washington, dc

The Democratic Strategist

Political Strategy for a Permanent Democratic Majority

Are Midterm Polling Averages Skewed Toward GOP by Junk Polls?

From a pro-Democratic candidates point of view, there’s a lot of Debbie Downer poll analysis being bandied about this week. So let’s take a peek at the other end of the spectrum, and see what the Pollyannas have to say. In “‘Red Wave’ Narrative May Be Built On Crap Polling? Color Us Shocked,” ‘Doctor Zoom’ sees it this way at Wonkette:

In the final weeks of the 2022 midterm campaign, national polling averages appear to show a number of close races for the House and Senate tilting toward likely Republican wins. Very serious analysis pieces attempt to explain what’s going on in the national mood — maybe it’s Republicans deciding to stick with their party as the election gets nearer? People getting tired of hearing about abortion rights? Anger over declining gas prices, maybe?

Or perhaps the polling averages are being skewed by a lot of garbage data from GOP-friendly polling groups that have injected polling results that don’t have much to do with actual voter opinion. Democratic strategist Simon Rosenberg notes on Twitter that there appears to be a “ferocious” GOP effort to “flood the zone with their polls, game the averages, declare the election is tipping to them.” He says that while it’s entirely possible for Republicans to win in many of the elections next weeks, the polling and early turnout numbers so far suggest there’s not really any sudden shift to the GOP — especially not if there’s strong turnout by young voters.

Rosenberg warned Sunday that media organizations are being “played” if they uncritically report polling averages like those from FiveThirtyEight and RealClearPolitics, given the number of GOP-aligned polls being added in recent weeks in key states.

Given the stakes of the outcome of the midterm elections, who among us would doubt that the G.O.P. would do such a thing? Doesn’t mean they did. But flooding the zone with garbage polls to jack up averages in their favor is not all that big a moral stretch for the party that winks at the January 6 thuggery and also the plots to kidnap a governor and the U.S. Speaker of the House. Dr. Zoom asks further,

How much might the influx of GOP polling be skewing the polling averages? Rosenberg notes that there’s a “3.3 pt difference between the generic on Real Clear [Politics] and one without any partisan polling.”

It’s not just Rosenberg, either; Tom Bonier, CEO of Progressive data firm TargetSmart, pointed out Friday that an “avalanche” of GOP polling — some relying on an “older, whiter, more male” sample of voters than in the actual electorate — was making it look like Republicans were moving ahead.

Sure, you might expect progressives to articulate such a viewpoint. Doesn’t mean they are wrong, though. Dr. Zoom continues, quoting Rosenberg from a transcript of his recent interview by Joy Reid on MSNBC:

“What’s really unfortunate is that the places we rely on to help us tell us what’s going on in the election have been corrupted by a flood of Republican polling in the last few weeks. Now, in six major battleground states, more than half the polls conducted in October have been conducted by Republican firms. That means basically we can’t trust the data on Real Clear Politics or FiveThirtyEight any longer. It’s essentially Republican propaganda,” Rosenberg claimed. […]

“Listen, these are junk polls. The Republicans, this is part of the information war. They’re trying to suppress Democratic turnout, create more negative sentiment for Democrats and more positive sentiment for them. What I think is disappointing, many of the people who do the analysis on elections, should’ve caught this. This is an unprecedented massive campaign by the Republicans to game the polling average. And it’s disappointing to me this wasn’t caught earlier by many of the people that do this that are on TV and do this for a living. But it has to be understood now that the polling averages have been corrupted. We now need to look in my view towards the early voting,” Rosenberg said.

All of this would be based on Republicans trying to create a ‘bandwagon’ psychology to hustle the press and depress Democratic turnout. Stranger things have happened, and they are pretty clever about manipulating the media. There’s often a lot of volatility in polls in the closing weeks, but gamblers would be wise to pay more attention to non-biased poll averages in the closing 2 or 3 days of the midterm campaigns.

Bonier chucks some interesting stats into the mix in a couple of his Monday tweets:

Just 4 days ago, the Dem margin among those ballots returned so far was 2 pts wider than the same point in 2020. Now it’s 3.6 points wider. Meaning the returns since the debate have gone even more solidly Dem than they were before….But the most astounding element of the PA early vote? Voters under the age of 30 returning ballots thus far are +69.2% D. At this point in ’20 that same age group was +51.9D.

The last thing I would add is that, whether you believe the Pollyannas or the Debbie Downers, the margins in key senate races seem unusually close. Either way, it’s time for Democrats to pour it on to minimize losses and win close races.

3 comments on “Are Midterm Polling Averages Skewed Toward GOP by Junk Polls?

  1. Paul E Alderette on

    One Republican polling firm had J.D. Vance beating Tim Ryan in Ohio by ten points. If you believe that, I’ve got some beachfront property in Arizona that I would like to sell you 😉

  2. pjcamp on

    538 weights by poll quality so I call BS on that one.

    Certainly polls have a lot to answer for with regard to accuracy in recent years, but I believe I heard Republicans singing the skewed poll song back when Obama’s reelection wasn’t going their way.

    I wouldn’t be at all surprised to see the polls turn out wrong but not for this reason. You can’t flood the zone with shit when the zone deliberately excludes shit from its polling average.

    • Michael J Purnell on

      “538 weights by poll quality so I call BS on that one.”

      And yet Michael Moore was more on target in 2016 than 538.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.