Tuesday, September 6, 2022

BTRTN House and Senate Midterms Snapshot: How Far Have the Democrats Really Come in the Last Month?

Tom with BTRTN’s latest on the race for control of Congress in the November 2022 midterm elections.

The headlines these days are breathless, indeed.  “Dems Flip 2022 On Its Head”…"Election Forecasters Rethink Their Ratings”…”House Up For Grabs.”  Between the SCOTUS Dobbs decision overturning Roe, Biden’s incredible, no-one-saw-it-coming hot streak, tumbling gas prices and Trump’s DocuGate fiasco, suddenly Democrats are downright jolly about November.  Why, look at the turnout for the Kansas abortion referendum, and Pat Ryan’s stunning win in the bellweather New York 19th district special election! With all that, surely, the electoral landscape has been upended, and the Democrats are rocking toward keeping their White House, Senate and House trifecta in 2022.  Right?

Well, not really.  Certainly not yet.  The real headline is this:  despite all of that Blue good news of late, the Senate and House race dynamics have only modestly changed in the last month.  

·        The Democrats are still the odds-on favorites to maintain the Senate, largely due to the ineptness of the GOP’s swing state candidates.  And those odds have indeed improved, and in this snapshot we have flipped another GOP seat to Blue.

·        But the Dems are long shots to hold onto the House, despite all that good news.  That, actually, does represent progress, because a month ago they had virtually no chance at all -- not enough to even qualify as a long shot.  But there remains much to be done -- many numbers to improve -- for the Dems to truly have a chance to hold onto the House.

Our proprietary BTRTN models, which calculate the odds of the Democrats maintaining control of the Senate and the House, show some positive movement on both, but hardly a sea change.

These odds could still change, of course.  Remember, these assessments are “snapshots,” not “predictions,” based on today’s conditions and data. We’ll get into the prediction business on November 7, the day before Election Day.  From here on in, for the next nine weeks, the Democrats have to translate all that good news into a coherent message and then sell the heck out of that message this fall. 

The best way to think about the recent set of positive events is, perhaps, as the beginning of a Biden/Dem comeback.  Only time will tell if they can complete the job.  But the numbers, those in the chart above and those you will read about below, have to take a sharp -- and far sharper than to date --upward trajectory, fast, for the Democrats to contemplate a double win in Congress come November.


For those of you unfamiliar with BTRTN, here is our track record on the House midterms and all Senate races over the past decade-plus.  We think (rather immodestly) that our record holds up well against all other forecasters.


To illustrate how little has changed, let us bring back a chart we published last week, which uses a variety of political and economic measures to quantify the electoral environment.  The data establishes that on measure after measure, Biden and the Democrats did make progress in August, but that progress, however welcome, has been modest, more in terms of stopping the bleeding than a sharp rebound.  For instance, Biden’s approval rating at 41% (net -14) is back to only that of June, short of January 2022 (44%, -7), and well below the 55% (+21) he enjoyed at the outset of his presidency.  The ratings of how Biden is handling key issues, and various economic measures, show varying degrees of “recovery.”  But there is only one really good number in this chart, and that is the unemployment rate.  Every other measure – from the approval rating to the price of gas – may be moving in the right direction, but there is still a long way to go.

So, with that stage set, let’s take a deeper look at where the House and Senate midterms stand – right now.


The most important number on the chart above, in terms of the House, is the generic ballot.  As we have said many times, it has been a remarkably accurate predictor of midterm outcomes.

You can see that (again, following other measures), the Democrats have made progress in the generic ballot in the last few months, pulling even with the GOP at 44% to 44%, meaningfully better than the 45% to 42% advantage the GOP held in June.  The Democrats are back, roughly, to where they were in January.

But getting back to January levels, back to being “even”, does not mean the Democrats suddenly have a 50/50 chance of keeping the House.  Due to small state population bias and aggressive gerrymandering, the House is disproportionately represented by Republicans, so a tie in the generic ballot is simply not good enough.  To get within a truly competitive range of holding on to their House majority, the Democrats need to have at least a five-point advantage in the generic ballot.  Biden and the Dems have to get back to September, 2021 levels, when they led the GOP by four points in the generic ballot – or better.  That is not impossible – they were there just a year ago -- but given the size of the gap and the timeline to the election, it is a tall order. 

