Monday, November 7, 2022

BTRTN 2022 Official Midterm Election Predictions: The GOP Will Take the House and Fate of the Senate Will Await a Runoff in Georgia

At last, here are Tom’s official BTRTN’s predictions (not a snapshot!)….but he reserves the right to come back tomorrow (again at 5 PM ET) if the final flurry of polls released in the morning reveal any material shifts.

Roughly a month ago, The New York Times made the following statement:  “…the final stretch of the 2022 midterms defies predictability.”  (Emphasis mine.)  For those of us who predict elections, that is a daunting statement.  They could not be more right – and that statement was made before a number of races tightened considerably in the last few weeks.

So it takes an awful lot of chutzpah for anyone to make predictions about the 2022 elections.  With increasing polarization and tight margins, it was already a reasonably difficult task, but the 2022 election cycle has brought its own bag of unique goodies to bedevil us.  There is redistricting coming out of the 2020 Census; the ongoing and thus far baffling polling challenge of finding and properly sampling those non-responsive Trump supporters, overlaid this year by a flood of GOP partisan polls; the advent of early voting and mail-in voting, so that this cycle resembles neither 2018 nor 2020; the potential swelling of the youth vote and suburban women turnout, in response to Dobbs, which could mean those segments have been undersampled in the polls; the mixed-bag nature of the economy, clouding the usually reliable “it’s the economy, stupid” conventional wisdom; and the “Trump Factor” of an ex-president looming large over the proceedings in a precedent-shattering manner.

All that on top of a polarized country that has pushed our representation to 50/50 status in the Senate and nearly even in the House as well, so all it takes is a few lost races to shift control of our Houses of Congress.  The governorships are not quite split in half but nearly so.  And finally, we have more truly razor-thin races than we have seen in recent times.

Whew!  Well, this is my one chance to cover my posterior for posterity, so I am taking full advantage.  I feel like I know how to throw the darts pretty straight, if only that darned dartboard on the wall would stay put.

So, if chutzpah, is what it takes, well, chutzpah we shall bring, as our predictions follow below.  We will forecast the winner of each and every one of the 35 Senate, 435 House and 36 gubernatorial races – 506 separate races in total. 

Why should you listen to us?  We’ve had a few tough ones over the years, notably Hillary in 2016 and the House in 2020, but overall we stand by our track record.

This adds up to a 96%+ accuracy rate, and in races that were decided by less than 5 points, about a 75% rate.  That means that, in general, we miss about two Senate races per cycle – enough, this year, of course, to swing control of the Senate.  A good rule in life is “success not perfection,” but in the case of election forecasting, anything less than perfection is viewed by some as a massive miss.

One note:  if any reader wants our handy-dandy semi-famous spreadsheet that allow you to easily track Election Night (and beyond) outcomes for each of the Senate, House and Governor races (and see how we are doing against our predictions), just email us at and we will send them along to you.


First-term presidents usually find themselves getting a comeuppance in their first midterms, a rude awakening experienced by Ronald Reagan, Bill Clinton, Barack Obama and Donald Trump, who all lost dozens (and dozens) of House seats.  Turning campaign poetry into governing prose is not easy.  The Bushes escaped the full wrath of the electorate because of seismic events that occurred in the first two years of office that helped them considerably, the Gulf War (for 41) and 9/11 (for 43). 

Joe Biden came into office under terrible circumstances, including Covid and the aftershocks of January 6 and The Big Lie, resulting in a large minority of the nation not accepting the legitimacy of his presidency.  But he enjoyed a strong honeymoon (with an approval rating over 50%) in the early going by passing bipartisan legislation, introducing science into Covid management, repairing damaged relations with allies and bringing decency, integrity and empathy back to the White House.  But with a messy exit from Afghanistan and then rising inflation due to supply and labor shortages, the honeymoon ended and his approval rating dropped, and, by the spring of 2022, he seemed destined to take the same kind of lashing in the midterms as most of his predecessors.

But the summer brought the Dobbs decision, a flurry of legislative triumph, foreign policy success in Ukraine, massive legal problems for his predecessor and falling gas prices, all of which (and more) gave rise to hopes among Democrats that the party could hold onto both Houses of Congress after all.  Biden and the Democrats did enjoy a brief surge in their prospects in September, aided by a lackluster set of Senate Republican candidates in key swing states, including a trio of B-List celebrities backed by Trump who were political novices and committed gaffe after gaffe upon winning their primaries.

But inflation proved to be extremely difficult to tame, and despite five rate increases by the Fed and otherwise sound economic data (including a 50-year low in the unemployment rate), the 8%+ inflation rate continued unabated through the fall, and the narrative and momentum shifted back to the GOP in October and into November.  There were no “October Surprises” that truly changed race dynamics to speak of, and none in November (to this point) as well.


First, our final percentage chance that the Democrats hold onto the Senate and House.  These percentages are derived from our various BTRTN regression/simulation models, with a healthy overlay of judgment.



And here we go (deep breath):

Senate.  The Democrats will emerge, at some point in the days after Election Day, when all the counting is done, with 49 seats and the GOP will have 50.  We are predicting that Georgia will go to run-off, as in 2020, as neither candidate will crack the 50% mark required for election.  The Georgia run-off – with the fate of the Senate at stake, again echoing 2020 -- will be on December 6. 

