The Final 2022 Midterms Polling Forecast

Republicans are defending 21 seats, six of which were left open by retirements. Just four of those races are rated competitive (either toss-ups or races leaning to a particular candidate) by the experts at the Cook Political Report. Democrats are defending just 14 seats, only one of which is “open.” Cook considers five of these races competitive. FiveThirtyEight’s projections based on polling and expert opinion now give Republicans 55 percent odds of flipping the Senate (this and their individual race projections are probabilities, not predictions); Democrats have a 45 percent probability of holding the Senate. Let’s take a closer look at the most uncertain races.

.

Pennsylvania: Fetterman and Oz are in a dead heat.

The contest between Democratic lieutenant governor John Fetterman and celebrity doctor Mehmet Oz began with Fetterman clearly favored and Oz derided as an amateur. It has evolved into one of the closest races in the country with Oz leading by 0.1 percent in the RealClearPolitics polling averages. But that figure could be slightly misleading, since the last two polls (each giving Oz a two-point lead) were from polling outfits (Trafalgar Group and Insider Advantage) that have been regularly showing above-average performance by Republican candidates across the country. FiveThirtyEight’s forecast based on polls and other data gives Oz a 56 percent probability of winning.

The seat these two candidates are battling over is currently occupied by retiring Republican Pat Toomey, so a Fetterman win could give Democrats a bit of a firewall against Senate losses elsewhere.

.

Georgia: Walker vs. Warnock could go to overtime.

In another Senate race where a Republican often mocked as a low-quality candidate has caught up with a Democratic front-runner, Herschel Walker now leads Democratic incumbent Raphael Warnock by 0.5 percent in the RealClearPolitics polling averages. Crucially, however, neither candidate is polling at or very near 50 percent in a state that famously requires general-election runoffs if no one wins a majority of the vote. Libertarian candidate Chase Oliver was at 4.5 percent in the latest poll of the race from Landmark Communications, which showed Walker leading Warnock by 1.4 percent with 46.8 percent of the vote. FiveThirtyEight’s projections give the Republican 63 percent odds of finishing ahead (though perhaps not with a majority).

Aside from the possible overtime contest, this race has been characterized by a drumbeat of revelations about Walker’s background that have kept him from fully benefiting from what is looking to be a Republican wave in the state. Warnock has run a steady and magnificently funded race, but it could clearly go either way now or in December.

.

Arizona: Kelly vs. Masters is a MAGA-movement bellwether.

A third Senate race featuring a Trump primary endorsee struggling to catch a national wave is in Arizona, where incumbent Democrat Mark Kelly has yet to trail Peter Thiel protégé Blake Masters in a single public poll. But the contest has unmistakably tightened with Kelly now leading in the RealClearPolitics polling averages by a single point. A late poll from Insider Advantage shows the race actually tied, and another from Remington Research gives Kelly a three-point lead. FiveThirtyEight projects that Kelly has a 66 percent probability of winning.

Arizona’s Senate race is best understood in the broader context of a highly competitive state in which this year’s GOP ticket is dominated from top-to-bottom by MAGA election deniers who are benefiting from border-control concerns along with the general sense that Democrats are losing their grip on Latino voters (some alienated by cultural progressivism and others just upset by inflation like everyone else).

Masters is, depending on your perceptions, a refreshingly frank outsider or or a creepy extremist stooge. You don’t sense he’d have much of a chance of beating the steady and genial centrist Kelly without a favorable national environment, but he likely has that going for him.

.

New Hampshire: The key to Senate control?

The contest between incumbent Democrat Maggie Hassan and MAGA candidate and retired general Don Bolduc heated up very late. While Hassan was considered vulnerable from the beginning of the cycle, Bolduc was far down the list of preferred challengers from the GOP; after a September primary win, he was struggling to overcome a history of wacky statements and poor fundraising skills. But sensing this race might be the key to control of the Senate, Mitch McConnell and other national Republicans have poured resources and messaging help into Bolduc’s campaign and made him a viable candidate.

