Monolith Analytics · Nebraska 2026

Nebraska 2026: the forecast

Who is likely to win Nebraska's Senate seat, its two contested House seats, the governor's and Secretary of State's offices and the Legislature, and what the early vote says so far. Every number comes from models run on the state's own election history and voter file, and the page updates as new data arrives.

Loading the latest run…

Read this first. The statewide, House, Secretary of State and county forecasts come from a model that builds each race up voter by voter and reads no polls. Tested on the two last midterms, it came within half a point of the 2022 governor's race and missed 2018's by about four, the year a former Republican ran as the Democrat. The Legislature model was tested on ten past elections, each held out in turn: it beats a naive baseline, but no more than a simple rule based on the May primary margin. Treat every number as an estimate with real uncertainty, not a prediction.

After the results are certified, we'll publish a public scorecard comparing every forecast on this page with the actual outcome, including the misses. Target date: within 30 days of certification.

One survey, and no polls

The only survey behind any forecast on this page is the Cooperative Election Study (CES), a large academic study of American voters whose answers are checked against the voter file. It tells us how party registration and age shape a vote; it holds nothing about this year's races. No poll of any 2026 race, and no national generic ballot, enters any forecast here. Each polled race's card also shows what the same model says with this year's polls folded in, and the polls themselves are further down, all for comparison.

That is the point of the page. It is a demonstration of how voting works in the United States: elections are largely predictable from information that has nothing to do with polling, such as who is registered, who has voted before, how each precinct voted last time, and what usually happens in a midterm year. That includes races whose ballots show no party. The Legislature and the OPPD board are forecast from their May primaries and the candidates' parties, and the one seat with no primary from how such races have followed the party ticket: party predicts them too.

So read a big miss as news. These forecasts say what a year looks like when voters behave the way they usually do. If the results beat them by a wide margin, the size of that gap is the measure of how historic the outcome is.

The races

Each card gives the side's chance of winning and its likely share of the two-candidate vote, with a 90% range. None of these races is forecast from polls, so none is shown surer than 90% either way.

The Legislature

Republican-registered senators after November, across simulations. Thirty-three is the number that can end a filibuster.
Every seat on the ballot

Douglas and Lancaster county offices

The contested county races in Omaha and Lincoln, each built from the same office's 2022 vote precinct by precinct and then from who turns out this year, the same way as the races above. There are no public polls of these races, so the ranges are wide and we show no chance above 90% or below 10%.

Not forecast: We don't publish a forecast for the Lancaster County Assessor/Register of Deeds race because Monolith Analytics works for a campaign in it.

Omaha Public Power District board

The three OPPD seats on the November ballot, all open. The ballot shows no party, so a candidate's party here comes only from public record. Subdivisions 2 and 3 are anchored on their May primary, moved as the Legislature's races have moved from May to November since 2002. Subdivision 1 had only two candidates and so no May vote: it is anchored on the governor's race in its own precincts, as our earlier statewide model forecast it without its polls, carried at the rate the Democrat in a nonpartisan race has followed the top of the ticket (). Each campaign's money since May moves every seat. There are no public polls of these races, so we show no chance above 90% or below 10%.

Lancaster County's early vote

Democrats' lead over Republicans among returned ballots at this point before election day (dark) and once all early ballots were in (light).

The polls

Public polls of the Senate race and the national generic ballot, beside the forecasts for comparison. No forecast above uses them: each race is built from the state's own voters and its own past results, and the difference between the two shows how much a model can know without a poll, and where polling's limits are.

How far to trust a poll: in the last three weeks of 2018, 2022 and 2024, about 1,200 Senate, governor and House polls landed within about three points of the result on average.

How it works

The statewide, House, Secretary of State and county forecasts estimate, for every registered voter, how likely they are to vote this year and how likely they are to vote for each side. Turnout comes from each voter's own history, party and age, fitted on the midterms since 2006 as the voter file stood at the time. How each kind of voter votes comes from the race's last result, precinct by precinct, with the pull of party and age measured in the CES. The two are multiplied and summed. One more term follows a pattern seen in almost every midterm: the party out of the White House gains, by 3.3 points on average in the national House vote of 1998 to 2014, and by as little as minus 2 or as much as 9; each simulation draws a year from that range. Added to each voter's odds rather than to the total, it moves these forecasts by one to two points, not three. No poll and no national generic ballot enters. The full methodology sets out every model, with and without the polls, and how each has done in tests. The Legislature forecast starts from each May primary, using how primary leaders have fared in November in 167 races since 2002. It is a model with stated assumptions, not a prophecy: a 1-in-7 chance happens one time in seven.

Data: Nebraska Secretary of State, county election offices, the state voter file, FEC and NADC filings, the Cooperative Election Study (Harvard Dataverse, doi:10.7910/DVN/II2DB6), and public polls as published, shown only. Questions: contact Monolith Analytics.

Monolith Analytics is run by Eric Aspengren, a Nebraska political data analyst and consultant. He is a Democrat and the former executive director of the Nebraska Democratic Party. His work covers political data engineering, geospatial analysis (including precinct boundary reconstruction and contract mapping for national advocacy organizations), and campaign finance research, including a database that resolves donor records across state (NADC) and federal (FEC) filings. The forecasts on this page are built independently from Nebraska election results, the state voter file and public filings. The methods are published so others can check them.

Model last updated —. Earlier versions are archived at monolithanalytics.org/forecast/archive.

Found an error or have a correction? Contact Monolith Analytics. We'll fix mistakes and note the change.