House.  The GOP will take control of the House of Representatives, adding 21 seats to come away with a 234 to 201 margin.

Governors.  The Democrats will come away with a net pick-up of one governor seat, flipping two to offset one flip by the GOP, but the GOP will still hold a slight majority of state houses, 27 to 23. 

We must be quick to add that the Senate has a wide range of potential outcomes.  With the 8 battleground races all within three points, and most even closer than that, the true range of final Democratic seats is anywhere from 46 to 54. 

Don’t expect an early resolution to this; no Election Night finality, even apart from Georgia.  None of the slowest counting states – Michigan, Pennsylvania or Wisconsin – changed their laws to allow for early counting of mail-in ballots, largely because the Republican lawmakers objected to such a change.  Had they made this change, we would eliminate the “red mirage.”   We will thus be confronted once again with a nail-bitingly slow count in Pennsylvania, for sure, where election officials have already said it could take several days to count the votes – and by implication, given the closeness of the Fetterman/Oz race, declare a winner.  As in 2020, please don’t panic if Oz pulls out to an illusory lead.  And don’t be surprised if Trump starts shouting “fraud” again as Fetterman creeps closer to Oz’s totals.  


Control of the Senate has come down to eight clear battleground states, and all of them are exceedingly close.  

The race has only tightened of late as Republican candidates have chipped away at some large Democratic leads in New Hampshire, Arizona and Pennsylvania – but we think the Democrats in each will hold on to win, with Pennsylvania being a flip.  Three other races all feature GOP seats in which the Democrats have been knocking at the door but have never surged ahead at any time, in Ohio, North Carolina and Wisconsin.  The most recent polls seem to be moving redder again in those states, and we see the GOP holding on in each.

That leaves the two tightest races.  In the last few weeks in Nevada, Adam Laxalt appears to have nosed ahead of Democratic incumbent Catherine Cortez Masto, and we think that surge will be enough to push him over the top.

Which leaves Georgia, Ground Zero for 2020/21 election fraud claims in the presidential race; the scene of legal proceedings against President Trump for his role in it, including the caught-red-handed-on-tape phone call Trump made to Secretary of State Brad Raffensperger, demanding enough votes to overturn the state; and the twin Senate runoffs that both Democrats won on January 5, 2021 to get to a 50/50 Senate split and hand the Senate gavel to Chuck Schumer.

It is Georgia’s time in the spotlight yet again.  We predict another runoff in Georgia, as Libertarian Chase Oliver secures enough votes to prevent either incumbent Democratic Senator Rafael Warnock or GOP challenger Hershel Walker to get to 50% of the vote.  This will throw the race into a December 6 runoff with everything on the line.  We will make another prediction about that race on December 5.  (In 2020, we called each Georgia race going to run-off on Election Day, and then called that the Democrats would win both runoff elections – a high point for BTRTN predictions!)

Note well that while we are calling the 49/50 scenario as the lead prediction, there are a full range of potential outcomes here.  All eight battleground states could go one way or the other, suggesting outcomes anywhere from 46 to 54 Democratic seats; but 86% of our simulations fall within the 48 to 51 range.

Here are all the Senate races, generally sorted by how quickly they might be called.  Thus at the top of the chart you will see states where the polls close earlier on Tuesday evening, featuring races that will not be very close, and states that generally count ballots quicker.  At the bottom you will find those with the opposite conditions, and all the battleground states.




We’ll cover the Governors next, otherwise they will be buried down below the 435 House races.  There are a whopping 36 gubernatorial races this year, and we see five of them as truly competitive at this point. 

We see the Democrats rather easily flipping Maryland and Massachusetts, two blue states where moderate GOP governors, Larry Hogan (Maryland) and Charlie Baker (Massachusetts) have termed out, and will almost surely be replaced by the Democratic nominees, Wes Scott and Maura Healey, respectively.  These will not be competitive races.

We also see Democrat incumbent Steve Sisolak getting swept aside in Nevada (along with Cortez Matso in the Senate) by Joe Lombardo.  Other than that, every other seat will be kept by the incumbent party.

Here are all the Governor races, which are generally sorted the same way as the Senate.


The key data point here is the “generic ballot”, in which pollsters ask voters whether they would be more likely to vote for a Democratic or Republican representative, without naming any candidates (hence, “generic”).  While there are other variables in our BTRTN regression models that predicts how many seats will “flip” in the overall election, this is the most powerful one.  It has been quite accurate in predicting midterms outcomes going back to 2010.

The generic ballot at this point shows the Republicans are up by +1-2 percentage points over the Democrats.  Their “true” lead is even greater since, due to GOP overrepresentation in the House (due to small states and gerrymandering), the Democrats actually need to be at least +3 points (or more) ahead in the generic ballot to be competitive in the House.  They are a good 4-5 points away from that. Thus our models show the Democrats will likely lose 20 seats or more (the point estimate is -21), leaving the GOP with control of the House by a 234/201 margin. 

All the races are summarized below, with the 22 "toss up" races first, followed by the 37 "leans," the 31 "likelies" and then races the Democrats and then Republicans will surely win (the "solids").





  1. I wonder who is gonna win the Georgia runoff, leaning towards Warnock??

    1. Yes, for now, butwe'll be back with a final prediction on December 5.


Leave a comment