Like her colleague Mark Kelly, Hassan is hanging on to a one-point lead in the RealClearPolitics polling averages after looking like a sure bet for reelection over many months. Two late polls from New Hampshire outlets illustrate how close the race has become with St. Anselm College giving Bolduc a 48-47 lead and the University of New Hampshire showing Hassan still up 50-48. FiveThirtyEight is still a bit bullish on Hassan, giving her a 72 percent probability of winning.

If early reporting New Hampshire gives Republicans a Senate gain on Election Night, it will likely be a long dark night for Democrats.

.

Nevada: Cortez Masto and Laxalt’s geographical puzzle.

Nevada’s contest between incumbent Democrat Catherine Cortez Masto and Republican famous-name scion Adam Laxalt has been close all along. Now it will come down to a familiar battle between Democrats generating a huge ground-game lead in Las Vegas and Republicans running up the score in rural counties. The polls have been leaning toward Laxalt, a former state attorney general, who leads in the RCP polling averages by 2.8 percent, but some of that may be the mix of pollsters. Nevada political journalism “dean” Jon Ralston has with no great certainty predicted a Cortez Masto win. FiveThirtyEight has the race almost exactly dead even with Laxalt having a 51 percent probability of prevailing.

Variables include not just the proper functioning of the Democratic turnout machine built by Cortez Masto’s mentor, the late senator Harry Reid, but also whether the incumbent can stem the erosion of Democratic support from her fellow Latinos and whether Laxalt can produce historic margins among non-college-educated voters in a sour economic environment.

.

Wisconsin: Can Johnson hold on yet again?

Perpetually underestimated Republican senator Ron Johnson has never been especially popular in Wisconsin but has a real knack for winning elections. In his bid for a third term, he leads Democratic state lieutenant governor Mandela Barnes by 2.8 percent in the RCP polling averages, mostly on the wings of a heavily negative set of attacks on Barnes for past comments the progressive Democrat made about policing and criminal-justice reform in a state where racial tensions are high. FiveThirtyEight gives Johnson a 78 percent probability of winning.

Barnes hasn’t led any public polls of this race since September, nor has RonJon built any big leads. The final survey from the generally reliable Marquette Law School outlet showed Johnson up 50-48, well within the margin of error. A win here against the only Republican incumbent senator who’s in serious trouble would be huge for Democrats.

.

North Carolina: Just out of reach for Democrats?

The battle for another Republican-held open seat in a highly competitive state that went narrowly for Donald Trump in 2020 may be remembered as a lost opportunity for national Democrats. Absent a major national intervention, Democrat Cheri Beasley can’t quite seem to catch up with Republican Ted Budd, who now leads by 5.6 percent in the RCP polling averages; the contest was very close in mid-September.

The preponderance of Republican-leaning pollsters in this state, however, leaves some hope for Democrats, although a non-Republican-leaning Marist poll in mid-October showed Budd leading Beasley 49-45. This may have just been the wrong year for Beasley to run. FiveThirtyEight puts Budd’s win probability at 82 percent.

.

More potential Senate upsets (that probably aren’t happening).

Several once competitive or potentially competitive Senate races seem to have fallen off the battleground map in late polling. In Florida, Republican incumbent Marco Rubio leads Democrat Val Demings by 7.8 percent in the RCP polling averages; a late Siena poll had Rubio up 51-43. In Ohio, Democrat Tim Ryan’s highly competent campaign kept him competitive in a strongly red-leaning state for a good while, but Republican Trump endorsee and Thiel-financed J.D. Vance now leads by 7.5 percent in the RCP averages; a late Trafalgar Group poll showed him up by 10. In Colorado, Republicans were bullish about outsider business executive Joe O’Dea, but Democratic incumbent Michael Bennet has led in every public poll (though polling has been sparse) and is up by 5.3 percent in the RCP averages.

An early-October Iowa poll from Ann Selzer raised eyebrows by showing seven-term Republican incumbent Chuck Grassley in trouble against Democrat Mike Franken, but the same pollster now shows the 89-year-old Republican comfortably ahead by 12 points, which is what you’d expect in red Iowa. Utah Democrats backed conservative independent Evan McMullin against Republican incumbent Mike Lee and created some initial excitement, but recent polls show Lee blowing out to a double-digit lead. And in Washington, Republicans have been excited by a few GOP-leaning polls showing their candidate Tiffany Smiley running not far behind five-term Democratic incumbent Patty Murray, but more objective polls and the strong majority Murray posted in the state’s nonpartisan primary suggest an upset there would be a real surprise.

