2/43 Our running total starts: 5D-8R (skipping AK, DE, ND, SD, RI, VT, WY, HI, ID, WV).
NB: I try to maintain communities of interest & relatively compact maps; but compromise for D-maximization (may not comply with local rules). And I apply VRA as >50% primary (not total) pop.
First off are some red states with competitive suburban seats e.g. Salt Lake City. I'd rate UT-04 as competitive by default, but would lean D with @RepBenMcAdams' incumbency. #ElectionTwitter
Combining Omaha and Lincoln creates a blue-trending district. But not yet safe outside of a wave, so would default it to toss-up (in a neutral environment). #ElectionTwitter
It is a stretch – but was so astonished that you can get 2 competitive districts out of KS that I had to go with it (Kansas City to Topeka, & K.C. to Wichita via Junction City). #ElectionTwitter
MO-03 would have been competitive a decade ago (like the old MO-03 before it was eliminated). But St Louis’ suburbs can definitely still support two solid Dem seats. #ElectionTwitter
It’ll be very interesting to see whether the trends in Fort Wayne and the Indianapolis suburbs are enough to convert IN-03 and IN-07 into Democratic-leaning seats. #ElectionTwitter
Clinton lost KY-04 by 8.6% but it’s a winnable toss-up thanks to Cincinnati's blue-trend. Not only did Beshear win (57.6-40.5%), so did Stumbo (2019 AG) by 90 votes (50-50%)! #ElectionTwitter
Moving South, some fascinatingly polarized geography. Ds can squeeze two close seats out of West TN; & Trump lost ground in the 3rd/Chattanooga-Knoxville & 4th/Murfreesboro. #ElectionTwitter
Charleston should push the 1st out of competitive; so realistically @RepCunningham would only be vulnerable in a primary. A fairly clean map; solid 4R-3D delegation. #ElectionTwitter
Another gerrymander that just gets Ds a proportional result. My only surprise was that both Dem districts can get over 50% African-American without getting too egregious. #ElectionTwitter
One rural (the Delta), one suburban (Jackson-Hattiesburg-Meridian), both >50% black, and relative clean lines. MS’s inelasticity makes their margins pretty secure. #ElectionTwitter
Amazingly, it is possible to get 3 fairly coherent Dem districts: New Orleans, Baton Rouge, and the Bayou (aka Mississippi and Red Rivers district). #ElectionTwitter
The last of our Mississippi River African-American districts. Promising Democratic trends in Conway (Faulkner County) should counter rural loses to the Republicans. #ElectionTwitter
Heading West, @RepKendraHorn’s seat would become fairly safe. Edmondson (18 Gov) won it 57.4-38.4%; also came close in the 1st (48.2-48.8%) but Tulsa isn’t as large as OKC. #ElectionTwitter
Much easier for @RepTorresSmall to defend this 22.8% Native American NM-02 district.
Would have to see where Pres third party vote goes before confident NM-03 is safe. #ElectionTwitter
CO-03 is the only toss-up, enabled by two very different R vote sinks: ‘Millennium Falcon’ CO-05 (suburban Colorado Springs) & CO-04 (rural Western Slope & Eastern Plains). #ElectionTwitter
@RepHorsford’s NV-04 gets shored up; and @RepSusieLee’s NV-03 stays competitive, but now leans slightly to the left rather than R+2 – a very good bellwether of the state. #ElectionTwitter
Only 4 R seats (far inland)! All 11 swing districts voted Clinton & are trending blue (San Joaquin Valley/10, 21, & Southern California/25, 36, 39, 41-42, 45, 47-48, 50). #ElectionTwitter
From one coastal Democratic bastion to another; this map shores up the 9th district – the only seat currently in any conceivable danger. But not under these lines. #ElectionTwitter
Sticking with the Northeast; Connecticut’s 2nd and 5th are reinforced, while ending the 1st’s fishhook tendrils (Always nice when your map is safer AND more compact). #ElectionTwitter
I figured I’d try to make both districts match the statewide partisanship as closely as possible. A little bit cleaner than the current map; both seats still toss-ups. #ElectionTwitter
I applied the same logic in Maine: each district got a share of R-trending rural and D-trending suburbs, which should cancel out and keep them both balanced. Cleaner too, imo. #ElectionTwitter