The reason our track record in past midterms is quite good is because of the strength of our BTRTN model.  The model takes into account the very powerful generic ballot, as well as which presidential term it is (it matters, first-termers always do worse), and how many seats the president’s party holds, which matters as well.  There are “only” 222 Democratic seats right now (221 actual members and one vacancy, Florida’s 20th, which Democrat Charlie Crist held before he resigned a few days ago).   That low number actually serves to limit the number of seats the Democrats can lose; when Barack Obama took a beating in 2010, the Democrats held 256 seats.  Ditto Bill Clinton in 1994, when the Democrats held 258 and lost 54.  As bad as it looked for Biden in early 2022, with such weak approval rating and gas prices soaring, there was simply no way the Dems was ever going to lose the 70 seats that some giddy Republicans once thought they might.

All the rating services (538, Cook, Politico, Sabato, BTRTN, etc.) agree that only a fraction of the 435 House races actually stand a chance of being competitive.  Most of these races are over once the primary is finished, as they are so heavily weighted, whether naturally or via gerrymander, to one party or another.  At most, about 100 districts are actually “in play,” that is they could be won by either party, but even that number includes about 40 that are very likely to go one way or the other, leaving only about 60 that are truly competitive.  Those 60 will decide the fate of the House.  To keep the House, the Democrats will have to win three-quarters of them, in a year of strong political headwinds and the precedent of midterm nightmares for first-term presidents.

Our view back in June was that the Democrats had virtually no chance – 1% -- of retaining the House.  The year seemed to be lining up as yet another solid repudiation of a first-term president in his first midterms.  Our current view, based on the modest improvement in the generic ballot, is that the odds are now roughly one in seven, or 14%, that the Democrats retain the House -- a long shot, for sure, but something to build on down the home stretch.  The races array as follows.


The Senate, as we and others have been saying all along, is a very different story.  Senate elections are far more driven by the quality of the candidates themselves (though they are not quite immune to the external fortunes that influence all politicians).  The Democrats’ odds of holding the Senate has grown from about 50/50 in our first in-depth look in June, to nearly 6o/40 in July, to just about 75/25 now.  The cause is clear and, at this point, well-established:  the GOP may blow what should have been an excellent chance of taking control of the Senate for the simple reason that they have nominated, largely at the behest of Donald Trump, a slew of simply awful candidates in the key races, many of whom are trying to replace respected GOP incumbents who have retired.

The GOP candidates in the nine battleground states are a motley crew indeed, consisting of the wildly underperforming TV doctor Mehmet Oz in Pennsylvania; the downright disturbing former Heisman Trophy winner Hershel Walker in Georgia; the Trump-savaging-turned-Trump-groveling (and thus Trump-backed) bestselling author J.D. Vance in Ohio; the surprisingly Trumpian (and thus unpopular) incumbent Senator in purple Wisconsin, Ron Johnson; a trio of uninspiring and occasionally gaffe-prone Trump-backed newcomers, Blake Masters of Arizona, Adam Laxalt of Nevada, and Ted Budd of North Carolina; the winner of a way-too-late, yet-to-happen primary (slated for September 13) in New Hampshire; and the incumbent Marco Rubio of Florida, the only one of the nine in a reasonably commanding position. 

The Democrats have countered with an impressive array of candidates, mostly centrists with deep political experience, who are pulling in money at astonishing rates.  Their solid personal record and campaign performance has doubtless received at least some presumed modest positive impact from the Biden surge, Biden’s suddenly tough-talking anti-MAGA posture, the post-Dobbs ire, Document-Gate and falling gas prices.   As we see it, the Democrats have improved their chances of holding the Senate from 58% in July to 74% today.  That is, if the elections were held today, the Democrats would likely keep the Senate. 

We have made the following five rating changes since July, all but one Senate race moving in a positive direction for the Democrats.

