Your 2026 election-watching guide

What polls and experts expect to happen. Real Clear Politics’ generic congressional ballot polling average has Democrats up 8.9% (52.0%-43.1%), and that margin has been in the 8-9% range for about a month. Such a lead is consistent with a wave election big enough that gerrymandering the districts shouldn’t matter. So just about everybody expects Democrats to win the House.

Things get interesting in the Senate, where 34 Democrats and 31 Republicans are not up for election this year. Cooks Political Report lists 9 Democrats and 11 Republicans as solid favorites, bringing the totals to 43 Democrats and 42 Republicans. Each party has two seats listed as “likely” and two more as “leaning” in their favor. So if you add those in, it’s 47-46 for the Democrats. That leaves 7 toss-up races, which is an unusually large number. So any outcome between a 54-46 Democratic majority and a 53-47 Republican majority (the same distribution we have now) is plausible. (Democrats have had fantasies about flipping seats in Mississippi and Florida, but that seems like a bridge too far.)

How that happened is fascinating. For months, Cooks has been shifting races towards the Democrats, but not across the board: Races might go from “likely Republican” and “leaning Republican” to “toss-up”, but the toss-up races would not shift to “likely” or “leaning” Democratic. (The NYT/Siena poll does seem to show Dems pulling away, but it could be an outlier.)

That appears to have happened for a few reasons: The worst impacts of Trump’s policies — high diesel prices, closing hospitals, data-center construction — have tended to fall in rural states that usually vote Republican: Iowa, Kansas, Nebraska, South Carolina, Alaska. Because these states weren’t supposed to be competitive, the GOP is stuck with some poorly chosen candidates. I mean, Roger Marshall is a sucky senator, but it’s Kansas, so who cares? Darlene Graham isn’t her late brother Lindsey, but it’s South Carolina, so anybody with an R next to her name was supposed to be safe.

Meanwhile in blue states like Michigan and Maine, Democrats have maintained small but fairly stable leads in the polls. In a shifting national environment, you’d expect them to start pulling away, but they haven’t. So the RCP average has Troy Jackson up 3.3% on Susan Collins in Maine, and Abdul El-Sayed up 5.3% in Michigan. Both margins were considerably smaller before averaging in yesterday’s NYT/Siena poll.

You especially can’t count on small margins in a mid-term election, because mid-terms are all about turnout. The difference between Obama’s landslide in 2008 and the Democrats’ “shellacking” in 2010 wasn’t so much that people changed their minds. It was that the 2008 Obama coalition of young voters and minorities didn’t come out to vote in 2010. 2010 was an older, whiter electorate, and it elected a way more conservative Congress.

So a big piece of the Blue Wave theory is that Democrats are more motivated this year than Republicans. Trump voters who believed he would lower prices, end foreign wars, and generally look out for the little guy (at least the White male little guys) are disillusioned and unmotivated. But Democrats who have watched Trump circumvent laws, enrich himself, and send his ICE thugs out to murder American citizens with no consequences — they’re motivated.

And that explains the Republicans’ closing arguments that center on fear and bigotry: Democrats will turn America Communist, let brown people run the country, change your child’s gender, take away your guns, and persecute Christians. El-Sayed is a jihadist, James Talerico and Adam Hamilton are fake Christians, and so on. Those arguments won’t convince anybody who cares about the price of gas or health care, but they’re not supposed to. They’re supposed to fire up the MAGA cultists to go out and vote.

The upshot of all this, in my opinion, is that the toss-up races will almost all go the same way. One party or the other will sweep at least six of the seven. The people who make up their minds at the last minute, the people who may or may not vote — they’re all going to do the same thing. I just can’t tell you what it is.

Last-minute events. The Michigan Senate race got roiled Sunday when Rep. Haley Stevens (who lost to Abdul El-Sayed in the primary) withdrew her endorsement of El-Sayed and made an appearance with Republican nominee Mike Rogers, whom she had previously described as “a rubber stamp for Donald Trump“. In a subsequent interview with CNN’s Anderson Cooper, she dodged and flailed when asked what changed her mind. Even as her social media message absurdly accused El-Sayed of campaigning on “hate”, she appealed to Islamophobia by referring to him only as “Abdul”.

Michigan observers question how much difference this will make. It has been a much bigger story in the national media than locally, and has focused attention more on Stevens herself than on any problem with El-Sayed’s candidacy. Her career as a Democrat is over now, and it’s hard to imagine her as a Republican given her record in Congress. Possibly she’s hoping for a job as a lobbyist with AIPAC, who channeled $30 million into her primary campaign and has also spent millions to support Rogers.

