Faulty reasoning signals

• Biased Writer Voice - 24.3%
• Appeal to Authority - 18.8%
• Negativity Bias - 18.4%

Article text

Graham Platner’s controversial social media history was first reported on October 16 , and he revealed the existence of his tattoo with Nazi origins on October 20. In the following days, two polls sampling Maine Democratic Senate primary voters were released. They told completely different stories. The University of New Hampshire poll, taken largely before the tattoo news broke, had the rookie candidate Platner leading the incumbent Governor Janet Mills by an astounding 34 points. Yet SoCal Strategies, in a poll taken just after the tattoo news, had Mills up by only 5 points. The tattoo scandal, we know now from yesterday’s Wall Street Journal , caused “an immediate drop in fundraising … with donations falling to less than half of what the campaign had been averaging daily.” What helped steady the campaign ship was the next poll, conducted by the Maine People’s Resource Center [MPRC] from October 26 to 29 and released on November 12. Not only did the survey show Platner edging Mills in the primary by 2 points, but it was also the first poll to project that Platner would defeat the incumbent Republican, Senator Susan Collins. And it had Mills losing to Collins by the same margin Platner was winning: 4 points. The presumption of who was the most electable option was being flipped. But the MPRC was not a disinterested operation. Come February, the progressive activist group would formally endorse Platner. In the month following the MPRC poll, three polls tracking the Maine Senate race were released. Pan Atlantic Research had Mills leading Platner by 10 in the primary , but indicated no statistical difference in how they stacked up against Collins (Platner up 1, Mills tied). Two others gave Platner huge primary leads of 15 and 20 points. One was from the Progressive Change Campaign Committee , which had just endorsed Platner. The other was from the Platner campaign’s own pollster , which also conveniently found that only Platner, and not Mills, was leading Collins. No other primary polls were released until February. No other major controversies blew up in that period either. Thanks to the Platner polling network, Maine primary voters could reasonably conclude that Platner had demonstrated the political skill to weather the controversy and was well positioned to win in November. Once a handful of independent polls went into the field in February and March, Platner was holding primary leads ranging from 5 to 38 points. A feeble, half-million-dollar ad buy from Mills—spotlighting Platner’s social media comments that blamed women who drink too much for getting raped—didn’t move the dial, and soon after, she suspended her campaign. Platner’s polling network didn’t manufacture Platner’s support out of thin air. Clearly, the Democratic electorate in Maine was predisposed towards a charismatic populist outsider and not a septuagenarian political veteran. But pumping out favorable polls at a sensitive phase of the campaign—especially in the absence of independent polls produced by well-read media outlets—warped perceptions about Platner’s political durability and made it harder for skeptics to warn about the possibility of bigger scandalous shoes dropping. That Platner’s campaign dominated the polling space is no scandal. Campaigns and their allies have long produced favorable polling designed to shape media narratives. But since high-quality polling has become more expensive, as it has become harder to get potential respondents to pick up the phone, fewer media outlets in an age of declining local news are producing their own surveys. Media executives may also be increasingly fearful of a big polling whiff that sullies their brand. That leaves more space for campaigns as well as fly-by-night outfits with little to no track record. Ginning up favorable polls is something any campaign or sympathetic group can, in theory, do, regardless of party or ideology. In Iowa’s recent Senate Democratic primary, not a single independent poll was sampled . The Hawkeye State was once home to the celebrated Des Moines Register poll by Ann Selzer, but she quit the business after she wrongly projected that Kamala Harris would beat Donald Trump in Iowa. And the paper didn’t get back into the polling game for this latest campaign. In the Register’s absence, all we got were polls tied to the campaigns. Two polls showed big leads for state Senator Zach Wahls, one from his campaign and one from Teamsters Local 238, which endorsed him . And three polls showed big leads for state Representative Josh Turek, one from his campaign and two from the Vote Vets Action Fund, which endorsed him . Only Turek’s polls proved accurate, as he won by 25 points. As of late, the polling push seems to be more heartily embraced by left-wing insurgents, understandably so. Since such candidates tend to face initial concerns about electability, injecting favorable polls into the discourse can help mitigate those concerns. The charitable case for it is that it helps draw attention to outsiders with great potential who might be unfairly overlooked. The downside risk is that it creates an unwarranted perception of strength. We are seeing the polling tactic used, effectively so far, in Michigan’s Senate Democratic primary by former public health official Dr. Abdul El-Sayed. El-Sayed, while he doesn’t embrace the socialist label, is nonetheless the great socialist hope in the 2026 midterms. Explicitly backed by the two most prominent federal socialist lawmakers—Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez—and tacitly backed by the Democratic Socialists of America , a victorious El-Sayed could prove that a candidate who supports Medicare for All, challenges the concept of Israel as a Jewish state, and once trafficked in “Defund the Police” rhetoric can win a statewide race in a perennial battleground. As we have no examples of a candidate with El-Sayed’s issue set