenched. They could draw an 8-1 map by “pizzamandering” Nashville (pizzamandering is what occurs when a Democratic county is split up into multiple slices, weakening its ability to influence an election outcome). Again, assume non-aggressive maps, with no changes. -7D, +5R.

Texas gains three seats. Republicans control the trifecta, and have shown a willingness to be aggressive in previous redistrictings. At the same time, the goal here is not to test the limits of what parties could do in their redrawings. Assume that Republicans largely consolidate their gains by making the 7th and 32nd districts heavily Democratic, add an additional Democratic seat around Dallas, and create two new Republican seats. -6D, +7R.

Utah has a new independent redistricting commission that will recommend maps to the legislature. The legislature, however, is not obligated to accept them. If Democrat Ben McAdams keeps his seat, the legislature may go along with a map that creates a heavily Democratic district based in Salt Lake City. If GOP challenger Burgess Owens wins the race – which is now pending – it is highly unlikely that Republicans will eliminate this district and instead will look to shore up Owens. Owens is leading right now, so assume no change. -6D, +7R.

Virginia’s map will be drawn by a commission. This would have led to big Democratic gains had Republicans’ original 8-3 Republican map held, but with Democrats controlling the delegation 7-4, the next map seems more likely to weaken a Democratic incumbent than a Republican one. Assume no change for now. -6D, +7R.

Washington: A proportional map here would increase Republican strength in the 8th District, reverting to a 6-4 Democratic majority in the state. Assume that the commission basically maintains the current map, however, leaving a 7-3 map with one Democratic district winnable by Republicans in the right environment. -6D, +7R.

West Virginia loses a district, and Republicans hold all three currently. -6D, +6R.

Wisconsin: Democratic Gov. Tony Evers holds a veto over Republican maps. At the same time, the geography of the state is what it is (as they say) and Democratic strength is heavily concentrated in Dane and Milwaukee counties. Incumbent protection maps are often a compromise when neither party controls the trifecta; such a compromise may involve shoring up Democratic Rep. Ron Kind in the 3rd District (Kind won narrowly against relatively weak opposition in 2016 and 2020). This would maintain the Republicans’ 5-3 edge in the state delegation. -6D, +6R.

Note that this is not a prediction, it is a thought experiment assuming no extreme gerrymandering by either party. The way things are going with the still-extant House races, this six-seat pickup alone might be enough to give Republicans control of the House even if they don’t pick up additional seats, as is traditional for the minority party in an off-year election.

Edit: I screwed up the count, fixed it from a +4 Republican pick-up to +6.

[Doug64]

And it looks like Sean Trende may have called it for Virginia:

Virginia unlikely to see major changes in redistricting

A relatively static population means Virginia’s legislative maps will be less affected by redistricting than any time since World War II, according to a leading member of the state’s new redistricting commission.

Sen. George Barker, D-Fairfax, told commission members at their first meeting Thursday evening that preliminary data show northern Virginia in line for a modest increase, and communities on the North Carolina border losing some representation, but nothing like the upheaval from past decades, when northern Virginia was home to some of the fastest growing suburbs in the country.

The relative stability in population means that redrawing the legislative boundaries “will not be as difficult to do as in the past when there have been these radical shifts,” Barker said.

Barker is one of 16 members on the newly created Virginia Redistricting Commission. Voters established the commission in a referendum last fall. It takes the once-a-decade process of redrawing legislative boundaries for Congress and the state Legislature out of the General Assembly’s hands in hopes of reducing partisan gerrymandering.

In an interview Friday, another commission member, Del. Marcus Simon, D-Fairfax, said that while population shifts might not be significant, that won’t automatically translate to a simpler job for the commission. He said the commission is tasked with drawing fair lines under criteria very different from the partisan motivations that drove past redistricting.

“The old maps were drawn on old criteria,” he said.

The 16-member commission is comprised of equal numbers of Democrats and Republicans. It includes eight legislators and eight citizens nominated by legislators.

The commission faces a daunting task in drawing new lines, given the fact that Virginia is one of only two states that has state legislative races in 2021.

The lines can’t be redrawn until new census data is provided to ensure that the state’s population is evenly distributed among the districts.

In past decades, the Census Bureau has given Virginia and New Jersey, the other state with 2021 legislative races, an early release of population data so the new lines can be drawn in time.

This year, though, census data has been delayed by lawsuits, difficulties completing the count during the coronavirus pandemic and battles over whether the census should address citizenship issues. That has called into question whether the new lines can be drawn in time for primary elections in the summer and the general election in November.

A legislative attorney assigned to work with the redistricting commission told the committee it was “not likely” the commission could complete its work in time to meet deadlines for new districts in the 2021 election.

Barker, who played a key role in drawing the state Senate maps in 2010, provided some hope it could be done. He told the commission that a member of President Joe Biden’s transition team reached out to him. He said the Biden team is well aware of Virginia’s needs, and he urged the feds to provide Virginia its full data by mid-April to early May, if possible.