Of the 20 GOP-held governorships up this year, two are almost certain to flip (Massachusetts and Maryland, where highly atypical moderates Charlie Baker and Larry Hogan are leaving office). Only two others are in competitive territory, according to the Cook Political Report. Meanwhile, seven of the 16 Democratic-held governorships up this year are at risk in competitive races. So if there is indeed a national Republican wave in these midterms, the GOP could reach or exceed historic highs in the number of governorships it holds (most recently 33 after the 2016 and 2017 elections). Here are the most competitive gubernatorial races.

.

Oregon: Normalcy may reassert itself.

Throughout the cycle, Republicans have had reason for optimism about winning the governorship in very blue Oregon for the first time in 30 years. Until very recently, relatively moderate GOP nominee Christine Drazan has led progressive Democrat Tina Kotek and more centrist ex-Democratic independent candidate Betsy Johnson by a narrow but consistent margin. But a late poll from Emerson suggests Johnson’s support could be collapsing (as often happens to independent candidates late in campaigns that they can’t win). This seems to be benefiting Kotek, who’s suddenly up by five points.

Still, the race is tied in the RealClearPolitics polling averages, and FiveThirtyEight has Kotek’s win probability at a mere 54 percent. A last-minute breeze in either direction could be crucial.

.

Wisconsin: No breathing room between Evers and Michels.

If Oregon is at the very last minute trending Democratic, Wisconsin probably has the nation’s closest gubernatorial race, involving incumbent Democrat Tony Evers and Trump-endorsed self-funding businessman Tim Michels. The Republican has a 0.4 percent lead in the RCP averages. The respected Marquette Law School’s final poll has them tied. Siena showed Evers up 2 percent, and Trafalgar Group had Michels up 2 percent. FiveThirtyEight has Evers’ odds of winning at 51 percent, which basically means the candidates are dead even.

The two candidates in this highly polarized state have almost nothing in common, though Michels did recently commit to accept the results if he loses, despite publicly doubting the legitimacy of Joe Biden’s 2020 win in the state and nationally.

.

Arizona: Lake-effect storm in the desert.

Despite getting a lot of props for outcampaigning her Democratic opponent, Katie Hobbs, Republican gubernatorial nominee Kari Lake, perhaps the purest MAGA candidate for statewide office anywhere, hasn’t been able to put away a win over Katie Hobbs in most polling. Lake does lead in the RCP polling averages by 1.8 percent, but Hobbs led in a final Marist survey 48-47.

In a state loaded with close races where the GOP ticket is pretty remote from term-limited incumbent Republican governor Doug Ducey, Lake remains the betting favorite, but it’s hard to calculate how much of the longtime local-TV anchor’s winner’s glow is hype and media savvy. FiveThirtyEight does give Lake a 63 percent chance to win.

.

Nevada: A Vegas edge for Lombardo.

Like the Nevada Senate race, the gubernatorial contest between Democratic incumbent Steve Sisolak and Republican challenger Joe Lombardo has been very close. But Lombardo is leading by 2.3 percent in the RCP polling averages; FiveThirtyEight gives him a 57 percent probability of winning.

Lombardo’s potentially crucial advantage is that he’s currently the sheriff of Democratic stronghold Clark County (Las Vegas) and may improve on the usual GOP vote there. That’s a key reason Nevada political wizard Jon Ralston predicts he should win narrowly.

.

Kansas: Can Kelly do it again?

Kansas’s Democratic governor Laura Kelly won in 2018 in no small part because of a split within the state’s dominant Republican Party. There’s no real split this year, but Kelly could still win reelection over State Attorney General Derek Schmidt. There hasn’t been much public polling in the state, but Kelly led by three points in the latest poll from Emerson.

