WDM Gerrymandering Analysis

I’m finally ready to do this. Things are really coming together. Say goodbye to traditional outlier analysis and S/V curves (like anyone liked S/V curves anyway). 1/

@ChazNuttycombe Using 2020-24 input for both maps, the proposed VA plan, with W ~ -16, is about as severe a gerrymander as they come, though states like NC, WI, and others have been measurably more gerrymandered. Your map is fair with a good cluster of competitive districts.

A lot of the anti-DEM gerrymander folks keep saying that the proposed VA congressional map would be the most extreme gerrymander in the country. So let's put that to the test using actual political science. The efficiency gap was the standard developed by political scientists during Rucho v. Common Cause (2019). It's not an entirely perfect measurement, as noted in Keena et al., "Gerrymandering the States," because it would technically rate the 2010s Maryland gerrymander as a fair map, for example (though it does show the 2010s Wisconsin gerrymander as an egregious gerrymander!). Keena et al. argue that the mean/median difference is sometimes a better practice; however, for example, it shows the VA gerrymander as a fair map. My standard for measuring how extreme a gerrymander is, or whether a map is gerrymandered at all, is pretty much one standard or the other, depending on which makes sense. In this case, the efficiency gap is the best standard. Using the 2024 presidential results, we can see in DRA 2020 what the efficiency gap is in the enacted/proposed congressional gerrymanders across the country. The closer to 0, the fairer the map using efficiency gaps. Negative favors DEMs, positive favors GOP. Let's take a look. VA: -24.02% NC: 22.55% MO: 17.65% OH: 15.34% CA: -14.92% IL: -14.60% TX: 14.05% NM: -13.98% FL: 11.79% So yes, the proposed VA DEM gerrymander would be the most extreme gerrymander in the country. In fact, it would be more extreme than the 2010s WI Gerrymander that Rucho was all about, which has a 19-point efficiency gap. Only North Carolina comes close to being an egregious gerrymander.

@nicholastomaino Wow. Looking forward to see the changes ahead as the Post shifts focus.

After four wonderful years at the WSJ, I started today as an editor with the Washington Post's editorial page. Grateful to Adam O'Neal and co. for the opportunity, and excited to join such a talented team.

@thomasjwood @electionstudies Interesting how all curves tend to widen over the years showing increased polarization in all groups. I noticed the curves are not normalized, that is, have more area underneath as time progresses. What does this represent?

Symbolic ideology and education, 1972-2024. Ideological polarization increasingly interacts with education. Data from @electionstudies CDF.

@JeffWice @RedistrictNet @mahoneyw @_rebeccaclewis @YanceyRoy @Danjanison @_aizaguirre @ccampy @davidgoldiner Statistically speaking, the current NY congressional map benefits Rs. That is not initially obvious by using a simple votes verses seats model. x.com/WDMelections/s…

Justice Sotomayor asks 4 Democratic responses 2 GOP effort 2 stop a new New York congressional map by February 19 ⁦@RedistrictNet⁩ ⁦@mahoneyw⁩ ⁦@_rebeccaclewis⁩ ⁦@YanceyRoy⁩ ⁦@Danjanison⁩ ⁦@_aizaguirre⁩ ⁦@ccampy⁩ ⁦@davidgoldiner⁩

@vpapupdates x.com/WDMelections/s…

New: See how the map proposed by Democratic legislators would impact your congressional district. Use VPAP’s district comparison tool to look up your address: vpap.org/redistricting/… Explore the proposed map and see where those districts would draw voters from: vpap.org/redistricting/…

@ElectionCenter_ x.com/WDMelections/s…

DRA bros you know what to do

@VoteHub here's how those data look plotted: x.com/WDMelections/s…

NEW — VoteHub has calculated the 2024 presidential results for the proposed VA map.

@QuiverQuant x.com/WDMelections/s…

BREAKING: Virginia State Democrats have reportedly reached a redistricting deal for a 10D–1R congressional map. This could give Democrats 4 additional House seats.

Here's the plan in dra. You can change the demographics used, 2016, 2020, 2024 etc by clicking the cog on the upper rhs of the page. #viewmap::bc930c25-236f-46a7-bbe9-d8d77e21d011" target="_blank" rel="nofollow noopener">davesredistricting.org/maps#viewmap::…

Virginia is proposing a Dem gerrymander of their congressional districts. Starting with the current plan (bottom), shifting districts closer to V0 brings more to the D side. Given that midterms are likely to shift dists further left, this will be a stable and very effective gm.