Simon, though, was pessimistic about meeting the timeline that would be required to allow elections this year under new boundaries. He said Barker “is the only person I’ve spoken with who is optimistic about getting the census data in time.”

Once the commission receives the needed census data, it is supposed to draw the new boundaries within 45 days.

[Doug64]

And just south, Maryland's Republican governor has announced that he is setting up a redistricting commission through executive order. I'm not a big fan of allegedly bipartisan commissions, but it'll be interesting to see how this one works out because right now there's definitely some gerrymandering in that state--Trump won only a single one of Maryland's eight House districts, when the 32% of the vote he received should have netted him two or three.

[Doug64]

So Monday the US Census Bureau has released the numbers for reapportionment, and it's good news for Republicans (and IMHO the nation), but not as good as expected. The expectation was that Republicans would be able to win back the House just because of reapportionment, now not so much. Here's Sean Trende's report:

Census Data Update: Will GOP See Apportionment Gains?

Going into this decennial reapportionment, it appeared that states’ congressional delegations were poised for widespread reshuffling of the deck. New York was on the cusp of losing two seats, while Texas and Florida were in a position to pick up three and two, respectively. Given the legislatures that control redistricting in these states, it seemingly offered substantial opportunities for Republicans to redraw the lines in ways that boosted their chances in the House significantly.

Instead, the reapportionment numbers announced by the U.S. Census Bureau on Monday were something of a wash. Only seven states lost seats while six gained seats.

There were notable outcomes here: California lost a seat for the first time in its history. Rhode Island – widely expected to be reduced to a single-member state – held onto its two House seats (in fact it wasn’t a terribly close shave). New York, even with COVID deaths pushing it toward a loss of two seats, lost just one.

There were significant close calls as well. Minnesota, thought to be losing a seat, held onto its eighth district by 26 people; Montana, seen as a guaranteed winner, received its second seat (which it lost in the 1990 reapportionment) by only 6,000 people. New York was just 90 people away from not losing any seats. Ohio was 11,000 people away (in a state of around 11 million) from holding onto all of its seats.

Even the quiet nature of this reapportionment was historic. The 14 total seats that will be reshuffled is the smallest number in the history of reapportionments. This is true even with newly admitted states excluded, as well as accounting for the growing size of the House during the 1800s.

What will this mean for redistricting? In November of last year I estimated that Republicans would probably gain six seats from a normal redistricting. In other words, without doing anything like trying to squeeze a 13-2 map out of Ohio (which is doable) or break apart Emanuel Cleaver’s Kansas City-based district in Missouri (again, doable), Democrats would lose control of the House.

How does this change things? Here are a list of states that have to be revisited given the reapportionment shifts:

Alabama was expected to lose a seat. Instead, it holds all of its districts, which the Republican legislature will redraw. Upshot: Republicans gain a seat over the old baseline.

Arizona, surprisingly, is not gaining a seat. That seat would have been drawn by an independent commission. The earlier analysis gave it to the GOP, though upon revisiting the state, it likely would have been a Democratic seat in the Phoenix suburbs. At the same time, Arizona’s 1st and 2nd districts will likely remain swing-ish. Regardless, a seat I’d awarded (probably incorrectly) to Republicans won’t be created. Upshot: Republicans fail to gain a seat from the previous baseline.

Florida only gains a single seat, instead of the two many expected. A lot here depends on how aggressive the GOP chooses to be, and how far the state Supreme Court will allow it to go under the fair-districts amendment. Regardless, Republicans likely lose out on a district they were slated to win.

Minnesota, as noted above, appears unlikely (for now) to lose a seat. I assumed that seat would be a Democratic-leaning one. Population shifts may yet force some Democrats into increasingly unfriendly territory, or they could end up with a favorable court-drawn map. Regardless, Democrats won’t lose the seat I’d projected they would.

Rhode Island was widely expected to lose a seat. It won’t. In fact, it didn’t come close to losing it. So Democrats keep a seat they certainly would have lost.

Texas was expected to gain three seats; it will gain two instead. As with other states, a lot here will depend on how aggressive the GOP would like to be. For now, I’ll assume that the party isn’t particularly aggressive (though it probably will be) and won’t draw a seat it otherwise would have won.

So Democrats likely will gain two seats net over their previous baseline, meaning that Republicans won’t be in a position to take the House simply through a “natural” redistricting. Of course, there is still some wiggle room here as there are cascading effects from these changes. As noted above, some of Arizona’s Democratic-leaning districts that likely would have become safe as a result of gaining a seat will likely have to take on some Republican-leaning territory and will be more marginal. There may be similar effects in Florida and especially Texas. Regardless, this is good news for Democrats in their quest to hold the House.

Note, that "Democrats likely will gain two seats net over their previous baseline" means they are likely to lose four instead of six through reapportionment, so Republicans--the party out of power in an off-year election--only need to pick up two more seats to take back the House.

And for the situation more generally, four of the six states gaining seats went to Trump while five of the seven states losing seats went to Biden. That cannot be a comfortable trend for Democrats.