Similarly, the would-be scandal of Democrats recruiting a second Dan Sullivan to run for the Senate in Alaska seems to be more of a national-media story than an issue on the ground in Alaska. The media-criticism blog Finding Gravity comments:

[T]he national news media is doing everything they can to make the central story of the last days of this election season an alleged political maneuver by Alaska Democrats that was, by all accounts, legal, consistent with past behavior by both major political parties, and unlikely to have any real effect on the outcome of the Alaska Senate race. America’s dominant news companies, led by the New York Times, have literally ignored the new rape allegation against the president of the United States while breathlessly hyping the Alaska maneuver, giving Donald Trump the distraction he desperately seeks and a phony premise for his anti-democratic efforts to effectively cancel a Senate election that could determine control of the entire U.S. Senate.

… At this writing, the New York Times has now run more than a dozen reports on this matter in just three days. The Times is in the kind of full-court press mode it saves for things like Joe Biden stumbling over his words. By my tally, the alleged Peltola maneuver has been a primary subject of six New York Times articles, four New York Times newsletters, a New York Times podcast, a New York Times video, and two letters to the editor the New York Times chose to run. At least nine New York Times journalists have bylines on these pieces. I can only imagine what the Sunday crossword puzzle has in store for us.

I have two observations:

  • If Republicans had done something like this, it would be covered as an infuriating-but-clever move to own the libs.
  • It’s a non-story, because Alaska has instant run-off voting. Even if you get confused about which Dan Sullivan is the Republican incumbent, despite the words “Republican” and “incumbent” appearing next to one Dan Sullivan’s name, you just rank the two next to each other. When the “fake” Dan Sullivan is eliminated, your vote will flow to the “real” one. So there is absolutely no chance this maneuver will affect the outcome of the election.

How the night unfolds. There’s a certain amount of tea-leaf reading you can do before any polls close. Exit polling goes on all day. The major media networks are committed not to release any of that information while the polls are open, but their talking heads have probably seen it. They’re generally pretty careful not to reveal anything directly, but I believe they shade their predictions so that they don’t look stupid later. So if MS-NOW folks are still talking about the Blue Wave at 5 p.m., it’s probably still plausible. Likewise, if Fox News pundits are making not-so-fast noises, that could be what the exit polls are showing.

The returns start coming in at 7 p.m. EST, when the polls close in Georgia, Indiana, Kentucky, South Carolina, Virginia, Vermont. (All poll closing times and a summary of the interesting races are in this NPR article. Some eastern time zone districts in Kentucky and Indiana will report at 6.) None of the toss-up Senate races are here, but you can get some insight into how the night will go by looking at margins. The South Carolina Senate race is supposed to go to Republican Darlene Graham by a small margin. How small? If this race stays too close to call for a long time, it’s a good night for the Democrats. If Graham wins with surprising ease, that’s good for Republicans. The WaPo’s Henry Olsen targets five points as the line to watch.

The Georgia governor’s race is also expected to be close.

Florida’s eastern time zone polls close. A number of House races are close, including some recently gerrymandered districts.

At 7:30: North Carolina, Ohio, West Virginia. The big race here is the Ohio Senate race, where Democrat Sherrod Brown has a small lead in the polls, but the race is considered a toss-up. Don’t expect that to get called early either way. The thing I’ll be watching for early is Black turnout in Cleveland and other big cities. The first bad sign for Hillary Clinton in 2016 was low Black turnout in Cleveland. Probably the Democratic early vote comes in first, so Brown should have a big early lead that shrinks as the night goes on.

Roy Cooper is expected to flip the North Carolina Senate seat to the Democrats. Polls have him up double digits. If that holds, and the race gets called early, that’s a sign that the polls are accurate. NC-1 is a toss-up House seat. Gerrymandering was supposed to push Democrat Don Davis out of Congress. If it doesn’t, that’s a bad sign for Republican gerrymanders across the country.

OH-7 is another toss-up race. This time Republican Max Miller is the incumbent.

At 8: Alabama, Connecticut, Delaware, Florida, Illinois, Maine, Maryland, Massachusetts, Mississippi, Missouri, New Hampshire, New Jersey, Oklahoma, Pennsylvania, Rhode Island, Tennessee, Washington, D.C.

