October 19, 2026 · By FutureBallot
In New Hampshire, Wisconsin, Michigan and Arizona, Republican majorities look safer on paper than they do in FutureBallot's forecast. Close districts with a known holder are concentrated on the Republican side.
Republicans hold 212 of the 400 seats in the New Hampshire House, 43 more than Democrats. FutureBallot's state-legislative forecast gives them a 52% chance of continuing to hold their majority after November, and gives Democrats a 47% chance of taking the chamber (New Hampshire House). A lead of more than 40 seats is really, in this forecast, closer to a coin flip than a lock.
And New Hampshire is not alone. Of the 88 chambers in 46 states with seats on the ballot this year, 11 are competitive: in each, no outcome has a probability of 85% or more (State legislatures). The line is a choice, not a natural break; the chambers just outside it are listed below. In several of the 11, the current split looks more settled than the forecast does.
Four chambers where the lead looks bigger than the race
Wisconsin Assembly. Republicans hold 54 of 99 seats, Democrats 45 (NCSL). Republicans can lose only four seats and still keep their majority. The forecast still gives Democrats a 30% chance of taking the chamber (Wisconsin Assembly). Ten districts are close (neither party above 85%), and Republicans hold nine of them.
Michigan House. Republicans hold 58 of 110 seats, with two to spare. Their chance of keeping a majority is 53%; Democrats' chance of winning one is 38%, and a 55-55 tie takes the rest (Michigan House). Of 12 close districts, Republicans hold 10.
Arizona Senate. Republicans hold 17 of 30 seats, one more than a majority. The forecast gives them a 44% chance of keeping their majority after November. It gives the Democrats a 38% chance, with an 18% chance of a 15-15 tie (Arizona Senate). All four close districts are held by Republicans.
New Hampshire House. The chamber has 400 seats, and 149 of them sit in the 67 districts the forecast rates as close. Many of those districts elect several members at once. Republicans hold 18 of them and Democrats eight; for the other 41, which include multi-member districts, the forecast records no single holder. NCSL lists 18 House seats as vacant.
The pattern is similar in each. The current seat count shows how many seats each party holds today; the forecast asks how many seats are genuinely in doubt this time, and which party holds them. In Wisconsin, Michigan and Arizona, almost every close district belongs to the majority, so the seats in doubt are the majority's to lose. In New Hampshire most close districts have no single holder, but of those that do, Republicans hold more than twice as many as Democrats. What that concentration does to a majority's chances is already in the forecast's numbers above; this article does not add a test of its own.
Where the majority has less than a 50% chance of keeping it
In two competitive chambers the party holding a majority today has less than an even chance of keeping it.
Wisconsin Senate. Republicans hold 18 of 33 seats. Only 17 seats are on the ballot; the other 16 are held over, and Democrats hold 10 of those, Republicans six. So Democrats need seven of the 17 seats being contested to reach the 17-seat majority, while Republicans need 11. The forecast gives Democrats a 60% chance (Wisconsin Senate). Thirteen of the districts it forecasts were redrawn since they were last on the ballot, so their starting point is rebuilt from 2024 presidential votes rather than taken from their last legislative result (review overview).
Arizona Senate, above, is the second: Republicans have a majority today, and they are still more likely than Democrats to hold one after November (44% against 38%), but an 18% chance of a 15-15 tie keeps their chance below even.
The chambers that are close, and look close
Five of the other six competitive chambers are narrowly divided now, and the forecast has them close too. The sixth, the New Hampshire Senate, is not narrowly divided (see below the table). The three columns are the forecast's chances of a Democratic majority, a Republican majority, and neither (a tie, or independents holding the balance); they are rounded separately, so a row can add to 99 or 101.
| Chamber | Seats now (NCSL) | Democratic majority | Republican majority | Neither |
|---|---|---|---|---|
| Pennsylvania House | 102 D, 100 R, 1 vacant | 73% | 27% | 0% |
| Minnesota House | 67 D, 66 R, 1 vacant | 71% | 23% | 7% |
| Minnesota Senate | 34 D, 33 R | 79% | 21% | 0% |
| Maine House | 75 D, 73 R, 3 other | 70% | 30% | 0% |
| Michigan Senate | 20 D, 18 R | 61% | 25% | 14% |
| New Hampshire Senate | 8 D, 16 R | 11% | 81% | 7% |
The New Hampshire Senate is competitive under this article's definition only because Republicans' chance of a majority, 81%, falls below the 85% line. Democrats' chance is 11%; the rest is a 12-12 tie.
Just outside the line are the North Carolina House, where Republicans hold 71 of 120 seats and the forecast gives them 86% (North Carolina House), and the Alaska and Pennsylvania Senates, at 89% each for Republicans. In the Alaska House, the forecast's likeliest outcome is that neither party wins a majority (96%), as in recent sessions run by bipartisan coalitions (Alaska House).
Chart: competitive chambers where one party holds a majority of seats today. Maine's and Minnesota's Houses, where no party holds a majority today, are left off.
How we did this
- Forecast. Every chance is FutureBallot's published 2026 state-legislative forecast (model
stateleg_swing/0.11, made October 19, 2026 at 04:05 UTC), as shown on /state-legislatures. Each district starts from its own last result, moved by the national House environment and by whether an incumbent is running; chamber chances come from 20,000 simulated elections that move the districts together (methodology). The forecast file this article was computed from is kept beside the analysis (forecast_20261019T040531Z.json.gz, SHA-2568735adb6…), so the counts can be rebuilt from exactly that input. - Competitive chamber: the likeliest outcome (a Democratic majority, a Republican majority, or neither) has a chance shown below 85%.
- Close district: a district on the ballot whose Democratic chance is shown between 15% and 85%.
- Current seats are NCSL's State Partisan Composition table, updated October 6, 2026, which counts party members and does not take coalitions into account; the per-chamber counts it gave are recorded in
figures.json. Who holds each close district is the forecast's own incumbent record (Open States). - Every chance quoted was checked against the live overview page on October 20, 2026; all 88 chambers matched. That check compares the chamber chances and seats on the ballot as displayed; it does not independently verify the district holders or the forecast itself. The script and its output are
compute.pyandfigures.jsonin the repository'sanalysis/state-chambers-closer-than-seat-counts/folder.
What this doesn't tell you
- This is an initial model, and its ranges have run wide. In tests on past elections it missed one of its five accuracy targets: in 2024 its 80% ranges held the result 92% of the time, so they were wider than they needed to be. If the model overstates uncertainty again, some chambers counted as competitive here are safer for their favoured party than these chances suggest. Those tests reconstruct how the model would have done on elections it was built with; 2026 is its first test on an election it has not seen (State legislatures).
- A majority of seats is not always control. Coalitions, power-sharing agreements and tie-breaking votes decide who runs a chamber in some states, and this article does not model them.
- "Close" is a count of districts, not a cause. The comparison between seat counts and chances is descriptive. It is not part of the forecast model, and no test here shows that the share of close seats a majority holds predicts a flip.
- Candidate lists change. A withdrawal or a court ruling after October 19 can move a district, and so a chamber.
Figures used (all from figures.json): competitive[*] (chances, seats now, cushion, close districts and their holders, seats needed), n_chambers_with_election, n_states_with_election, n_competitive, near_misses_85_to_90, neither_likeliest, sources, site_checks.
FutureBallot publishes one analysis a week until the election. Get the briefing by email, or follow the Analysis page.