That leaves the overall scorecard showing the Democrats, if the election was held today, coming away with a 52/48 hold on the Senate.  You might think this may not mean much if the Democrats lose the House, but it will mean a ton in terms of federal judgeships, including SCOTUS nominees, which require only 50 votes in the Senate (plus VP Kamala Harris if needed).  And it will prevent Mitch McConnell from launching insane investigations into who knows what (the House, under GOP control, will do their share of that).

The overall scorecard now looks like this.  There are  nine “battleground” states and the Democrats are ahead in six of them, including two that are currently held by Republican incumbents (Pennsylvania and Wisconsin).


And the chart below lists all the races, with a focus on the nine battleground between the two purple lines.



Some readers look to articles like this to decide where to focus their personal energy in the election, including donations and/or volunteering (making phone calls, writing postcards, canvassing, etc.).  To assist in those “resource allocation” decisions, we have listed them in order of our sense of priority to help make those choices.  Remember, the Democrats have to win four of these nine races to hold the Senate.

Georgia.  This is a very close race, and a “must hold” seat for the Democrats.  Raphael Warnock is facing former Georgia Bulldog football star Herschel Walker.  One of these people is a serious politician.  Walker, like Oz, is a Trump-backed celebrity with no political experience.  He also has a history of mental illness, claiming multiple personalities, one of which happened to abuse his wife.  And yet, this being America, this is a close election.  The polling has been very even, first favoring Warnock by a nose, and now Walker by the same slim margin.  Thus we lean to the incumbent and maintain our BTRTN Rating as Toss Up Democratic.

Pennsylvania.  This is the Democrats’ best chance of flipping a state, and nothing should be spared in that effort.  Lieutenant Governor John Fetterman, who is a good fit with mainstream Pennsylvanians, won the Democratic nomination just days after suffering a stroke.  But he is back on the campaign trail and giving the inept and underfunded Dr. Oz quite a thrashing thus far.  We have changed our BTRTN Rating from Lean Democratic to Likely Democratic, and this would be a flip for the Democrats. 

Nevada.  Democratic incumbent Catherine Cortez Masto is locked in a tight race with former state Attorney General Adam Laxalt, an erstwhile mainstream Republican (he’s the grandson of former Governor, and Reagan pal, Paul Laxalt who is now a full-on Trumpster, having led Trump’s Nevada efforts to overturn the 2020 election outcome}.  This is a close race with Cortez-Matso slightly ahead in the  polling.  We continue to maintain our BTRTN rating as Toss Up Democratic.

Wisconsin.  This has emerged as another flip opportunity for the Dems.  GOP incumbent Ron Johnson is an outright Trumpster, a trafficker in conspiracy theories, a vaccine skeptic, and is notoriously dismissive of January 6 critiques (“largely a peaceful protest”).  These are unusually hard line positions to take in a purple state that went for Biden. His approval rating has suffered accordingly, down to 36% (as of April).  Lieutenant governor Mandela Barnes will face him, and while we are not indifferent to the difficulties in unseating an incumbent, it is hard to ignore the fact that Barnes has been ahead of Johnson in every public poll since her primary win.  Thus we are changing our BTRTN Rating from Toss Up Republican to Toss Up Democratic and a second Democratic flip.

Arizona.  Democratic incumbent Mark Kelly, has turned from Space Shuttle astronaut to fundraising juggernaut, and has been leading GOP nominee Blake Masters, a Trump-backed tech entrepreneur , by near double-digit leads in the polls.  Masters fell in line with the GOP gaffe-brigade with this doozy:   when asked about gun violence, his tone-deaf assessment was as follows: “It’s gangs.  It’s people in Chicago, St. Louis, shooting each other very often, you know, Black people frankly.”   We have changed our BTRTN Rating from Lean Democratic to Likely Democratic.

New Hampshire.  The New Hampshire primary is the last in the nation, not until September 13, and that will not give the GOP much time to consolidate around their nominee to challenge Democratic incumbent Maggie Hassan.  The likely winner, former Army Special Forces outsider Don Bolduc, has trailed Hassen in the polls by mid-single digits in this purple state.  Hassan maintains a gigantic funding advantage.  We maintain our BTRTN Rating at Lean Democratic.

