On October 17, 2022, I published an updated set of projections for who will have the majority in the United States House of Representatives (“House”) following the 2022 elections, as well as who will win the 35 elections for United States Senate (“Senate”) and 36 elections for governor. Since then, an additional 17 generic ballot polls – which I use to estimate the probability Democrats retain their House majority – have been released, along with 50 new Senate polls and 55 new governor’s race polls.
Entering the 2022 election cycle, there were two truisms. One, Republicans were expected to have a very good year because it was the first midterm elections of a new Democratic administration; Joseph R. Biden, Jr.’s middling approval rating would only exacerbate that problem. Two, polling had substantially underestimated Republican strength in the past three election cycles.
The first is undoubtedly true, at least historically: the non-White-House party has gained an average 23.7 seats in the nine such elections since 1962, jumping to 37.5 in the last four. And Biden’s approval rating has hovered between 37.7 and 45.5% over the last 12 months, settling down to 41-43% since Labor Day.
The second, however, is more problematic. My own analysis, following the 2020 elections, was that I overestimated the final state-level margin between Biden and President Donald J. Trump by an average 3.4 percentage points (“points”) across all states, and by 5.9 points in states won by Trump. The story was similar for Senate races: Democratic overestimate of 5.8 points overall, jumping to 8.3 points in states with Republican victors. Essentially, polling severely undercounted Republican voters in the most Republican states, while performing reasonably well elsewhere. At the same time, David Wasserman of the Cook Political Report calculated polls undercounted Democratic voters in Arizona, Nevada and Texas in the 2016 and 2018 elections for president, Senate and governor by an average of 2.9 points; I have not seen a similar calculation for 2020.[1]
Table 1 lists time-adjusted averages for every pollster to conduct at least three generic ballot polls and whom FiveThirtyEight.com calculates overestimated Democratic margins by at least 2.1 points in recent elections; this is the amount I adjust their polls when calculating weighted-adjusted polling averages (WAPA).
Table 1: Time-weighted generic ballot averages by pollster
| Pollster | Rating | Bias | Pollster Average | Overall Average* | Difference |
| Change Research | B- | -2.7 | |||
| Democratic | 46.8% | 43.3% | D+3.5 | ||
| Republican | 42.2% | 41.1% | R+1.1 | ||
| Cygnal | B+ | -2.1 | |||
| Democratic | 45.9% | 44.2% | D+1.7 | ||
| Republican | 46.6% | 43.5% | R+3.1 | ||
| Data For Progress | B | -3.3 | |||
| Democratic | 44.9% | 45.1% | D-0.2 | ||
| Republican | 47.3% | 44.1% | R+3.2 | ||
| GQR Research | B | -2.1 | |||
| Democratic | 49.2% | 44.2% | D+5.0 | ||
| Republican | 47.2% | 43.4% | R+3.8 | ||
| Monmouth University | A | -2.1 | |||
| Democratic | 44.9% | 44.5% | D+0.4 | ||
| Republican | 48.3% | 44.1% | R+4.2 | ||
| Morning Consult | B | -2.9 | |||
| Democratic | 45.6% | 44.4% | D+1.2 | ||
| Republican | 43.0% | 43.9% | R-0.9 | ||
| SSRS | C | -2.7 | |||
| Democratic | 47.0% | 44.4% | D+2.6 | ||
| Republican | 46.0% | 43.8% | R+2.2 | ||
| AVERAGE | B | -2.6 | |||
| Democratic | 46.3% | 44.3% | D+2.0 | ||
| Republican | 45.8% | 43.4% | R+2.4 |
*Over the same time period as polls conducted by this pollster
These seven pollsters gave Democrats a 0.5-point lead on the generic ballot, while other pollsters during the same period gave Democrats a 0.9-point lead, meaning they were biased 0.4 points in favor of Republicans. However, I have been deducting an average of 2.6 points from the Democratic margin on these polls. That is, I have been underestimating Democratic strength by a full 2.2 points just from these seven pollsters. This strongly suggests they – and especially Data For Progress (R+3.4) and Monmouth University (R+3.8) – have “corrected” their past Democratic lean, even as I continue to deduct from the Democratic margin in their polls.
Using the 10 generic ballot polls whose field data median centered on October 1, 2022 Democrats led by 0.9 points. Using the 10 polls centered on October 16, 2022, though, Democrats trail by -2.4 points. This, along with confirmation bias and media incentive for electoral drama has caused the conventional wisdom to congeal around a “Republicans surging!!” narrative.
