10
Loading video...
By Daniel Orton and Dan Gooding
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
Nate Silver's midterm elections forecast shifted slightly toward Republicans Wednesday, after he removed a series of fake polls that he said had been roughly 1 point more favorable to Democrats than the broader consensus.
The famed polling analyst then took a deeper dive into what had transpired with Wedgewood Polling in his newsletter Thursday, writing that vetting was thorough across the industry, and that there were real risks from fake polls looking ahead to future elections.
The biggest change Wednesday came in Michigan, where Democratic U.S. Senate nominee Abdul El-Sayed's lead fell from 3.6 points to 3. Silver noted that the movement also reflected the addition of a Marist poll to the model that showed El-Sayed 2 points ahead of Republican former Representative Mike Rogers.
Read More on News
"Overall, our forecast is little changed, though that's partly because some strong generic ballot polls entered this AM for Democrats," Silver posted on X on Wednesday morning.
The fake polls came from Wedgewood Polling, which revealed on Tuesday that it had fabricated most of its polls. Newsweek has corrected and retracted stories that included data from Wedgewood.
"I don't know what the value of the 'social experiment' is in the two recent examples of faked data operating under the guise of polls," Michael Traugott, a political science professor with the Institute for Social Research at the University of Michigan, told Newsweek.
"There is no indication in either the Wedgewood Polls or Median Strategies case that the intent was to embarrass the polling industry or to support a claim of fake polls by specific candidates or political parties," he said. "The principals involved in these 'experiments' disappear upon disclosure, and we don't learn from them what they intended."
...
Silver Bulletin Adjustments: What Has Changed?
Silver said Wedgewood had been permanently banned from his aggregations moving forward, noting that the removal of the polls, which tended to be roughly 1 point more favorable to Democrats, had altered numbers in favor of Republicans in some cases.
The most notable change was the Michigan Senate race. Some other races saw a slight dip compared to Tuesday, including Republican Senator Susan Collins against Democrat Troy Jackson in Maine. Both saw their support dip slightly by 0.1 and 0.2 percent.
The overall generic ballot average was largely unchanged, with Democrats falling from 51.7 percent Tuesday to 51.5 percent Wednesday, and Republicans going from 42 percent to 41.9 percent.
Overall, between 15 and 17 polls fabricated by Wedgewood had been cited by various sites before being removed Wednesday.
"Previously, Wedgewood responded to inquiries from multiple polling aggregators and answered questions about their identities, methods, and funding, which is why their polls were subsequently aggregated," Silver wrote on his site Wednesday.
Related Story
Newsweek Webinar: How the Politics of AI Could Shape Midterms
Wedgewood Polls: What to Know
On Tuesday, Wedgewood released a statement acknowledging that the vast majority of its polls were not based on genuine survey data. The group said it wanted to demonstrate how easy it was for a new polling outfit to gain credibility with media organizations and polling aggregators.
The admission came after FiftyPlusOne researchers examined purported raw data from surveys published by Wedgewood, including a Mississippi Senate poll, and found irregularities that raised questions about its authenticity. When confronted, Wedgewood issued a public confession.
Major polling aggregators and media organizations began removing Wedgewood data from their databases and coverage on Wednesday.
Races that had been included in the polls released by Wedgewood included the Michigan Senate seat, Louisiana's Republican Senate primary and the Mississippi Senate race.
"These actions do harm the collection of survey data which support a number of valid social observations and reporting on important population trends about policy preferences or measures of economic and social status and wellbeing," Traugott said. "If they contribute to lowering faith and confidence in the data collection method, specifically in lower response rates that can affect the representativeness of samples, society suffers from the lower quality data that result."
Silver Says Fake Polls a Real Problem
In a Thursday update, Silver wrote that fake polls had become a very real problem in reporting on American politics for several years, citing previous instances where he had been led to question data stretching back to 2009.
While Wedgewood's explanation for its work was that it was an “experimental project to demonstrate the vulnerabilities, lack of oversight, and overall low barrier to entry in the modern political polling industry," Silver said he did not believe those behind the project.
"My opinion is that Wedgewood’s explanation is bull****, a pathetic attempt to get ahead of the story after FiftyPlusOne was set to publish a story exposing the fraud," Silver wrote. "Whoever was responsible for fabricating the data probably shouldn't be trusted with anything election- or poll-related again."
Silver then turned to the newsletter to examine the polls that had been released by Wedgewood, including multiple fabricated surveys in Maine, Michigan and Texas, saying there had been a slight lean toward Democratic candidates in the results.
He also called some of the choices made unusual, as surveys were purported to have been run in races with very little other polling data already, making it difficult to determine averages alongside other more established pollsters.
"Overall, compared with past fake pollsters — which often tried to fit in rather than draw attention to themselves — the Wedgewood “polls” were a bit more narrative-shaping. Though I also wouldn’t overstate this," Silver said.
Silver also defended the vetting of polls, saying that his site often draws from The New York Times' polling site, which he said was staffed by highly skilled professionals who do carry out a lot of vetting—and had led to other fake polls being removed or never posted in the first place.
He said that no poll vetting policy would ever be 100 percent accurate, as having too strict a policy could lead to false positives or other biases in trends.
Looking ahead, Silver said there was a risk moving forward that fake polls could be used to influence prediction markets or to discredit the industry overall.
"Still, the proprietors of poll-collection and forecasting websites are less naive than Wedgewood alleges," Silver said. "People in the industry know the data really well, and from Strategic Vision to Wedgewood, everyone has caught their share of scams at some point. Our instincts about when something doesn’t pass muster are usually pretty good. Or at least, so we’d like to think. The smartest cheaters are the ones who never get caught."
Contact Newsweek editors on this story: Jason Lemon, Dave Siminoff and Edward T. Cummins
Submit Correction
View Editorial & AI Guidelines