The Maine Senate race is one of the closest in the country. Chris Pappas is supposed to be ahead in New Hampshire; if he’s not, that’s a bad sign for Democrats. If the Mississippi Senate race is close, then the toss-up races are trending Democratic.

In Pennsylvania the focus is on House seats, where Democrats are expected to flip several. Florida is another test of Republican gerrymandering.

Josh Shapiro and Jay Pritzker will easily get reelected as governors in Pennsylvania and Illinois, but the margins matter if you think of them as a presidential contenders in 2028.

Somewhere in the 8-9 range, expect to hear projections that Democrats will take the House. The majority of races won’t be called, but enough toss-up races should come in to determine the outcome.

8:30. Arkansas. The only real news here will be if the Republicans don’t sweep.

9: Arizona, Colorado, Iowa, Kansas, Louisiana, Michigan, Minnesota, Nebraska, New Mexico, New York, North Dakota, South Dakota, Texas, Wisconsin and Wyoming

Lots to chew on here. Michigan is a toss-up Senate race with last-minute events. Cook has the Iowa Senate race as a toss-up and the polls are very close, but I have a hard time believing Iowans won’t come home to the Republicans.

Texas and Kansas test the liberal Christian message. Will Christian voters stick with MAGA, or do the teachings of Jesus still mean something? Will Christian voters prioritize anti-abortion and anti-gay rights (two issues Jesus never mentioned) or helping the poor, healing the sick, and welcoming the stranger (which he talked about all the time)?

NPR says this about Texas:

Talarico could take an early lead as early voting data and mail in ballots are counted but Paxton will probably start to gain ground as counties will start reporting their totals. While the governor’s race between Republican Gov. Greg Abbott and state Rep. Gina Hinojosa is also close, Abbott is still favored.

New York has some Republican House seats that Democrats need to flip.

By now, you shouldn’t be looking for tea leaves to read any more. The night will have developed momentum one way or the other.

10: Montana, Nevada, Utah. Not a lot of excitement here. Mostly watch for close races from earlier states to begin resolving.

11: California, Idaho, Oregon, Washington. No major Senate races. California is a test of Democratic gerrymandering. But California results take a long time to come in, so we may not know much until Wednesday, if then. Ditto for California ballot questions, like the wealth tax.

12: Hawaii.

1: Alaska. If the Senate is close nationally, it could all come down to Alaska. Put some coffee on.

Post a comment or leave a trackback: Trackback URL.

Comments

  • Joel N.'s avatar Joel N.  On November 2, 2026 at 10:50 am

    you have mis-spelled Talarico.

  • fmanin's avatar fmanin  On November 2, 2026 at 11:19 am

    I’m not convinced that all these tossups are actually tossups. I kind of suspect that pollsters are hedging, since polling is so difficult these days. The New York Times poll (which is transparent and consistent about its methodology) actually shows a big spread among these races: AES is winning decisively, Dems are clearly ahead in Maine, Ohio, Texas and Kansas, Iowa and Alaska are true tossups, South Carolina is actually close, Nebraska and Florida are not (where many others show Nebraska as tied). By the way, Nate Silver has not had Michigan in the tossup column for a long time, and has written about how “traditional” pundits often underestimate the signs of a wave. Wikipedia has a nice summary chart here:
    https://en.wikipedia.org/wiki/2026_United_States_Senate_elections#Predictions

    • ldbenj's avatar ldbenj  On November 2, 2026 at 12:20 pm

      The results for close races tend to resolve in the same direction. If Keisha Lance Bottoms wins in Georgia, James Talarico will probably win Texas, and vice versa.

  • Unknown's avatar Anonymous  On November 2, 2026 at 12:17 pm

    Recruiting candidates with the same name goes back a ways. One fairly well-known example was in 1946 in Massachusetts, when a Democratic primary candidate named Joe Russo was running against another promising Democrat. That Democrat’s campaign manager found another man named Joe Russo to run in the primary. It didn’t matter because the fake Joe Russo only got 1000 votes, while the promising Democrat won with 26,000 votes.

    The “promising Democrat” was John F. Kennedy and his campaign manager was his father, Joseph Kennedy. Joe learned this trick from his father-in-law John Francis Fitzgerald, better known as “Honey Fitz,” Congressman and Boston mayor. There have been many other examples of this in recent years of candidates with the same or similar names, the most confusing being Jose E. Serrano running against José E. Serrano.

Trackbacks

Leave a Reply