This narrative is – misleading – for several reasons.
First, polls measure hypothetical votes, not actual votes. In fact, if either party is “surging,” it is the Democrats, who are banking millions of early votes, according to data collected by University of Florida Professor Michael McDonald. Overall, more than 7.5 million votes have already been cast; if recent patterns hold, they will be overwhelmingly Democratic. Indeed, of the 4.1 million votes cast in the 17 states which release partisan registration data, 50.4% were cast by registered Democrats, 30.2% by registered Republicans and 19.4% by non-party registrants.
Second, Republican-leaning pollsters have “flooded the zone” lately. Thus, the 3.3-point shift toward Republicans on the generic ballot includes a Monmouth University poll showing Democrats trailing by six points – from which I subtracted and additional 2.1 points (instead of adding 3.8 points). It also includes polls from extremely Republican-leaning Rasmussen Reports (R+2.5), as well as from Public Opinion Strategies (R+1.5) and McLaughlin & Associates (R+1.6). Finally, an October 18-19, 2022 Emerson College poll showing Democrats down by five points follows a September 21-22, 2022 Emerson College poll showing a tie. A five-point swing Republican, right? Well, yes and no. The Democratic percentage dropped from 45 to 41, while the Republican percentage increased from 45 to 46, meaning most respondents switched from Democratic to undecided, not to Republican.
Finally, the overall distribution of Senate seats and governors I project after Election Day has barely changed since Labor Day. For example, the first Senate forecast I published gave Democrats an 82.0% chance to maintain their majority. The current forecast gives them an 82.1% chance, albeit centered around a smaller net gain. And while the likelihood Democrats retain the House has declined somewhat, it is hardly in line with a huge Republican surge.
Basically, everyone who expected a Republican surge is seeing exactly that – if they look at the polls just the right way. They also likely assume the 10% or so of the population yet to commit in the polls will break decisively for Republicans. It is true that, over time, undecideds have broken a bit more Republican on the generic ballot: the correlation between Republican and undecided percentages in the rolling 10-poll average is -0.81. But there is no reason to think that trend will continue. Given how cross-pressured potential voters appear to be – torn between concerns about inflation and crime on one hand and abortion and democracy on the other – I might argue they are more likely to break essentially even, perhaps even slightly Democratic given the poor quality of many Republican candidates.
In other words, the “Republican surge” is something like one part genuine shift in voter preference and two parts artifactual/methodological, to be taken with many grains of salt.
***********
The House. As of October 23, 2022, I estimate Democrats lead Republicans on the generic ballot by 0.2 points, 44.3% to 44.1%. Based upon this value, I estimate the probability Democrats retain a House majority is 27.5%. This is a decrease of 2.2 points in one week. I estimate Democrats will lose 10.2 seats, putting them at 211 or 212 (compared to 222 following the 2020 elections), with a 95% prediction interval of -31.9 to +10.9.
Using only the 89 generic ballot polls conducted entirely after Labor Day, though, Democrats lead by 0.5 points, 45.2% to 44.7%, which equates to a 30.5% chance of retaining their House majority and an average loss of 9.2 seats (95% CI: -29.2 to +13.5). These values are down noticeably since last week. For Democrats to be nominal favorites to retain their House majority (p>50%), they need to win nationally by at least 2.3 points.
For the first time, my estimates diverge somewhat from those from FiveThirtyEight.com, whose probabilities range from 20% (Deluxe) to 27% (Classic), with 211 seats assigned >50% probability of being won by the Democrat in the methodologically-conservative Deluxe model. However, if Democrats win every seat with probability≥35%, they maintain a bare House majority across all three models.
Bottom line: A narrow House Republican majority remains more likely than not, with Democrats’ position clearly declining for the first time.
***********
The Senate. As of October 23, 2022, 393 publicly-available Senate election polls have been conducted since November 1, 2021. These include five polls assessing the likely final matchup between two Republicans in Alaska: incumbent Senator Lisa Murkowski and former state official Kelly Tshibaka; Murkowski currently edges Tshibaka 51.7% to 48.3%. Only the Senate elections in Hawaii – which incumbent Democrat Brian Schatz is a near-lock to win – and in Alabama, Idaho and North Dakota – which Republicans Katie Britt, Michael Crapo and John Hoeven, respectively, are near-locks to win – do not yet have publicly-available polling.
Table 2 shows updated projections for all 35 Senate elections. I made one small change to how I calculate WAPA, weighting 50% lower any polls whose Democratic and Republican percentages sum to 100%; this ignores undecided and third-party voters.
