Wedgewood and the Rise of Fake Polling

State Sen. Mallory McMorrow (D-MI)

As a general rule, I try not to re-litigate Democratic primaries right before an election where I'm hoping the party is coming together to support the nominee, but there's a news story I feel I want to talk about on this blog, and it involves one specific primary (Michigan US Senate) it's hard to ignore at this point.  Until early Summer, the Michigan US Senate race was relatively even-keeled.  It had become apparent that something needed to break as polling showed that Dr. Abdul El-Sayed was going to take the lead from the Far Left if either Rep. Haley Stevens or State Sen. Mallory McMorrow didn't drop out.  But it wasn't clear which was the better candidate, and which made sense to drop out.  That was until a poll from Wedgewood Polling in late June that showed McMorrow with 9% support in the primary basically eviscerated her campaign.  McMorrow tweeted about this earlier this week, but the poll essentially caused severe donation drought, volunteers thought they were suddenly working for a lost cause campaign, and McMorrow was pushed by the press to drop out, which she eventually did.  But here's the thing: Wedgewood Polling never actually conducted that poll-they made it up.

Wedgewood Polling revealed in tweets that have since been deleted that they essentially made up the polling, for (as they claim) a way to prove that it was far too easy in modern political culture to essentially broadcast and make up a poll to influence a political race.  In that, they weren't wrong.  FiftyPlusOne and The New York Times both included Wedgewood polls in their aggregate reporting, and the polls were discussed in Newsweek, and on MS NOW & Decision Desk HQ.  These completely fabricated polls became an actual news item in the mainstream media, and were reported on in the same fashion as legitimate polling companies like Pew Research or Gallup.

It wasn't just in Michigan that this happened.  In Wisconsin, a Wedgewood poll in July showed State Rep. Francesca Hong significantly stronger than virtually any other polling outlet did, in many ways making it look like the Democrats had a new frontrunner, and hurting other Democrats in the process.  Here, if their goal was to help Hong, it actually did the opposite-the establishment was able to coalesce around a different candidate to beat her, which they did.  This was not the only fake poll in this year's Wisconsin gubernatorial election.  An outfit by the name of Median Strategies had also published a poll in August showing Hong up by 20-points, a lead that didn't remotely materialize in the results, when Hong ended up losing later that month by just under a point.  A third polling company (SoCal Strategies) has disappeared on social media in the wake of the Wedgewood scandal, and again-this is a polling outfit that messed in Democratic primaries, specifically the unusual 17-point poll lead in early August that showed Abdul El-Sayed destroying Haley Stevens, something that the results (Stevens lost by less-than-a-point) didn't pan out, but did result in an enormous amount of momentum for El-Sayed.  Again, like McMorrow, a fake poll hurt an establishment candidate.

The big blame here, ultimatelym is on the news organizations that run this.  Strapped newsrooms without as many researchers seem to have missed that these companies were not real, and were in fact lies designed to disrupt the electoral process.  But I will note that the news media is only partially to blame here.  Social media, where a lot of "in the know" political accounts have developed savvy followings, including from politicians & news reporters, is designed to not just tweet out about these polls without really checking their authenticity, it also is designed to blast it out and make pronouncements.  Twitter's search algorithm is a nightmare, but I can say with confidence that I knew about these polls (I remember seeing them), and I would not be surprised if I had tweeted about them myself, and certainly believed them to be legitimate.  The system is designed to allow bad actors like this to flourish.

Which is why I am tepidly happy that FCC Chair Brendan Carr is looking into this, but given the Trump administration's attacks on polling that does not favor the Republicans (see, for example, the lawsuit that Trump brought against pollster Ann Selzer & The Des Moines Register that was officially dismissed last month), I worry that they will be more willing to go after reputable pollsters like Selzer rather than just made up ones that are trying to shift an election.

Because ultimately that's what this feels like-trying to use polling mechanisms to hurt candidates.  It does not feel like a coincidence that all three of these pollsters were used to help bolster two Far Left candidates who had been seen to be harder sells in the general election (Francesca Hong & Abdul El-Sayed).  There is no evidence that either of their campaigns are linked to the pollsters, for the record, but also it's possible that supporters of them, or more likely, supporters of Republicans who might want easier-to-beat candidates, seemed to have as their aim getting them elected through perpetrating this deceit.  In El-Sayed's case, I don't totally buy that McMorrow was taken down solely by this poll, as there were other polls at the time that showed her losing ground to both El-Sayed and Stevens, but given the common phenomena of poll herding (where other pollsters sometimes withhold some results to match other pollsters and not be seen as outliers), having fake poll companies can do damage to campaigns.  It's hard not to look at the output from established, reputable polling house Susquehanna at the same time as Wedgewood's poll of Michigan that showed 49% of voters still undecided, and wonder if (without the Wedgewood poll) McMorrow would've had more room to grow with voters before Wedgewood framed this as a two-person race.  This is a deeply damaging moment in American politics given how much polling shapes news coverage of election races, and unless these companies are put on trial for perpetrating a fraud on the public, this will only get worse.