Did Republican Pollsters Shift the Midterms?

Sen. Maggie Hassan (D-NH)

It has become a continual refrain in recent years that "polling sucks."  This maxim has been around for years (polling is, by its very nature, an imperfect science), but became a rallying cry after polls did not predict the surprise victories of Donald Trump in 2016 in Wisconsin, Pennsylvania, & Michigan.  In 2020, people doubled down on this refrain as Republicans who had been in "tight" Senate races in South Carolina, Maine, & Montana won gargantuan victories, hundreds of millions of small-dollar donors' cash wasted on races that went nowhere.

At first glance coming out of the 2022 midterms, you'd be forgiven for assuming that this was the case.  Going into election night, if you asked most pundits whom they thought would win, they'd have said the Republican Party, and of course while they were right in some cases, the size of the Republican victory was nowhere near what they assumed.  The GOP barely won the House, and very possibly had a net loss in the Senate (we'll know more on Tuesday when Georgia finishes its vote).  But looking at polling, this wasn't really the case.  Assessing averages from 538, most of the polling averages ended up being spot-on, if maybe slightly overestimating the Republican position.  North Carolina Senate was very close to what 538 was predicting, as was Georgia Governor.  Ohio Senate was almost exactly right, and Wisconsin Senate wasn't far off.  While there were races that were clearly wrong (Pennsylvania Senate underestimated John Fetterman's position, and Democrats in Colorado & New Hampshire way outdid what was expected of them), the polls were right...

...at least the nonpartisan polls were.  It's worth noting that a number of Republican pollsters, specifically Trafalgar, were dramatically off this past year.  Trafalgar made a name for itself in 2020 for being one of the few pollsters to understand that Donald Trump was doing better in key swing states than a lot of the national pollsters were giving him credit for.  This has given the organization, whose methodology is suspect, a sheen of respectability which it clearly did not earn.  Maggie Hassan won the New Hampshire Senate race by 9-points-Trafalgar had her down by one point the week before the election.  Michael Bennet won in Colorado by 14-points...Trafalgar had him up by just 2-points.  Patty Murray won her election in Washington by 15-points...Trafalgar had her up by just one-point headed into the election.  Those are ridiculously inaccurate polls, even in a "the polls are wrong" universe.

Sen. Patty Murray (D-WA)

Trafalgar's name and reputation after this election is basically mud, and I doubt they retain the "A-" rating they got from 538...honestly, I wouldn't be surprised if they aren't counted at all going forward.  But it's worth asking the question of whether or not a glut of Republican polling, which came out repeatedly throughout October and early November, was intended to try to create a wave.  There's something to this.  Democrats are wont to say "run like you're ten points behind" which is about the worst advice you can give...no one wants to vote for a loser, and so if you're trying to juice races like those of Hassan, Bennet, & Murray, you need to show polling that the race is actually competitive.  Trafalgar, InsiderAdvantage, & Moore Information Group, all conservative polling outfits, put out polls in the waning days of the election showing Republicans in close positions in these three states, even though that wasn't reality, and wasn't borne out by their nonpartisan competitors, particularly in Washington.

Bad polling is one thing, but I honestly do think there's some credence to the accusations that they were trying to create a wave where there wasn't one, and trying to distract from races that were obviously winnable.  It's hard to gage this without getting inside the strategies of these campaigns, and distraction is a legitimate campaign tactic.  But in the final days before the campaign, no one was acting like the Republicans would be winning Wisconsin & North Carolina by smaller margins than the Democrats were going to have in Arizona & Pennsylvania...even though that's what happened.  In fact, you would've been laughed off of the internet if you'd have said that.  Saying that control of the House was going to come down to a couple of seats that were "foregone conclusions" for the Republicans that instead were won by less than a point like IA-3 or CO-3 would've been absurd.

It's hard to tell if this impacted how campaigns spent money.  It's clear based on leaked reporting that Democrats did not have confidence in their performance that night (the Fetterman campaign, for example, had internal polling showing them down the week before the election, and reports from a number of House Democratic campaigns claim they were surprised by their victories).  It's also not entirely certain that the Republicans had too much confidence; Arizona's Governor's race may well have been different if Republicans had taken Katie Hobbs seriously as the threat she ended up being.  But I have to wonder if we'd be looking at an even better night for Democrats had we known how close this would've been.  Would the Democrats have been able to hold the House if they had had accurate enough reporting to know the House majority just required winning a handful of seats in Biden districts on Long Island?  Would Democrats be in a much better position in 2024 for the Senate if they'd gone full throttle for Cheri Beasley & Mandela Barnes, potentially landing a surprise victory on their unexpected strengths?  These are questions you ask after every election (there's always close races), but given the importance of Republican pollsters in shaping the idea of a red wave, you have to wonder if Democrats & the pundit industry fell for a trick that cost the Democrats key races.