Table 2: Democratic Senate win probabilities and projected final margins, 2022
| State | Fundamentals | Polling (WAPA) | Poll Wt | Final Projection | |||
| Democratic seat | Margin | P(D win) | Margin | P(D win) | Margin | P(D win) | |
| Hawaii | 34.7 | 100.0% | n/a | n/a | n/a | n/a | 100.0% |
| Vermont | 29.1 | 100.0% | 22.3 | 100.0% | 0.84 | 23.3 | 100.0% |
| Maryland | 31.9 | 100.0% | 23.0 | 100.0% | 0.70 | 25.7 | 100.0% |
| California | 26.9 | 100.0% | 23.6 | 100.0% | 0.87 | 24.0 | 100.0% |
| New York | 25.9 | 100.0% | 17.7 | 100.0% | 0.95 | 18.1 | 100.0% |
| Connecticut | 19.6 | 99.6% | 12.6 | 100.0% | 0.83 | 13.9 | 99.9% |
| Washington | 19.4 | 99.6% | 8.4 | 100.0% | 0.92 | 9.3 | 100.0% |
| Illinois | 19.0 | 99.5% | 16.6 | 100.0% | 0.85 | 17.0 | 99.9% |
| Oregon | 15.8 | 98.3% | 14.0 | 100.0% | 0.89 | 14.2 | 99.8% |
| Colorado | 11.4 | 93.3% | 7.8 | 99.8% | 0.87 | 8.3 | 98.9% |
| New Hampshire | 6.9 | 77.7% | 6.7 | 100.0% | 0.87 | 6.8 | 97.2% |
| Nevada | 5.2 | 72.7% | -0.3 | 46.9% | 0.85 | 0.5 | 50.8% |
| Arizona | -4.1 | 39.7% | 6.2 | 100.0% | 0.86 | 4.8 | 91.5% |
| Georgia | -4.5 | 38.3% | 2.1 | 74.6% | 0.88 | 1.3 | 70.1% |
| Estimated Democratic seats lost = 0.9 (range = 0.3 to 1.6)* | |||||||
| Republican seat | |||||||
| Pennsylvania | -2.1 | 47.0% | 5.6 | 99.8% | 0.87 | 4.6 | 92.9% |
| Wisconsin | -5.5 | 34.6% | -2.1 | 22.3% | 0.90 | -2.4 | 23.5% |
| Florida | -8.6 | 24.7% | -5.3 | 0.3% | 0.83 | -5.9 | 4.4% |
| North Carolina | -5.6 | 34.4% | -1.8 | 31.2% | 0.89 | -2.2 | 31.6% |
| Ohio | -9.6 | 22.0% | -0.4 | 45.4% | 0.86 | -1.7 | 42.1% |
| Iowa | -12.9 | 13.9% | -8.6 | 0.0% | 0.90 | -9.0 | 1.4% |
| Alaska | -18.9 | 5.0% | n/a | n/a | n/a | n/a | 0.0% |
| South Carolina | -19.0 | 4.9% | -18.5 | 0.0% | 0.72 | -18.7 | 1.4% |
| Missouri | -18.8 | 5.2% | -12.0 | 0.1% | 0.90 | -12.7 | 0.6% |
| Indiana | -22.7 | 2.3% | -2.3 | 42.4% | 0.76 | -7.3 | 32.6% |
| Kansas | -24.4 | 1.6% | -14.9 | 0.0% | 0.73 | -17.5 | 0.4% |
| Louisiana | -25.4 | 1.2% | -23.2 | 0.0% | 0.84 | -23.5 | 0.2% |
| Utah | -30.7 | 0.3% | -8.5 | 1.5% | 0.79 | -13.2 | 1.3% |
| Alabama | -29.0 | 0.5% | n/a | n/a | n/a | n/a | 0.5% |
| South Dakota | -32.7 | 0.2% | -20.0 | 0.0% | 0.81 | -22.5 | 0.0% |
| Kentucky | -33.4 | 0.1% | -15.5 | 0.0% | 0.39 | -26.4 | 0.1% |
| Arkansas | -33.4 | 0.1% | -17.6 | 0.0% | 0.73 | -21.9 | 0.1% |
| Idaho | -37.9 | 0.0% | n/a | n/a | n/a | n/a | 0.0% |
| North Dakota | -38.5 | 0.0% | n/a | n/a | n/a | n/a | 0.0% |
| Oklahoma (1) | -40.9 | 0.0% | -17.9 | 0.0% | 0.77 | -23.3 | 0.0% |
| Oklahoma (2) | -40.9 | 0.0% | -13.4 | 0.0% | 0.77 | -19.8 | 0.0% |
| Estimated Republican seats lost = 2.3 (range = 1.5 to 4.6) | |||||||
| Overall expected Democratic seats gained/lost = +1.4 (range = -0.1 to +4.3) |
*Adding 4 or subtracting 3 points to all WAPA, including generic ballot estimates
Italics indicate open seats, while boldface indicates a likely “flip.” “Fundamentals” is the sum of state’s partisan lean (3W-RDM), current generic ballot estimate (D+0.3) and incumbency advantage (Senate values: 5.5 points for Democrats, 3.3 points for Republicans); I assign one-third incumbency advantage to Democrats Mark Kelly of Arizona and Raphael Warnock of Georgia because they first won their seats in 2020 special elections. Essentially, this is what one would expect the Democratic margin to be for a generic Democratic vs. a generic Republican. The “fundamentals” probability the Democratic candidate wins is the likelihood of a Democratic margin of ≥0.0000001 given a normal distribution, with mean of +1.3 (mean historic “miss”) and standard deviation of 10.7.