Rounding off the 2-CDs club, our new entrant (replacing RI). MT's western 1st is a toss-up; ironically, without containing 1st & 3rd biggest cities (Billings & Great Falls). #ElectionTwitter
Remaining Northwest, the Portland suburbs can fit another Dem district, while strengthening both @RepSchrader’s OR-05, & @RepPeterDeFazio’s OR-04 (though still competitive). #ElectionTwitter
Easy enough to get 7 Dem seats; limited options to cleanly improve WA-03. Managed to shave off a few points with Yakima, but Herrera Beutler is still a strong incumbent. #ElectionTwitter
Continuing with states dominated by a major city: Chicago & the collar counties can support 12 seats. The 13th & @RepCheri’s 17th were trickier, but got them Dem-leaning. #ElectionTwitter
And of course NYC. 3 Republican vote sinks upstate; the rest of the districts lean Democratic (albeit 3 toss-up seats only narrowly so). Plus my NYC 10th & 12th are tidier. #ElectionTwitter
Good to discover that Maryland Dems can get a map that both delivers even more Democratic seats, yet is much more compact than the current spaghetti junction of districts. #ElectionTwitter
Transitioning to the Midwest; felt bad cracking Bucks (worth it). Sacrificing Wilkes-Barre + dividing Pittsburgh (rather than stretch to Erie) avoids dummymanders. #ElectionTwitter
Reversing the current 5R-3D delegation requires splitting Dane (Madison), joining the BOW counties (Green Bay-Appleton-Oshkosh), & relying on water contiguity (WI-08). #ElectionTwitter
I bit the bullet and settled for 2 D-leaning seats. Trying to push the 1st from lean-R to true toss-up just leaves the 2nd and 3rd vulnerable to more erosion. #ElectionTwitter
This 7D-4R-2? map is a very satisfying reversal. NB: I’d expect African-Americans would make up a majority in the Democratic primaries for both MI-09 and MI-13. #ElectionTwitter
I focused on safeguarding 5 seats with the Twin Cities; but @collinpeterson's MN-07 also wound up bluer. Always impressed how Obama did in rural Midwest, in retrospect. #ElectionTwitter
OH-03 & OH-10 restored to their previous (2002-10) positions: Dayton & Cleveland, respectively. The former is the only innate toss-up, but Mike Turner seems to defy gravity. #ElectionTwitter
Another traditional swing-state. 3 of the 5 toss-ups are getting safer: Jacksonville (4) & Tampa (14+15). 2 are trending-R: St Petersburg (13) & 18th (but Nelson 51-49%). #ElectionTwitter
Onto the New Purples! 25-14 D map; 12 seats were right-leaning in 2008: Waco-Austin-San Antonio (20, 39) Fort Worth-Dallas (24, 26, 32, 38) & Houston (2, 7, 10, 22, 29, 37). #ElectionTwitter
The only toss-up is the (surprisingly high Hispanic) GA-07. And with Atlanta’s suburbs trending blue, I imagine it’d just get more left-leaning. #ElectionTwitter
NC is an interesting one: Charlotte will continue to cement NC-08 and NC-10. Dems would have African-American-majority *primaries* in 5 of 9 seats (1, 2, 9, 12, 13). #ElectionTwitter
Remarkably, every district was more R than the nation in 2008 (McCain’s home-state bonus?). Now, the only toss-ups: AZ-01 (Sinema 51.7-45.0%) & AZ-10 (Sinema 52.6-44.5%). #ElectionTwitter
Finally, my sentimental favorite. Assuming wins in the 3 D-trending toss-ups (1, 5, 11); this is a 9-2(!) map. Yes, I’m quite proud of myself. #ElectionTwitter
NB: Methodology: I combined results by year-sets e.g. 2012-16 (Sen), and 2013-16 (Gov).
I used the wonderful ourcampaigns.com to pull county results going back to 1957 (i.e. stopped when Hawaii and Alaska were admitted as states + data quality issues that far back). 3/40
Tested it out with this new KY map. Based on 2016-2019 results, intrigued how it would do in 2020. The #KY04 (R+7 PVI) seems more competitive than the real #KY06 (R+9), with the movement in the Cincinnati and Lexington suburbs. 2/27 davesredistricting.org/join/d389d8e2-…
However, I also did a bunch of Democratic gerrymanders after 2012 redistricting to compare with the real Republican ones. Thought I’d review how my designs performed in light of 2016. The contrast shows just how much of a difference partisan control makes (spoiler: a lot). 3/27