President-elect Donald Trump lawsuit It may be the first time in a modern U.S. presidential election that he is up against veteran pollster J. Ann Seltzer with an off-the-charts poll released just before last month's election.
When I looked into my book, I found thatLost to Gallup”, which tells a history of high-profile polling failures in presidential elections since 1936, I did not uncover anything resembling Trump’s case accusing Selzer. “Brazen election interference”
The case is likely to be a tough legal challenge and may not survive. But its apparently unprecedented nature could have worrying implications for pollsters. Checkered record of accuracy. One can imagine how wary and cautious pollsters would be if they faced the risk of litigation if pre-election polls went in an unexpected direction.
The focus of Trump's lawsuit is Selzer pollannounced three days before the Nov. 5 election, shows Vice President Kamala Harris extending a 3-point lead in Iowa that Trump easily led in the 2016 and 2020 elections. Ta. Selzer said of the poll results that Harris has “clearly jumped into the lead.”
The meaning was clear. If Harris had actually made headway in pro-Republican Iowa, her chances of winning the battleground state and, by extension, the presidency, would have been much better. Ultimately, this poll confirmed Selzer's obscurity and her stellar reputation for accuracy in Iowa. In fact, this year's fellow pollsters called her “oracle of iowa”
Then Trump won Iowa. 13 percentage pointsThis means Selzer's poll was wrong by 16 points. It was a humiliating misfire for the veteran and well-respected pollster.
But Trump's lawsuit says the poll was not a “mistake” and that by projecting a “false narrative of inevitability” about Harris' prospects, he “attempted to influence the outcome of the 2024 presidential election.” ” claims to be a thing.
Trump also claims that Selzer's “huge platform and following” gave her “a significant and influential opportunity to deceive voters.”
While not directly challenging the merits or logic of Trump's lawsuit, the American Institute for Public Opinion Research, the nation's largest polling organization, said: statement: “Differences between poll results and election results can and often do occur for reasons unrelated to fraud or fraud. Such fluctuations It highlights the complexity of the issue and the importance of interpreting polls within their limits.”
seltzer, recent interviews On a PBS station in Iowa, she denied that her poll was intentionally off the mark. She said it would go against her ethics. Selzer also said he doesn't know why his poll was so wrong. Thinking about it “keeps me awake in the middle of the night,” she said, adding, “We don't know. Should I have known? Yes, I would have.”
Immediately after the election, Mr. Selzer said he had previously decided to retire from polling and had informed his major sponsor, the Des Moines Register, of his decision.
One line of defense for Selzer and her co-defendants, the Register and its owner Gannett, may be found in a lazy but longstanding investigative cliché: The idea is that pre-election polls are just a “snapshot at a certain point in time.” “Snapshot” advocates may argue that the poll was accurate when it was taken in late October, and that Trump gained 16 points in the days that followed.
As far-fetched as it may seem, it's unclear how President Trump would refute such claims.
Perhaps a stronger line of defense lies in the protection of the First Amendment, and pollsters, like journalists, should do so. well-intentioned mistakes are forgiven. Without such flexibility, or as the U.S. Supreme Court in 1964 said, “breathing space” can stifle active public discussion and debate.
Criticism of election pollsters has been frequent and even harsh since George Gallup, Elmo Roper and others began developing quasi-scientific polling methods in the mid-1930s. The polling technique now looks like this: more sophisticated Decades later, even with the precision of polling, It was uneven.
For example, polls generally underestimate the extent of support for Trump in each area. three recent presidential elections – in spite of correction The goal was to reach out to and interview previously elusive Trump supporters.
One research organization found that this year's final pre-election polls underestimated President Trump's approval rating by an average of 2.4 percentage points. Analysis by NBC News. in 2020polls overall underestimated President Trump's support by 3.3 percentage points, his worst overall performance since 1980.
NBC reported that this year's modest showings of support for Trump “extended to polls in states across the political spectrum.” Furthermore, President Trump's approval rating was generally underestimated in public opinion polls conducted in the United States. Each of the seven battleground states Where elections have changed.
Although public opinion polls are being conducted, general skepticism For years, taking pollsters to court over presidential election results was unheard of until President Trump filed suit against Selzer.
Eighty years ago, Gallup went before a House committee and testified about his polling methods and the discrepancies between his polls and the results of that year's presidential election. Gallup predicted that the race between President Franklin D. Roosevelt and Republican Thomas E. Dewey would be closer than it actually was.
As he wanted to do, Gallup engaged in what today would be recognized as a.spin” he told a House committee during a heartfelt hearing. 1944 public opinion poll “…they came through the election with flying colors. Although their record falls short of absolute accuracy, it exhibits a degree of accuracy found in some fields outside the exact and physical sciences.”
But four years later, Gallup and other polling organizations made a huge mistake in predicting Dewey's sure victory over Roosevelt's successor, President Harry S. Truman. probably, The biggest shock election Truman won reelection by a 4.5 percentage point margin, the most in presidential history. Gallup predicts Mr. Truman will lose by 5 points, meaning his poll has a margin of error of 9.5 points.
that errorNo doubt it was on Gallup's mind during a lecture he gave in Cleveland in late December 1948. “Doctors can bury their mistakes,” he said. A lawyer can justify his claim. …But the researcher of public opinion must stand naked before the world, and his shame will be recorded for posterity. ”
Exaggerated or not, it's an observation that has resonance today.
W. Joseph Campbell He is a professor emeritus at American University and the author of the following books: 7 nonfiction books On a variety of public relations and media history topics.