Based solely on the “fundamentals,” Kelly and Warnock are modest underdogs, while Democratic incumbents in New Hampshire and Nevada are only modest favorites. On the flip side, the open Republican seat in Pennsylvania – vacated after two terms by Pat Toomey – is essentially a toss-up, albeit with the Republican candidate barely favored. Two other Republican-held seats – in North Carolina and Wisconsin – tilt Republican. On average, I now give 18% weight to fundamentals and 82% weight to WAPA, unchanged from last week.
Recall these “projection categories”:
Safe Republican (<5%)
Solid Republican (≥5 to <10%)
Likely Republican (≥10 to <20%)
Lean Republican (≥20 to <33%)
Tilt Republican (≥33 to <45%)
Toss-up (≥45 to ≤55%)
Tilt Democratic (>55 to ≤67%)
Lean Democratic (>67 to ≤80%)
Likely Democratic (>80 to ≤90%)
Solid Democratic (>90 to ≤95%)
Safe Democratic (>95%)
The difference between “Safe” and “Solid” is not especially illuminating for Republicans, so let us combine them:
Safe/Solid Republican (n=15). Incumbents John Boozman (AR), Crapo, Chuck Grassley (IA), Hoeven, John Kennedy (LA), Jim Lankford (OK), Jerry Moran (KS), Rand Paul (KY), Tim Scott (SC) and John Thune (SD) are near-locks to win. Republican nominees Britt, Markwayne Mullin and Eric Schmitt are also prohibitive favorites to win open seats in Alabama, Oklahoma and Missouri, respectively. Republican Senator Mike Lee is heavily favored – for now – to defeat Independent former CIA operative Evan McMullin (effectively the Democratic nominee) in Utah, while I project Murkowski has the edge over Tshibaka.
Likely Republican (n=2). Two polls have been conducted in the Indiana Senate election between Republican incumbent Todd Young and Democratic Hammond Mayor Thomas McDermott, Jr. These mediocre (B-/C+, using FiveThirtyEight.com pollster ratings) polls, one skewed Democratic and one skewed Republican, suggest Young leads McDermott by only 2.3 points, when he “should” be winning by more than 20 points. Moreover, neither candidate tops 42% on average, which combined with Libertarian candidate James Sceniak (6% in the most recent poll) suggests extreme volatility – which is why McDermott has a roughly one-in-three chance to win.
On paper, Florida’s Senate contest is Lean Republican, which is supported by incumbent Republican Marco Rubio leading overall by 5.3 points (n=31 polls; B-/B), and by 5.2 points in nine polls conducted since Labor Day (B/B+). What makes this election my “sleeper” is that Rubio still struggles to top 48.0%, meaning a Democratic upset IF undecided voters break about 80-20 for Demings.
Lean Republican (n=2). Incumbent Republican Johnson is now favored to win reelection in Wisconsin, in line with fundamentals. He still leads Democratic Lieutenant Governor Mandela Barnes 48.8% to 46.8%, in 17 polls (B+), and 49.2% to 46.4% in 13 polls released since Labor Day (B+).
