On the menu today: As much as I may feel in my heart that the top story this morning is the New York Jets beating the previously 6-1 Buffalo Bills, our attention must refocus on tomorrow’s midterm elections. This is just about the latest possible date for final predictions in the battle for control of Congress, so below are the final calls on control of the House and the big races in the Senate. (Tomorrow, I’ll go through the governor’s races.)
Red Dawn
I guess I’ve put off my election predictions for as long as I can. I usually prefer to hold off until right before Election Day, because you never know when some October surprise — or more accurately, a November surprise — will come along and change the dynamic, or give those last few undecided voters a reason to break a certain way.
But I think one of the big stories of the 2022 midterms is how much the dynamics haven’t changed. This looked like a big GOP wave year back in the spring; it looked like a more modest GOP wave year back in midsummer; and it’s steadily looked more and more like a big GOP wave year as autumn progressed. I also think a lot of the “Democrats will keep the House and gain seats in the Senate” talk since summer amounted to Democratic and media wish-casting, seeing what they wanted to see instead of what was there.
Throughout the summer and fall, I thought Republicans were going to win 20-25 House seats and finish with at least 51 Senate seats — a “pretty good” wave year. (At least, I think that’s the answer I remember giving on The Editors podcast.) Now that we’re at the election’s eve, I think we’re on the higher end of a red-wave year, and approaching “red tsunami” territory.
In some circles on the right, it’s become almost trendy to dismiss all polling as skewed and too inaccurate to be of any use. I believe that polls can give you a ballpark sense of the state of a race. Yes, every once in a great while, all of the public polling will be wildly off base, such as in the Maine Senate race in 2020. Sometimes, you’ll have one particular pollster that is way off from everyone else, such as when Quinnipiac kept showing a tie race in the South Carolina Senate race in 2020, and that helps shape the narrative in a not-so-accurate direction. But we remember those races because they stood out for being so far away from the final results. Democrat Mark Kelly led almost all of the polls in Arizona in 2020, and he won — albeit by about three percentage points fewer than his final result in the RealClearPolitics polling average. Republican Joni Ernst led most, but not all, of the polls in Iowa and ended up winning by 6.6 percentage points — about four points higher than her RealClearPolitics polling average.
By and large, if a candidate is the leader outside the margin of error by the end of the race, they’re going to be the winner when all the votes are counted. The challenge is when they have a lead within the margin of error, such as Cal Cunningham’s 2.6 percentage-point lead in the final RCP average of the 2020 North Carolina Senate race. Republican Thom Tillis eked out a win by 1.2 percentage points. Polls are going to have a tougher time accurately showing which candidate is ahead when the race is close.
“But Jim, the polls always understate the Republicans support for the Republican candidate,” you say? First, not always, and second, the public polls are already pointing to a big GOP wave. So even if you think there’s a systemic problem of reaching enough white, blue-collar voters — a GOP-leaning demographic — this points to a GOP wave so large, not even a systemic polling flaw could obscure it.
In the Battle for the House . . .
The Cook Political Report has 212 seats rated solid, likely, or lean Republican. Larry Sabato’s Crystal Ball says the GOP has 219 seats that are rated safe, likely, or leans Republican. RealClearPolitics has 227 seats rated solid, likely, or lean Republican.
For the past week or so, my back-of-the-envelope math envisioned a GOP House majority somewhere between 229 and 241, and I’m sticking to that. Give the Republicans the 212 seats in Cook Political Report, with two-thirds of the 35 races in the toss-up category, and you end up with 235 Republicans and 200 Democrats, so put those down as my final prediction numbers.
The only specific U.S. House races I will make predictions for are three in my home state of Virginia, in part because they provide useful barometers of the scale of the GOP wave, and partially because Virginia polls close relatively early — 7 p.m. Eastern time. (Also note that Virginia does not have a senatorial or gubernatorial election this year, so there is no “top of the ticket.” House members must win or lose on their own merits.)