winning a general election in Michigan before, his nomination would carry a degree of inherent risk. However, it’s certainly possible that the current electorate has moved ideologically in his direction. Before the hypothesis can be tested, he must win the primary. In the past month, media coverage has increasingly referred to El-Sayed as the primary’s “frontrunner.” But on what data is that title based? Five primary polls sampled in April through mid-May showed wildly varying results with no clear frontrunner. An average of the five polls put El-Sayed ahead, but only by less than 2 points over U.S. Representative Haley Stevens, a moderate who is vocally pro-Israel (albeit critical of Israeli Prime Minister Benjamin Netanyahu ) and, according to Bridge Michigan, “supports creating a [health insurance] public option but has not introduced legislation in Congress to do so.” However, Stevens was the leader in three of those five polls. Also of note, in four of the polls, the undecided share of the electorate was at least one-third, indicating the collective mindset of voters was far from settled. Also, during this April to May period, we have three polls with general-election trial heats against the presumptive Republican nominee, former U.S. Representative Mike Rogers. In two of these, El-Sayed fared the worst among the Democrats, and in the one where he fared the best, he was still 1 point down. This fed a narrative that El-Sayed had the biggest electability problem of the Democratic field. Then, on June 1, El-Sayed’s campaign released its own internal poll showing a 3-point lead over Stevens, with State Senator Mallory McMorrow coming in a weak third. Four days later, Tulchin Research, a polling firm hired by the newly formed pro-El-Sayed super PAC Fighting for Michigan, not only found El-Sayed leading a second-place Stevens by a whopping 18 points, but also—unlike most other polls up to that point—he would beat Rogers in November by five points (no data was provided about McMorrow and Stevens.) Then in mid-June, the Common Defense political action committee —which has endorsed El-Sayed —published general election trial heats with the 41-year-old doctor winning by three, slightly better than his Democratic rivals. Another internal poll, conducted in late June by Fighting for Michigan, showed a 19-point primary lead. These polls from the El-Sayed network were buttressed in June by three other independent polls with directionally similar results—two of which showed only El-Sayed beating Rogers. This run of good El-Sayed polls largely came at McMorrow’s expense, with her weak showing prompting her to drop out of the race earlier this week. But two of these pollsters, Mitchell Research & Communications and Quantas Insights, have middling grades in the Silver Bulletin Pollster Ratings , and the third, Wedgewood Polls, just entered the polling scene last month. Its first poll of the Louisiana Republican primary runoff was wildly wrong. Meanwhile, only one June poll came from a “ Select Pollster ” as deemed by The New York Times —meaning the pollster met two of three basic criteria: a track record of accuracy, membership in a professional polling organization, and use of probability-based sampling. That poll, from Susquehanna Polling & Research, found El-Sayed had a scant 2-point lead over Stevens , well within the margin of error. It also found that 49 percent of the electorate was undecided, more than double El-Sayed’s 22 percent. (Only one other pollster working in Michigan has met the Times’ Select Pollster standard, and that’s Emerson College, which hasn’t been active in the state since April.) El-Sayed’s polling streak took a slight hit this week when a Tavern Research poll tested a two-way primary matchup and a three-way matchup including McMorrow (who is still on the ballot). El-Sayed won by three when McMorrow was included, but lost to Stevens by one in the head-to-head. Tavern Research is a Democratic political consultant shop that hasn’t been polling long enough to be rated by Silver Bulletin. What’s missing from this picture is credible polling from respected media outlets. It may well be that El-Sayed is more electable than Stevens in November—either because he is more polished on the stump or because the Michigan electorate has become a good match for his economic populism and determination to cut off aid to Israel. But we lack good data on which to make such an assessment. In theory, since all campaigns are free to pay for their own polls, competing attempts at poll-based spin could end up being a wash—which is what we saw in Iowa. One might presume that a campaign that isn’t releasing its own polls is losing. But another explanation could be that a campaign doesn’t think that’s a good use of limited campaign finance resources. Stevens released one internal back in February with her up two. Then McMorrow did one in March with her up five . Neither did much to further their preferred narratives, and for whatever reason, they didn’t keep going back to that well. For a good long while, many people groused about poll-driven “horse race” coverage from mainstream media outlets that ignored policy substance and left news consumers poorly informed. And as polling is always an inexact science, media outlets took it on the chin whenever a poll they sponsored was off the mark. But the decline of professional polling from media outlets leaves us more vulnerable to spin polls from campaigns. That vulnerability was cannily seized by the Platner campaign, to the detriment of Maine Democrats who perhaps wanted to believe a little too much what the numbers seemed to say. While we may all tend to cheer the polls with results we like and dismiss those we don’t, the decline in independent surveys by pollsters with extensive track records should make us extremely cautious about presuming too much about the future based on the data we have today. The post The Platner Collapse Is a Cautionary Tale About Polling appeared first on Washington Monthly .