On paper, Republicans are modest favorites to retain the Senate seat being vacated by Richard Burr in North Carolina (R+5.8). Unfortunately, recent polling suggests Democratic former State Supreme Court Chief Justice Cheri Beasley has lost ground against House member Ted Budd, trailing now by 1.8 points (43.8% to 45.6%). Moreover, she trails by 2.4 points in 11 polls conducted since Labor Day (B/B+). That said, Budd only averages 46.3% in these polls, meaning if Beasley wins undecided voters roughly 70-30, she wins.
Tilt Republican (n=1). Republicans should also be modest favorites to retain the Senate seat being vacated by Rob Portman in Ohio (R+9.8). And, in fact, Democratic House member Tim Ryan now trails Republican venture capitalist J.D. Vance 45.1% to 44.7% overall, and 45.6% to 44.6% points in 16 polls conducted since Labor Day (B+/B). Still, the fact Vance still cannot climb above 45.6% in a red state should be worrying to Republicans – as is the fact Ryan only needs to win about five of every nine undecided voters to win.
Toss-ups (n=1). In Nevada, six new polls show Democratic Senator Catherine Cortez Masto narrowing the gap with Republican former state Attorney General Adam Laxalt. Laxalt only leads 43.8% to 43.5% overall, and 45.5% to 44.4% in 13 polls conducted since Labor Day (B/B-). With neither candidate topping 46% at any point – and with 10% of the electorate potentially up for grabs, this truly is a toss-up, though one in which incumbency and a history of Democratic candidates outperforming polling in Nevada give Cortez Masto perhaps the tiniest edge.
Tilt Democratic (n=0).
Lean Democratic (n=1). Ten polls (B+) have been conducted since October 3, 2022, the day the Daily Beast reported ardently pro-life Republican former National Football League star Herschel Walker had reimbursed his then-girlfriend for an abortion in 2009; that same day, Walker’s son Christian tweeted about his father’s history of abusive and violent behavior. In these polls, Warnock leads Walker 47.3% to 45.6%, with an average 3.3 points for other candidates. In eight polls (B/B+) conducted just prior to the revelations, Warnock led 46.9% to 45.6%, suggesting Warnock gained 0.4 points while Walker held steady. Overall (n=46, B/B+), Warnock leads 47.1% to 45.0%, and is ahead 47.3% to 44.9% in 20 polls (B+) conducted since Labor Day. Georgia is still a Republican-leaning state (R+6.5), though, so a runoff occasioned by neither candidate winning a majority on Election Day remains a possibility. If the race goes to a December runoff, FiveThirtyEight.com projects Walker to win an average 67% of the time, essentially unchanged since last week.
Likely Democratic (n=1). The Senate seat most likely to change partisan hands is that held by retiring Republican Pat Toomey in Pennsylvania. Democratic Lieutenant Governor John Fetterman leads Republican television celebrity Dr. Mehmet Oz overall 48.0% to 42.4% (n=34, B). However, Fetterman’s stroke and some consolidation of Republican voters by Oz are likely why the race has plateaued at a lower margin: in 16 polls (B+) conducted since Labor Day, Fetterman leads 47.4% to 43.5%, a 0.6-point drop for Fetterman and a 1.1-point increase for Oz; using only post-Labor-Day polling, Fetterman still has a win probability of 83.8% (with a projected margin of 3.3 points).
Solid Democratic (n=1). Kelly remains very close this week to winning the Arizona (R+6.1) Senate race. He now tallies 48.5% to 42.3 for Republican venture capitalist Masters overall (n=31, B/B-), and 48.2% to 42.5% in 21 post-Labor Day polls (B). I expect Kelly to win by four-to- five points.
Safe Democratic (n=12): Incumbents Michael Bennet (CO), Richard Blumenthal (CT), Tammy Duckworth (IL), Maggie Hassan (NH), Patty Murray (WA), Alex Padilla (CA), Chuck Schumer (NY), Schatz, Chris Van Hollen (MD) and Ron Wyden (OR) are near-locks to win reelection. Murray was effectively reelected on August 2, when she won 52.3% of the vote in an all-candidate primary; all Democrats combined won just under 55% of the vote. Meanwhile, House member Peter Welch is all-but-certain to win the Vermont Senate seat being vacated by seven-term Democrat Pat Leahy.
Bottom line: There is an 82.1% chance (80.4% using post-Labor-Day polls) Democrats hold at least 50 Senate seats – and thus maintain the majority. There is also a 66.9% chance Democrats net add Senate seats, with one being the most likely outcome (19.4%). There remains a non-trivial (13.2%) chance Democrats net lose one seat, giving Republicans a narrow 51-49 majority. Overall, there is a 69.5% chance Democrats land between a net loss of one and a net gain of two Senate seats, for an over/under of +0.5.
