I keep thinking I should write about this, and the midterms are almost upon us, so I guess it’s now or never.
Fake polls! Or just bad, right leaning polls with evil crosstabs and poor samples. Everyone hates them. When Trafalger or Rasmussen or whoever comes out with another right leaning poll, twitter Democratic partisans descend like a pack of feral analysts, hacking about them with slide rules and irony. Also sarcasm. And sometimes rage.
Amongst their weaponry!
Where was I?
Right! Hacky right polls intended to show that the Republicans are ahead. Everyone hates them.
And those polls are irritating! There is in fact an ecosystem of right partisan pollsters who may or may not follow some sort of rigorous methodology before spewing out an infinite slush of good numbers for Republicans. (You can read about the debate here.)
Pollsters and analysts are skeptical of these polls and worry they’re bad science. But people on social media are often worried that the polls are part of a plot to tip the election towards Rs. After all, if you’re a Republican pollster, you want Republicans to win. So why would you release good poll numbers unless it would help Republicans?
The argument is generally that good numbers for Republicans create a positive narrative for Republicans, which inspires partisans and encourages donations. Thus, inflating polling averages is part of a campaign to win the election for the GOP.
I don’t think this is very convincing though. After all, the last few elections, polls have overestimated D chances, and that hasn’t helped Ds. On the contrary, in 2016, the good poll numbers for Ds led people to believe Trump couldn’t win, which demobilized Dem voters, and led to the absolute worst. Similarly, in 2020, D overperformance in polls just left D partisans disheartened on election day.
You could argue that R pollsters are trying to boost donations by encouraging R voters to donate in races they might otherwise ignore. But this isn’t a good rationale either. You usually want accurate polls precisely so that people can make the best decisions about where to send their money. Encouraging Rs to send a bunch of donations to the Colorado Senate race with a series of R leaning polls, for example, is the opposite of helpful if that race isn’t actually close. Deceiving your own voters and donors makes it harder to implement good strategy, not easier.
So, if crappy R-leaning polls don’t help Rs win, why do R-leaning pollsters put them out?
The obvious answer is that they put them out because they help the pollsters, even if they don’t help candidates.
R partisans love to see good numbers for Rs. Pollsters who put out good numbers for Rs get a lot of attention in the right wing media ecosystem. That can translate into name recognition, and eventually into donations and money, through small or larger donors.
In short, the interests of R pollsters and the GOP as a party don’t necessarily align—just as the interests of the GOP as a party doesn’t always align with the right wing mediasphere. (Right wing media loves Trump; Trump obsession doesn’t necessarily boost the chances of other R candidates or of the party as a whole.)
Trafalgar running crap polls is irritating, and a potential problem for poll aggregators who want to try to make their forecasts as accurate as possible. But Trafalgar isn’t pushing the country rightward with polls like that. If anything, if their polls are inaccurate, they may be hurting Rs at the margins by focusing attention on the wrong races.
The GOP remains a threat because partisanship is strong and they have a lot of structural advantages. But not everything that Rs do is genius or an effective strategy. In fact, the party is a huge mess, and they regularly make crap decisions—like nominating Mehmet Oz or Herschel Walker to run for races that should be easy wins for a competent Republican candidate. Rs may do well in the midterms. But if they do, it won’t be because a bunch of R pollsters massaged the numbers. It might well be in spite of that.