North Carolina.  The GOP primary to determine the successor to the retiring Richard Burr was won by U.S. Representative Ted Budd, with Trump’s endorsement.  Former state Supreme Court justice Cheri Beasley is atop the Democratic ticket, and so far the polling has been extremely close, with Budd slightly ahead in the early going, but Beasley inching ahead of late.  That is not enough, as yet, for us to change our rating, even with Beasley out-fundraising Budd by a 3:1 margin.  So we are maintaining our BTRTN Rating as Toss Up Republican. 

Ohio.  Hillbilly Elegy author and Trumpster J.D. Vance seems to be finding his footing in this tight race, drawing even in the polls of late in his bid to replace retiring GOP Senator, Rob Portman.  The Democrats have another strong candidate, U.S. Representative and former presidential candidate Tim Ryan.  Ryan has been running an excellent race, but Ohio has quickly evolved from a purple to red state.  Ryan is crushing Vance in fundraising, though, and running a center-right campaign, so this will be one to watch.  For now, we maintain our BTRTN Rating as Toss Up Republican.

Florida.  The Democrats are running a terrific candidate in U.S. Representative Val Demings, the one-time police chief of Orlando and more recently a House impeachment manager in the first Senate trial of Donald Trump.  But Florida has been a disappointment for the Democrats in many a high profile race in recent years.  There has been surprisingly little polling, and it is mixed.  But unlike in other battleground states, Rubio and Demings are about even in fundraising.  We are maintaining our BTRTN Rating at Likely Republican.

Missouri.  The Democrats’ hopes for stealing a clear red state seat lay in the comeback attempt of the disgraced former Governor Eric Greitens.  Mainstream GOP attorney John Wood had entered the race as an independent to give Missouri Republicans a respectable choice, despite the clear potential to split the GOP ticket and give the Democrats and opening.  But Republican Attorney General Eric Schmitt won the GOP primary, Wood dropped out, and Schmitt has a safe, double digit lead over the Democratic nominee, Trudy Busch Valentine (an heiress of the Busch family).  We have thus changed our BTRTN Rating from Likely Republican to Solid Republican, dropping Missouri from battleground state status.

Stay tuned.



  1. I think your house ratings are way too low. Been many reports of Republicans being short on money already, and Biden's approval rating is irrelevant because of the effect of disapproval from the left. It's not like those people are going to run out and vote for a Trumper election denier.

    I think your Senate forecast makes much more sense and is very good.

    The biggest ? IMO is whether the election better reflects registered voter polls or likely voter polls. Based on what it looks like so far in special elections and such, I think it will be closer to registered voter (and some states you can register on election day).

    I just hope people in states where Republicans don't want people to vote that don't vote every election, or even that don't look at every piece of junk mail they get, that people are checking their registrations and making sure they aren't surprised to learn that they've been purged on election day.

    We will not have free and fair elections in red states. We need to win anyway.

  2. I continue to be dubious of "Likely Voter" screens.

    I took all of the Real Clear Politics "Generic Ballot" polls that have been in the field since the Dobbs decision (a broader time frame than 2 weeks being used by the site)

    Looking at all of them, here's what I found

    screen Democrats Republicans Total
    Likely ............3.0 .............89.0 .................92.0
    Registered.........99.0 .............20.0 ...............119.0
    (empty) ............0.1 ...............3.0 ...................3.1
    Total ................102.1 ...........112.0 ...............214.1

    In other words, it largely takes the "special sauce" of a polling firm's definition of "Likely Voter" to keep the overall margin close.
    The "Registered Voters" surveys show 9 giving Republicans the edge, averaging 2.4%. There are 24 giving Democrats the edge, averaging 4.1%/ There are 5 that "tie" the parties. Putting them all together, Democrats have a 2.1% edge.

    If I look at the Registered Voter polls in the time frame that RCP is using -- there are 11, and the Democrats have a 3.1% bulge.

    Short version -- it is going to be a turn-out election. The closer the voters match the surveys' "registration" claims, the more likely Democrats are to win.

  3. it would be interesting to see your list of "toss up" House districts.

    1. Hey John, I included them on the most recent article, posted today (September 30). Best to you, Tom.


Leave a comment