Figure 1: 2022 Senate election distribution
***********
Governors. As of October 23, 2022, 420 publicly-available gubernatorial election polls have been conducted since November 1, 2021. Only those in the open seat in Hawaii – which Democrat Josh Green is a near-lock to win – and in Alabama, Idaho, Tennessee, and Wyoming – which Republicans Kay Ivey, Brad Little, Bill Lee and Mark Gordon, respectively, are near-locks to win – do not yet have publicly-available polling. A net gain of three governor’s mansions by Democrats would give them 25, giving them parity for the first time since 2010. Table 3 shows updated values for all 36 gubernatorial elections in 2022.
Table 3: Democratic gubernatorial win probabilities and projected final margins, 2022
| State | Fundamentals | Polling (WAPA) | Poll Wt | Final Projection | |||
| Democratic seat | Margin | P(D win) | Margin | P(D win) | Margin | P(D win) | |
| Hawaii | 29.2 | 94.0% | n/a | n/a | n/a | n/a | 94.0% |
| California | 35.5 | 97.1% | 23.5 | 100.0% | 0.87 | 25.0 | 99.6% |
| New York | 20.4 | 86.1% | 9.6 | 100.0% | 0.88 | 10.9 | 98.3% |
| Rhode Island | 22.0 | 87.9% | 13.3 | 91.6% | 0.85 | 14.5 | 91.1% |
| Connecticut | 24.5 | 90.3% | 11.7 | 100.0% | 0.86 | 13.6 | 98.6% |
| Illinois | 23.9 | 89.8% | 13.7 | 100.0% | 0.81 | 15.6 | 98.1% |
| Oregon | 10.3 | 70.7% | -0.8 | 44.5% | 0.82 | 1.2 | 49.2% |
| New Mexico | 16.9 | 81.5% | 7.4 | 97.5% | 0.94 | 8.0 | 96.6% |
| Colorado | 16.3 | 80.6% | 14.1 | 100.0% | 0.88 | 14.3 | 97.6% |
| Maine | 15.1 | 78.8% | 9.9 | 100.0% | 0.81 | 10.9 | 95.9% |
| Minnesota | 12.4 | 74.4% | 6.5 | 99.1% | 0.94 | 6.8 | 97.7% |
| Nevada | 10.1 | 70.3% | -1.0 | 40.6% | 0.86 | 0.6 | 44.9% |
| Michigan | 9.9 | 69.9% | 7.6 | 100.0% | 0.85 | 7.9 | 95.5% |
| Pennsylvania | -2.1 | 45.3% | 9.4 | 100.0% | 0.89 | 8.1 | 93.7% |
| Wisconsin | 8.2 | 66.7% | 0.6 | 55.8% | 0.90 | 1.4 | 56.9% |
| Kansas | -10.7 | 28.2% | 3.2 | 65.7% | 0.78 | 0.1 | 57.3% |
| Estimated Democratic seats lost = 2.4 (range = 1.2 to 3.4)* | |||||||
| Republican seat | |||||||
| Vermont | 15.2 | 79.0% | -35.7 | 0.0% | 0.84 | -27.6 | 12.5% |
| Maryland | 26.4 | 92.0% | 27.7 | 100.0% | 0.87 | 27.5 | 98.9% |
| Massachusetts | 26.3 | 91.9% | 25.1 | 100.0% | 0.92 | 25.2 | 99.4% |
| New Hampshire | -12.5 | 25.0% | -17.1 | 0.0% | 0.87 | -16.5 | 3.2% |
| Florida | -19.2 | 15.1% | -7.5 | 0.0% | 0.83 | -9.5 | 2.5% |
| Arizona | -5.9 | 37.4% | 0.0 | 50.1% | 0.88 | -0.7 | 48.6% |
| Georgia | -13.2 | 23.9% | -6.1 | 0.0% | 0.89 | -6.9 | 2.7% |
| Iowa | -23.5 | 10.4% | -16.7 | 0.0% | 0.97 | -17.0 | 0.3% |
| Ohio | -23.5 | 10.4% | -17.9 | 0.0% | 0.90 | -18.5 | 1.0% |
| Texas | -25.7 | 8.4% | -7.3 | 0.0% | 0.86 | -9.9 | 1.2% |
| Alaska | -29.5 | 5.7% | -8.6 | 0.0% | 0.78 | -13.3 | 1.3% |
| South Carolina | -29.6 | 5.6% | -7.3 | 0.0% | 0.73 | -13.2 | 1.5% |
| Nebraska | -24.9 | 9.1% | -5.2 | 30.7% | 0.76 | -9.9 | 25.5% |
| Tennessee | -40.9 | 1.4% | 1.4% | ||||
| Alabama | -42.9 | 1.1% | 1.1% | ||||
| South Dakota | -43.3 | 1.0% | -4.0 | 45.1% | 0.74 | -14.3 | 33.5% |
| Arkansas | -30.1 | 5.3% | -19.4 | 0.0% | 0.82 | -21.3 | 0.9% |
| Idaho | -48.5 | 0.5% | 0.5% | ||||
| Oklahoma | -51.5 | 0.3% | -3.7 | 22.6% | 0.78 | -14.3 | 17.7% |
| Wyoming | -61.2 | 0.1% | n/a | n/a | n/a | n/a | 0.1% |
| Estimated Republican seats lost = 3.5 (range = 3.0 to 4.4) | |||||||
| Overall expected Democratic seats gained/lost = +1.2 (range = -0.4 to +3.2) |
*Adding 4 or subtracting 3 points to all WAPA, including generic ballot estimates
Gubernatorial incumbency advantages are 10.4 points for Democrats and 13.9 points for Republicans. On average, I now give 15% weight to fundamentals and 85% weight to WAPA, an increase of 1.0 points in polling weight since last week.