Virginia’s second district, which includes Virginia Beach and Norfolk, is arguably the most competitive in the country, and I think that if this turns out to be a normal wave year, incumbent Democrat Elaine Luria will fall to Jennifer Kiggans. As of November 1, Republicans and their allied groups had spent almost $2 million more than the Democrats had in this race. If Luria hangs on, it will be an early indicator that Democrats are doing better than expected. A year ago, Glenn Youngkin beat Terry McAuliffe here, 55 percent to 44 percent.
In Virginia’s seventh district, I suspect Abigail Spanberger will hang on against Yesli Vega, partially because redistricting made this district a bit more Democratic-friendly, and partially because Democrats and their allied groups dumped enormous resources into keeping this seat — nearly $20 million by November 1. (You used to be able to afford to run a Senate campaign with that kind of money!) Last year, Youngkin won here 52 percent to 47 percent. If Spanberger loses, it’s an early indicator that Democrats are doing worse than expected.
In Virginia’s tenth congressional district, one of my favorite GOP challengers of the cycle, Hung Cao, is trying to beat incumbent Jennifer Wexton in a tough district — under the new lines, Youngkin was just short of McAuliffe last year, and Democrats usually win with well more than 50 percent. Cao’s a good candidate in a wave year, with Wexton winning the ad war by a large margin, working in a fairly blue district, If Cao wins, we’re moving from “red wave” to “red tsunami.”
In the Battle for the Senate . . .
My first prediction is that late on Election Night and into the wee hours of Wednesday morning, we won’t know which party will control the Senate. Our Brittany Bernstein wrote up a much-needed review of when we are likely to hear the results of key Senate races.
Pennsylvania and Wisconsin do not allow election officials to begin validating mail-in ballots until Election Day, which slows down the counting process compared to other states. Leigh Chapman, Pennsylvania’s acting secretary of the commonwealth and top election official, told the New York Times that she expects “results to take at least a few days.” This is exactly what you don’t want to hear when you want to dispel conspiracy theories that an election is being stolen or results are being altered by behind-the-scenes shenanigans. Smaller, more sparsely populated Republican counties will report first, and larger, more densely populated Democratic counties and cities will report later. The early returns will show Mehmet Oz leading by a dramatic margin, and then Philadelphia and Pittsburgh will show John Fetterman gaining quickly and likely taking the lead. Many Republicans, likely egged on by Donald Trump, will contend that the election was “rigged” or stolen and that there’s no way Oz could have lost to a man who can barely communicate.
In other words, I think Dr. Oz will fall just short in Pennsylvania. He’s run a steadily improving campaign, evolving from a Ford Pinto into a Little Engine That Could. This is the call I’m least confident in, because I think there’s a chance that a small but decisive percentage of Pennsylvanians are telling pollsters they’re voting for Fetterman and just don’t want to admit they’re voting against the guy who had a serious stroke.
In Wisconsin, Riley Vetterkind, a spokesman for the Wisconsin Elections Commission, told the New York Times that, “Given that unofficial results in both the 2018 and 2020 general elections weren’t largely complete until the early morning hours after Election Day, it would be reasonable to assume it may take until then for unofficial results to be tabulated and posted this year.” I think Ron Johnson will win reelection over Mandela Barnes by a surprisingly comfortable margin, but we may not see an official call until Wednesday afternoon.
New Hampshire officials are warning that an anticipated increase in write-in votes could lead to results being reported later than normal on Election Night. In New Hampshire, localities decide when polls open and close, but they cannot close before 7 p.m. I think Don Bolduc will beat incumbent Democrat Maggie Hassan, but it is likely to be very close, and we will not see a declared winner until later in the week.
In Ohio, polls close at 7:30 p.m. Eastern, and I expect J. D. Vance will quickly be declared the winner over Tim Ryan. Vance is pulling away in the polls, and the worries of early August are long forgotten. It turns out that Ryan only looked like a strong candidate when he had the airwaves to himself.
In Colorado, most ballots are sent by mail, and results are reported by the counties to the state’s reporting system at 7 p.m. local time, which is 9 p.m. Eastern. I think Republicans will get their hopes up as it takes longer than expected for organizations such as the Associated Press to call a winner, but incumbent Democrat Michael Bennet will indeed beat Republican Joe O’Dea. (If O’Dea does win, remember the Democrats’ odd desperation to tie O’Dea to Trump in the final weeks of the race.)