One major sign of hope for Kelly was the smashing defeat of an anti-abortion constitutional-amendment ballot initiative in August that she opposed and Schmidt favored. FiveThirtyEight calculates Kelly’s chance of winning at 66 percent, but it’s obviously a state where a red breeze could matter a lot.

.

Michigan: Whitmer’s star could rise or fall.

Until mid-October, Democratic incumbent governor Gretchen Whitmer was cruising toward an easy win over Republican Tudor Dixon, a Trumpy far-right commentator who initially struggled to raise money. But the polls have shown a tighter race recently (Whitmer is up by just 4.2 percent in the RCP averages), and Dixon managed to raise enough money to run a batch of ads.

This is one contest where polls really do diverge: Republican-leaning Trafalgar Group and Insider Advantage recently showed the race tied. But Michigan based EPIC-MRA gave Whitmer an 11-point lead in its final survey. FiveThirtyEight clearly finds the latter more credible, giving national star Whitmer an 89 percent probability of victory.

.

Georgia: Abrams needs a turnout miracle.

Normally, you’d figure Republican incumbent Brian Kemp had safely put away a second challenge from Democrat Stacey Abrams; he has an eight-point lead in the RCP polling averages. I include this race in the competitive category for two reasons: Abrams’s expertise in voter mobilization and turnout mechanics, and Georgia’s majority-vote requirement for general-election victories.

If Kemp ultimately underperforms a bit, he could be knocked into a December 6 runoff that would give Abrams a second chance at a comeback victory.

.

More close governor’s races (where an upset is less likely).

There are an array of gubernatorial races in which initial optimism about an upset has given way to reality — unless things get really weird. Hyperoptimists for a big Republican wave have pointed to isolated polling showing New York’s Lee Zeldin (challenging Democratic incumbent Kathy Hochul), New Mexico’s Mark Ronchetti (challenging Democratic incumbent Michelle Lujan Grisham), and Minnesota’s Scott Jensen (challenging Minnesota Democratic incumbent Tim Walz) leading. It’s possible a GOP wave that’s higher than expected could produce surprises in these contests. But more likely the partisan leanings of these three blue states will determine the outcome. Early on, Republican also hoped former two-term Maine governor Paul LePage could engineer a comeback against his Democratic successor, Janet Mills. But he’s run an uninspiring campaign and Mills has led in every public poll. And unlike his MAGA counterparts in other states who got some high-life national help, Pennsylvania Republican nominee Doug Mastriano was abandoned to his fate against Democrat Josh Shapiro, who is expected to win easily.

Similarly, Democrats had high hopes that in Texas Beto O’Rourke could throw a serious scare into Republican incumbent Greg Abbott; that Charlie Crist could trip up Florida Republican incumbent Ron DeSantis, and even that Joy Hofmeister could upset Republican incumbent Kevin Stitt in profoundly red Oklahoma. O’Rourke and Crist both faded down the stretch, and late polls show Stitt beginning to restore order in Oklahoma;FiveThirtyEight now gives him a 91 percent probability of winning.

According to the Cook Political Report, 64 of the 435 U.S. House races are competitive. Republicans are favored to flip ten specific Democratic-held seats, while Democrats are favored to flip two seats now held by Republicans. More importantly, of the 36 races rated as toss-ups, 26 are for seats currently held by Democrats, with just ten in Republican-held districts. So Democrats are very much on the defensive with just a five-seat margin going into Election Day.

Based on both history (only twice since 1934 has the party controlling the White House made midterm gains) and public opinion, the odds are high that Republicans will gain control of the House (FiveThirtyEight calculated an 84 percent probability that will happen). The public-opinion factor, however, isn’t that one-sided. For the best predictor of the national House popular vote, the generic congressional ballot, the RealClearPolitics polling averages show a significant but hardly overwhelming 2.6 percent Republican advantage (by comparison, Democrats won the House national popular vote by 8.6 percent in 2018 and flipped 41 seats).

One of the credible and objective outlets willing to make specific House projections, Larry Sabato’s Crystal Ball, currently estimates Republican gains of 24 net seats. Similarly but more cautiously, Cook Political Report projects “a GOP gain of between 15 and 30 seats, with more uncertainty than usual.”