Safe/Solid Republican (n=16). Incumbents Mike DeWine (OH), Mike Dunleavy (AK), Gordon, Ivey, Lee, Little, Henry McMaster (SC), Kristi Noem (SD), Kim Reynolds (IA) and Chris Sununu (NH) are near-locks for reelection. University of Nebraska Board of Regents member Jim Pillen and former White House spokesperson Sarah Huckabee Sanders will be the next Republican governors of Nebraska and Arkansas, respectively.
Meanwhile, Democrats Stacey Abrams (GA), Charlie Crist (FL) and Beto O’Rourke (TX) trail incumbent Republican governors Brian Kemp, DeSantis and Greg Abbott, respectively, by an average 7.5 points. Moreover, since Labor Day, all three incumbents average between 49.9% and 51.0%, effectively ending these elections.
Likely Republican (n=2). Popular Republican governor Phil Scott will win reelection, despite Vermont’s D+28.9 status.
Meanwhile, four mediocre post-Labor-Day-polls (C+/B-) give Stitt “only” a 3.6-point lead over Oklahoma (R+37.8) Democratic state Superintendent of Public Instruction Joy Hofmeister, 44.9% to 41.3%, with fully 13.8% not choosing either. Exclude a Republican-leaning American Viewpoint poll which had many more undecideds at the expense of Hofmeister, and the lead vanishes: 44.1% to 43.8%. Stitt should still win by nearly 20 points – something similar played out in the polls in 2018, when he won by 12.1 points. Except – single polls out of Nebraska and South Dakota also show closer than expected elections. The safe bet is for Republicans easily to win all three elections, but more data is clearly warranted.
Lean Republican (n=0).
Tilt Republican (n=0).
Toss-ups (n=3). Three governors who first won in the 2018 Democratic wave find themselves facing a tough reelection fight: Laura Kelly in Kansas, Steve Sisolak in Nevada and Tony Evers in Wisconsin. In Nevada, Sisolak now trails Clark County Sheriff Joe Lombardo 45.0% to 44.0% overall (n=24, B/B-), and 45.8% to 43.7% in nine polls (B/B+) conducted since Labor Day. And while I project him to win by 0.6 points based on incumbency, his inability to crack 44% should make Democrats very nervous.
The open governor’s race in strongly-Democratic Oregon (D+10.3) is a pure toss-up right now because of Independent Betsy Johnson, a former Democratic state senator, and a dearth of quality polling. Only 13 polls (B-) have been conducted overall, and they show Republican former State House Majority Leader Christine Drazan edging Democratic former State House Speaker Tina Kotek 37.1% to 36.3%, with Johnson at 16.3%. In nine polls conducted since Labor Day (B-), Drazan’s lead is a bit wider: 38.2% to 36.9%, with 15.2% for Johnson. As of now, Johnson voters appear to breaking evenly between Drazan and Kotek, who I still project to win by about one point.
