For months, polling and prediction markets have forecast that Democrats will retake the House, but that Republicans will retain control of the Senate. However, recently, the prediction markets have moved to control of the Senate being a toss-up. Here is how they got there.

The current “consensus” map on the 270-to-Win website shows 50 Republican likely wins, 46 Democratic, and 4 toss-ups. The four toss-ups on their map are Maine, North Carolina, Michigan, and Ohio.  

I am not sure whose consensus they are divining, but to call Maine, North Carolina, and Michigan toss-ups at this point seems pretty out of touch. Here are the prediction markets and the RealClearPolling average showing the Democratic advantage for those states.

Polling in Michigan is still in flux because of a hotly contested primary that will likely result in a runoff. All we have right now are a couple of hypothetical matchups, which do show a toss-up. However, according to Morning Consult, Trump is underwater by 9 points in Michigan. The prediction markets strongly forecast that the polling will swing toward the eventual Democratic nominee.

Probably the toughest race of the four toss-up states is Ohio. That is a special election to fill J.D. Vance’s unexpired term. Vance won the seat in 2022 by 6 points. The RCP polling shows the Republican candidate, Husted, with a small lead, within the margin of error. The prediction markets are giving the likely Democratic candidate, Brown, a 7-point advantage.

Of course, even if the Democrats were to win all of four of these seats, that would leave a 50-50 Senate, and Vance, as vice-president, would cast the tiebreaker for the Republicans to retain control. So, for the Democrats to win control, they would have to win all four of these and pick up one other seat where the Republican is currently favored. Here are the possibilities:

Alaska

In Alaska, Sen. Dan Sullivan is running for his third term. The traditional wisdom has been that since Trump won by double digits and Sullivan has the incumbent advantage, this is a safe seat for Republicans. However, the Democratic challenger is Mary Peltola, who was elected to the House in 2022 but lost the seat in a close election to a Republican in 2024. Since Alaska only has one House seat, those elections were statewide.

The only RCP polling from the pollster, Data for Progress, leans Democratic. Its two polls this year show the race as a toss-up. However, in March, the prediction markets began to break hard in Peltola's favor. Polymarket currently shows her as a 68% favorite, and Kalshi is not far behind at 60%. Keep in mind that Alaska has ranked-choice voting, which tends to favor more moderate candidates.

Iowa

Because of Joni Ernst’s retirement, this is an open seat this year. Like Alaska, most politicos consider it to be a safe Republican seat. However, some recent polling has shown the Democrats have a chance, and prediction markets have seen the odds narrow from over 80/20 to under 60/40. Both parties have primaries in early June. The leader in the Democratic primary is an interesting candidate who is a Paralympic gold medalist who is running on a populist platform.

Nebraska and Montana

Both of these states should be safe Republican seats, but in each, the Republican candidate will face a strong independent bid. In Nebraska, Dan Osborn will make his second bid as an independent populist after losing to Deb Fisher in 2024 by only 7 points. There will probably not be a Democratic candidate in the race.

Montana could be even more interesting. There is not much data to go on in this race because of the last-minute withdrawal of the incumbent senator, Steve Daines, in an apparent attempt to engineer a succession by a close political ally. Of course, this is always a politically unpopular maneuver and may be especially so in independent-minded Montana.

It may also be a boon to the nascent campaign of independent Seth Bondar. He is a West Point graduate, former Green Beret, Rhodes Scholar, and president of the University of Montana. He should not be taken lightly by the Republicans. This is a link to a long-form interview he did with Michael Smerconish.

Texas

Of course, we cannot leave the subject of Senate control without mentioning the Democrats' Holy Grail quest to win a Texas U.S. Senate seat. Of course, that quest has been repeatedly dashed for the last four decades. But could it be different this time?

The polling certainly seems to indicate so. The RCP polling average shows Talarico slightly ahead of both Cornyn and Paxton. However, a University of Texas poll released this week showed Talarico leading Paxton by 8 points and Cornyn by 7 points. One word of caution about this poll: it surveyed registered voters, not likely voters. Because turnout patterns differ between the parties, registered-voter samples can sometimes lean more Democratic than likely-voter models. This could be especially true in this election because of a rift between Talarico and black voters.

The prediction markets are certainly hedging the earlier bets. The likelihood of Talarico winning has gone from 20% to 45%.

I think you can see from all of this why the prediction markets are calling the control of the Senate a toss-up. Neither party has a clear path. Democrats must run the table on the seats currently rated as toss-ups and win, at least, one other seat where Republicans are currently favored.

On the other hand, Trump’s approval rating has become an anchor around the necks of down-ballot Republicans. He has suffered a catastrophic loss of support among the two most important swing-voter groups – independents and Latinos. And his approval on the inflation – that great presidential support killer – is now a negative 39%. Unless Trump can turn around his numbers or down-ballot Republicans can somehow distinguish themselves from him, the 50% on the prediction marketing betting on Democratic control might just be right.