The gerrymander map of 2026
Since August 2025, fifteen states have tried to redraw their U.S. House maps in the middle of the decade, or had them redrawn by a court. Nine new maps are in force for Nov. 3. We rank all 44 multi-district states, measure what the redraws did to voters, and set out how the damage could be undone.
Every number carries one label. SOURCED = taken from a published source (linked). ANALYST = Catalyst calculation or model, method stated. GAP = we do not have reliable data and show none. Hover or tap any chart for detail.
1 Which states are most gerrymandered?
The spine of this page is the New York Times Upshot ranking published Oct. 26. It compares each state’s current congressional map with 5,000 computer-drawn, compact maps per state that ignore partisanship, about 220,000 in all. The simulations were taken from or generated with Harvard’s ALARM project, and updated for Callais.
Seats gained is how many more seats the map is expected to give the party in charge than the typical neutral map would. Expected seats are a weighted mix of an even national environment and six-point waves in either direction. Aggressiveness (0–99) is how far out on the tail of the simulations the real map sits.
The NYT ranks the states on both measures together. Texas and California come first and second; Tennessee is the most aggressive map in the country.
Full ranking of all 44 states
Rank, seats gained, aggressiveness, tag and year drawn: SOURCED NYT, as of 2026-10-26 (article). Who controls the drawing is the NYT’s own classification. The NYT tags some states “Watching” (risk of a new map), “Fine” or “Balanced”; we show these where the NYT gives them.
2 The 2025–26 redistricting war, state by state
The fight began in Texas in August 2025 at the President’s urging. California answered with Prop 50. After the Supreme Court’s Louisiana v. Callais ruling on April 29, 2026 narrowed Section 2 of the Voting Rights Act, three Southern states took apart Black-opportunity districts within weeks.
The waterfall adds up the seat shift each in-force map was designed to produce. That is drawers’ intent, not a forecast: in a strong Democratic year, some “safe” new seats can still flip.
Drawer, method, legal status and designed shift: SOURCED Wikipedia redistricting tracker (compiles official actions and rulings), as of 2026-10-28. 2024 delegation: SOURCED certified 2024 results. “Designed shift” is the change in seats the map targets, not the outcome.
3 Before and after: the district lines
Real boundaries only, from Census and state-legislature shapefiles (sources under each map). New-map districts are shaded by Catalyst’s Nov. 3 win probability where our forecast rates the race (ANALYST, as of ); grey districts are ones the forecast treats as not in play. Where we could not obtain an official shapefile for a new map, we leave the panel blank and say so.
4 What it takes to win the House now
The median district decides the majority. Under the 2024 lines it was Arizona’s 1st, which Trump won by 3.1 points. Under the 2026 lines it is Wisconsin’s 1st, which he won by 4.5 (The Downballot). Nationally he won by 1.5.
Under a uniform swing, Democrats would need to win the national House vote by roughly the median district’s lean relative to the nation. That is about 1.6 points on the old lines and 3.0 points on the new ones. The bar Democrats must clear has risen by about 1.4 points.
Not every source agrees. Sabato’s Crystal Ball put the 2026 median at Trump +4.9, but concludes the overall bias is still no worse than in 2018. A full seats-votes curve would need 2024 results for every new district, which we have not verified. That curve is marked as a gap.
5 What it does to voters
Districts hold nearly equal populations, so the share of districts is also roughly the share of Americans. Going into Nov. 3, 44% of districts sit in states whose map was drawn to help Republicans, and 26% in states whose map was drawn to help Democrats (The Downballot).
The redraws removed a net 13 competitive districts (seats decided by under 10 points in 2024). Each one is a community whose general-election vote mattered and now mostly doesn’t.
Minority representation took the most targeted blow. After Callais, Tennessee split Black-majority Memphis (TN-9), Louisiana dismantled its second Black-majority seat (LA-6), and Alabama redrew the court-ordered AL-2. North Carolina’s map also targets NC-1, about 40% Black.
GAP How many voters were moved between districts nationally: no published, verified count yet. We will add it once one exists.
Our forecast in the redrawn states
Catalyst’s House forecast gives Democrats a chance of the majority, with a median of seats (ANALYST). Below are the rated races in states with new maps. Several seats Republicans drew to win are still rated competitive or better for Democrats in this environment. Their gerrymanders matter most in a neutral or Republican year.
6 How to unwind it
Five strategies, each scored on feasibility (1 = remote, 5 = achievable now) by Catalyst. The ratings are ANALYST judgments, not data.
The escalation thesis: gerrymander back until Republicans want a ceasefire
The argument. A party that profits from gerrymandering won’t vote to give it up. So Democrats should draw maximal maps wherever they control the pen, until the national map stops favouring Republicans. At that point, a federal ban becomes the deal both sides prefer.
Precedent. California’s Prop 50 is already this strategy in miniature. Its ballot title reads “Temporary Changes to Congressional District Maps in Response to Texas’ Partisan Redistricting”. Before it passed, Governor Newsom wrote to President Trump (Aug. 11 2025) that California would stand down if other states did. That is a conditional retaliation paired with an offer to disarm (source).
The arithmetic. Sourced Democratic headroom adds up to seats in six states. Several more (NY, WA, IL) have no published estimate. Almost all of it is blocked by commissions or courts that Democrats themselves created. Republicans control the process in 16 states that could redraw again for 2028 (The Downballot).
The game. This is a repeated prisoner’s dilemma. If one side disarms alone, it loses the most. If both disarm, competitive maps follow. If both arm, safe seats rise and voters lose. Tit-for-tat, matching each escalation and offering to stop, is the strategy that best produces cooperation in repeated games, but only if the threat is credible and the offer is real.
The matrix ranks outcomes for each party from 1 (best) to 4 (worst). The ranks are ordinal Catalyst judgments, not seat estimates.
Risks of escalation, stated plainly
- Voter backlash. Virginia’s amendment passed by just 3 points. Maryland’s Senate refused to act. Voters in commission states chose those commissions.
- Courts. State constitutions are the real limit. Virginia’s and Colorado’s supreme courts have already stopped Democratic efforts, and Missouri’s blocked a Republican one.
- Legitimacy. Democrats campaign on democracy. Breaking reforms they championed costs credibility and gives cover to worse Republican maps.
- Asymmetry. Republicans hold more redraw-ready states (16 vs about 12), and after Callais can unpack Black-opportunity seats in a way Democrats cannot match.
- Competition. Every maximal map, by either side, destroys competitive seats. An arms race ends in a House decided by primaries.
Catalyst recommendation
Conditional escalation tied to a binding exit. Democrats should retaliate, but only in proportion, with a sunset, and always paired with a standing offer to disarm. Matching Republican seat gains state by state, with every new map expiring automatically when a national standard takes effect, makes the threat credible without making gerrymandering permanent. Prop 50’s structure is the template: temporary, voter-ratified, explicitly reactive.
At the same time, run the other three tracks: state-court litigation under state constitutions (fast, proven in Utah); ballot-initiative commissions in Republican states that allow initiatives; and a federal ban ready to pass the first time Democrats hold the trifecta, with the filibuster carved out or reformed for election law. Sequence it so each step makes the next more likely.
Sources and method
Geometry: Census cartographic boundary files (2024 lines) and the official state shapefiles listed under each map. Basemap: U.S. Census via us-atlas. All charts are drawn in your browser from data.json, which carries a label, source and as-of date for every value.