The open governor’s race in Arizona – the first of three states in which a Democrat is at least even money to replace a retiring Republican governor – remains a toss-up. Democratic state Attorney General Katie Hobbs has a 48.6% chance to defeat Republican television personality Kari Lake in this R+6.1 state, dropping to 45.2% in post-Labor-Day polling; both values are up slightly this week. Lake continues to bump against a ceiling of 46.8% in those post-Labor-Day polls (n=17, B/B+), vs. 46.3% for Hobbs. Overall (n=26, B), the race is a pure tie: 46.4% to 46.4%.
Tilt Democratic (n=2). The good news for Kelly in heavily Republican Kansas (R+21.3) is that in two polls conducted at least in part after Labor Day, she leads Schmidt 46.9% to 42.6%, putting her much closer to victory; overall, in four polls, she leads 46.6% to 43.4%. I project her to eke out a narrow victory, one closer to three points than 0.6 points.
The story is similar in Wisconsin, where Evers’ lead over businessman Tim Michels has dropped to 47.4% to 46.8% overall (n=16, B+), and to just 47.2% to 47.0% since Labor Day (n=11, B+/A-). I still expect Evers to win by just over one point, but a narrow loss is also quite possible.
Lean Democratic (n=0).
Likely Democratic (n=0).
Solid/Safe Democratic (n=14). The other states which Democrats are near-locks to flip are Maryland and Massachusetts, with both former Army Captain Wes Moore and state Attorney General Maura Healey projected to win by around 25 points. Two other Democratic statewide officials are heavy favorites to win governor’s mansions held by retiring Democrats: Lieutenant Governor Green in Hawaii and Attorney General Josh Shapiro in Pennsylvania.
That leaves nine incumbent Democrats who are at least 9-1 favorites to win reelection: Kathy Hochul (NY), Ned Lamont (CT), Michelle Lujan Grisham (NM), Dan McKee (RI), Janet Mills (ME), Gavin Newsom (CA), Jared Polis (CO), J.B. Pritzker (IL), Tim Walz (MN) and Gretchen Whitmer (MI).
Bottom line: Democrats remain modest favorites to net at least one governor’s mansion (58.2% – 57.5% using post-Labor-Day polling), most likely (37.9%) between one and three. However, there is also a 28.5% chance Democrats net lose governor’s mansions, jumping to 31.5% using post-Labor-Day polling, with a loss of two being the likeliest outcome overall (18.6%). There is a 74.9% chance Democrats land somewhere between a net loss of two and a net gain of three governor’s mansions, for an over/under of +0.5.
Figure 2: 2022 gubernatorial election distribution
***********
While Democrats find themselves in a decent position – at least modest favorites to net Senate seats and governor’s mansions, while limiting losses in the House to 10 or so seats – it would only take a polling miss of 3 points in Republicans’ favor (relative to WAPA) to flip the script. In this scenario, Democrats still likely gain the Senate seat in Pennsylvania (83.5%), while losing in Nevada (27.6%); Senate control would hinge on Georgia (39.2%). Democrats would still win back governor’s mansions in Maryland and Massachusetts, while likely losing in Nevada (23.5%), Kansas (44.8%), Oregon (32.8%) and Wisconsin (32.6%). And the likelihood of retaining House control would drop to 7.0%, centered around a loss of 19-20 seats.
However, it is equally plausible Democrats overperform polls by a similar amount – an increase of 4 points relative to my WAPA. In this scenario, Democrats would be favored (61.1%) to retain House control, essentially holding serve or losing one seat. Not only would Democrats flip Pennsylvania while holding all their Senate seats (with Nevada closest at 86.3%), they would also add North Carolina (77.9%), Ohio (82.9%) and Wisconsin (82.7%) for a gain of four seats – with Florida (42.9%) and Indiana (45.8%) now in play. They would still top out at netting three governor’s mansions, albeit with Nebraska (37.6%), Oklahoma (41.0%) and South Dakota (37.4%) possibly in play.
As of mid-October, though, I still prefer to drive straight down the middle, while awaiting new polling, especially on the generic ballot, to see if there are signs how undecided voters are breaking.
If you have read to this point, a) thank you very much and b) you are clearly a close student of American politics, meaning there are two folks I urge you to follow on Twitter. One is Wasserman (@redistrict), and the other is Michael McDonald (@ElectProject). As the latter is the tabulator of early votes, I note that I voted on Sunday at the Brookline (MA) Town Hall.
Until next time, if you are not already registered to vote, I urge you to do so immediately. And if you like what you read on this website, please consider making a donation. Thank you.
[1] I believe this sheet was attached to Wasserman’s Twitter account in October 2020.
Just Bear With Me... 
2 thoughts on “2022 Elections Update: Taking Partisan “Surges” with a Grain of Salt”