@RedistrictNet @DrewSav It gets even bluer using 2020 Pres results

@DrewSav Yes they did tweak the map to even out the populations. Here is the map in DRA: davesredistricting.org/join/3dbd6a84-…

#NEW: The MD House has voted 99-37 to pass the proposed Maryland congressional map. It now heads to the MD Senate. mgaleg.maryland.gov/mgawebsite/Leg…

The pizzamander: With its vote share far from 50% (V₀ = 35%), we apply the BV method to MD, which flags the plan as gerrymandered, with every district benefiting Democrats. Note the fingers extending from red areas, pulling enough D votes from cities to create blue districts.

@BFNationalist @ZacharyDonnini "if one side goes out of their way to shape districts just to give them an edge...that's gerrymandering" How can you tell if they shaped them to give them an edge. Most states have a ~70% D district and a ~70% R district. Are all these gerrymandered?

@WDMelections @ZacharyDonnini Maryland has red areas, like the Eastern Shore, which is a discernable area. You're literally arguing against the meaning of words ffs.

Here’s a link to the Maryland map I created in DRA, based on the plan recommended by the Redistricting Advisory Commission and reported by WBAL. davesredistricting.org/join/acdfce26-…

@ZacharyDonnini @BFNationalist Correct. The average of thousands of randomly generated plans gives Rs ~ 18.9% of the seats which is a 6.5-1.5 plan. The proposed MD plan (reposted below) gives Rs ~ 6.7% of the seats, or a 7.5-0.5 plan giving a one seat-difference between the plans.

@WDMelections @BFNationalist You also have to consider the actual political geography of a state. If you draw a natural MD map from a COI and geographical standpoint you get 6-2. One GOP district each on the east and west sides of the state

@BFNationalist @ZacharyDonnini A. I am saying there are x percentage of gop voters in MD. Just reverse the numbers. B. Explain my assumption that is wrong. C. Show me a proposal where all VA seats are Dem and I'll analyze it. Give me the vote shares of all districts in that plan.

@WDMelections @ZacharyDonnini Hell, why not say that there are x percentage of GOP voters in MD, so they can get a proportional number of seats? The problem is they're drawing districts to give them the most number of seats. By definition, that's gerrymandering.

@BFNationalist @ZacharyDonnini This thread is about MD, not VA. Using simple statistics, an 8-dist plan with districts distributed normally (ie gaussian), you can easily argue such a heavily Dem state, MD, should have all 8 districts favoring Dems, so no gerrymandering. in excel, this is done with norm.inv()

@WDMelections @ZacharyDonnini By redistricting a congressional district held by a GOP Congress member so it'll go Democrat. Btw, is the proposal to make VA have all Democrat seats gerrymandering in your opinion?

@BFNationalist @ZacharyDonnini so how does the plan favor one party?

@WDMelections @ZacharyDonnini I would have hoped an account calling itself "gerrymandering analysis" would know what the word meant but here we are.

@BFNationalist @ZacharyDonnini no

@WDMelections @ZacharyDonnini They're literally redistricting in order to gain seats. That's the actual definition of gerrymandering is it not?

@grok @Emidio89684915 @twitdetoo @DemocraticWins your mistake is that you claim a district just above 50% Dem will elect a Dem 100% of the time. This snippet of DRA vote shares for the eight district show four districts close enough to 50% to have a decent chance of electing an Rep so claiming the map is 8-0 is false.

Yes, that makes sense—Maryland's recent elections show Republicans averaging around 32-35% of the statewide vote (e.g., 32% in 2020 presidential), aligning with Dave's Redistricting data for the proposed map. This differs from registration (~25% GOP), as independents often lean Dem.

BREAKING: Maryland’s Redistricting Advisory Commission has just recommended a new 8 Democrats 0 Republicans congressional map. Let’s go!

@grok @Emidio89684915 @twitdetoo @DemocraticWins V0=35% is the average Republican state vote share (65% Dem). It comes from the daves redistricting map of the actual MD proposed plan in question. Does that make sense?

I based it on the map's design to yield 8 Democratic seats in a state with ~25% Republican voters (per 2025 data), resulting in no GOP representation—similar to critiques in sources like The Hill and Washington Post. Your WDM plot shows mild bias; what vote data does VO=35% use? Views differ, with some calling it a fair counter to other states' maps.