Washington State votes mostly by mail as well, but some voters choose to cast ballots in person on Election Day, with polls that close at 8 p.m. local time, which is 11 p.m. Eastern. Tiffany Smiley has run about as good a campaign as any Republican can expect, and if Washington Democrats had been caught napping, Smiley might have pulled off the upset against longtime incumbent Patty Murray. Alas, Democrats spent a lot to ensure Murray serves her sixth term in the Senate.
In Arizona, the secretary of state’s office begins posting results at 8 p.m. local time, which is 10 p.m. Eastern, and state law permits the tabulation of early ballots prior to Election Day. More than 1.4 million Arizonans have cast ballots already. To me, this race and Pennsylvania are the two most difficult in which to project a winner: purple-ish states with incumbent Democrats who looked safer a month or two ago, against a Trump-endorsed Republican with all the momentum. My prediction is that Democrat Mark Kelly hangs on against Blake Masters by the skin of his teeth, but it wouldn’t surprise me if it was a narrow win the other way. It’s easy to envision a scenario where GOP gubernatorial candidate Kari Lake ends up winning the governor’s race by a large margin and pulls Masters over the top.
In Nevada, polls close at 7 p.m. local time, which is 10 p.m. Eastern. Certain Nevada localities have voted to do a hand-count of the ballots, and then use a machine count to verify the results. Republican Adam Laxalt has led consistently, and Catherine Cortez Masto’s share of the vote is terrible for an incumbent — as low as 43 percent in certain polls. I know the unions and the old Harry Reid machine can help Democrats overperform their final numbers, but the headwinds just feel too strong for Cortez Masto to overcome. (Nevada’s inflation rate is worse than the national average, and it’s the dominant issue on people’s minds.)
In Georgia, I now think that Herschel Walker will finish ahead of incumbent Democrat Raphael Warnock, but I think he will finish just short of 50 percent, triggering a December 6 runoff, which I think Walker will win. (Also remember that libertarian candidate Chase Oliver has been getting 2 or 3 percent in polls, and that’s within the normal range for a third-party candidate.)
By midnight Eastern, we probably won’t have sufficient results to declare a winner in Pennsylvania or Wisconsin, and Georgia is likely to go to a runoff. We might have a strong indication there’s going to be a GOP Senate majority if New Hampshire has enough of the votes tabulated, pointing to an insurmountable lead for Don Bolduc.
The under-discussed stories of this year are the Senate races that Democrats hoped would be competitive that just never became real races. In Florida, two-term incumbent Marco Rubio will beat Val Demings handily, which should make people ask whether those major mainstream-media columns touting Demings’ chances in late September were wishful thinking.
In Utah, two-term incumbent Mike Lee will beat Evan McMullin handily, which should make people ask why they were still seeing major mainstream-media columns touting McMullin’s chances in mid October.
Beating an incumbent senator is difficult. Beating a two-term senator is even more difficult. Beating a two-term senator in a state that usually favors that senator’s party is really difficult. And beating a two-term senator in a state that usually favors their party in a wave year that favors that senator’s party is just short of impossible. Whether or not you like to hear something does not affect its truthfulness.
In Iowa, that Des Moines Register poll in mid October that showed incumbent GOP senator Chuck Grassley only ahead by three points was probably just an unusual sample; the newest DMR poll showed Grassley up by twelve. Yes, that’s the closest race Grassley has run since 1986. No, that factoid doesn’t really matter that much, and yes, that should make people wonder why they were seeing major mainstream-media columns in mid October touting this as “the sleeper race of the year.”
With Bolduc, Laxalt, and Johnson winning, I come out to a 51–48 GOP advantage by the end of the week, with Walker and Warnock headed to a runoff. It wouldn’t shock me if Oz or Masters or both won, giving Republicans a 53- or 54-seat majority.
ADDENDUM: In case you missed it yesterday, you are likely to see some progressives scapegoating Joe Biden’s on-and-off “bipartisanship,” or Democrats’ supposed tendency to be “too nice” if the expected GOP wave materializes tomorrow for the expected GOP wave. But thankfully, you’re not the kind of reader who would buy into those excuses.