History Exposes the Perils of Relying on Polls

Polls are a centerpiece of our political news, especially as we get closer to Election Day. Journalists and forecasters will lean on these surveys to answer the question that will define the next five-plus weeks: Who will win? Kamala Harris or Donald Trump?

The pressure to break the news first creates a danger represented most infamously by a three-word Chicago Tribune headline: “Dewey Defeats Truman.” This blunder stemmed from the certainty that journalists felt, based on the polls, that New York Governor Thomas Dewey would defeat President Harry Truman in 1948.

Washington, D.C.-based business journalist W. M. (“Kip”) Kiplinger was one of the many who got the race wrong in his influential publications, the Kiplinger Washington Letter and the Kiplinger Magazine. He was so confident that Dewey would win, that he ordered an entire issue of Kiplinger Magazine focused on the new administration’s policies, with a cover that read, “What Dewey Will Do.” 

When Dewey lost, angry readers cancelled subscriptions, showing the risks today’s journalists face in trying to predict the neck-and-neck contest between Harris and Trump. If anything, the danger of incorrect predictions might be higher in 2024 than in 1948 due to the unreliability and proliferation of polls, the pressure to publish first in our 24-7 news cycle, and the rampant disinformation circulating on social media.

Kiplinger had gained prominence due to his Kiplinger Washington Letter, a pioneering political and economic newsletter that had become a must-read during the New Deal for business leaders seeking to understand the direction of government policies. Flush with money in 1947, he launched the Kiplinger Magazine, which grew into one of the leading personal finance magazines. Kiplinger was a moderate conservative, but his publications were non-partisan.

Besides being an excellent journalist, Kiplinger also developed a national reputation as a forecaster who could articulate political and policy trends before they became a reality. These forecasts became Kiplinger’s brand, and they developed from careful reporting. Yet in the months leading up to the 1948 election, the magazine published some strong pro-Dewey forecasts that were particularly confident.

Read More: How to Read Political Polls Like a Pro

In the magazine, the “Months Ahead” column in August 1948 declared: “Election campaign, resulting in victory for Dewey and Warren, giving the Republicans an excellent chance of being in office for eight years, through 1956.” The following month, the column declared: “In November, the election of Dewey as next President…In December, you will see Truman preparing to leave office with the good feeling of the American people.”

Even so, when it became time to plan for a post-election issue, editorial chief of staff George Bryant suggested producing two issues, one analyzing policies under a Truman Administration and the other for a Dewey Administration. “We would put out two magazines for November and we would mail the right one,” Bryant recalled years later.

Kiplinger initially adopted that plan, but the burden of creating two magazine issues was great. Then, the three major pollsters of the day — Gallup, Roper and Crossley — all released surveys showing that Dewey was well ahead of Truman. It prompted Kiplinger to decide to put together only one November issue, which would focus on Dewey. The publisher was competitive and eager to get a journalistic scoop. “Willard had dreams of how he was going to beat everybody on Dewey,” Gale Kiplinger, the journalist’s younger brother, later recalled.

There was some unease about making a Dewey prediction before the votes had been counted. In a staff meeting before publication, senior editor Clarence Marshall argued against committing to an entire Dewey edition. He warned against repeating the mistake of Literary Digest, which had incorrectly predicted that Herbert Hoover would win reelection in 1932. 

The staff’s reaction to Marshall’s warning? “They laughed at him,” recalled John Hazard, a senior Kiplinger Magazine writer and author of an unpublished company history. 

Meanwhile, Kiplinger doubled down and took out a full-page advertisement in TIME to run after the election, touting the “What Dewey Will Do” issue.

The staff’s apprehension was visible on Election Day; chief of staff Bryant recalled being “nervous as a cat” and pacing through the office. Kiplinger told Bryant to leave because he was worsening the newsroom jitters. Later, Kiplinger editors were rattled when Robert Ames, the circulation director, reported a long line at his polling place in heavily Democratic Virginia. Ames asked Kiplinger if he should hold back on mailing 15,000 copies of the issue, half of the press run. Kiplinger said, “'Mail them. You go home and go to bed, and by 2 o’clock in the morning, why, everything will be lovely,'” Ames recalled.

On election night, Kiplinger and his wife LaVerne hosted a dinner party, but it was interrupted by worried phone calls from his staff. Early news reports showed the vote trending toward Truman. Gale Kiplinger recalled his big brother telephoning him close to midnight, saying Dewy would lose. “I said, `I guess you're going to have to rewrite it,’” Gale Kiplinger recalled. Kip told him the magazine had already been mailed. “I said, `My God, what the hell are you going to do now?’ He said, `Damned if I know.’” About this time, Ames received a call from Kiplinger, asking if the Dewey issue had indeed been mailed. It had.

After the results were clear the next morning, Kiplinger sat in his office for some eight hours “and (did) not see anybody or talk to anybody,” Bryant recalled. “Kip was irritable. It hurt his pride, and he was a man with a lot of pride.” 

Some investors sent telegrams or telephoned, complaining about the wrong prediction. The Kiplinger staff tried to kill the full-page TIME ad, but it was too late.

Eventually, Kiplinger composed a memo apologizing to readers. “Dear Subscriber: We’re off base on Dewey — we fell flat on our editorial faces. We have no excuses, no alibis. The November issue was full of Dewey. It’s a fine job of reporting and analysis of what Dewey will do. Except for one little imperfection: Dewey wasn’t elected.” 

Ames and others admired the letter’s frank tone. Kiplinger didn’t blame anyone else, including the polls, the dozens of his confidential sources such as leading Democrats who also thought Dewey would win. “I think his letter and the personal tone it had, and the way he ate crow, was very well received by people who were basically his friends anyway,” Ames said.

Read More: Why the Polls May Be Feeding Liberals Another Blue Mirage

Nonetheless, the Kiplinger franchise suffered temporary damage. Subscriptions to the Washington Letter fell by 23,369 — 13% — over the next year. Ironically, the Kiplinger Magazine, still in its high growth start-up phase, actually gained subscribers the following year, and it more than doubled its circulation by 1950, to almost 75,000. Because many other media outlets, including the New York Times and NBC Radio, made similar Dewey predictions, “the disaster was far less than you would ever imagine,” Ames said.

Kiplinger “took no solace in the fact that he had plenty of company in underestimating Truman, and he promptly refunded the subscription fees of many irate subscribers,” as an internal company history chronicled.

For today’s reporters, the 1948 election debacle offers a clear warning about the fallibility of polls, which have underestimated Trump’s support among the white working class in both the 2016 and 2020 elections. 

There’s also a danger in the sociology of the newsroom, where intelligent journalists lock into a viewpoint and aren’t as receptive to competing evidence. The Kiplinger reporters privately had spent the fall ridiculing Dewey's campaign. Yet, as Bryant recalled, they never stopped to think that voters might be doing the same thing.

The Kiplinger case also shows another age-old newsroom dynamic: standing up to the boss isn’t easy, especially when the competitive juices are surging. On election night, there will be plenty of competition and an urge to be first. But journalists should listen to the veterans of the 1948 campaign: “The mail bags should have sat on the platform until the votes were counted,” Bryant said. 

Hazard, the newsroom’s unofficial historian, later asked why Kiplinger didn’t wait one day: “It wouldn’t have made a damn bit of difference.” 

Rob Wells, Ph.D., is an associate professor at the Philip Merrill College of Journalism at the University of Maryland. He is the author of The Insider: How the Kiplinger Newsletter Bridged Washington and Wall Street (University of Massachusetts Press, 2022).

Made by History takes readers beyond the headlines with articles written and edited by professional historians. Learn more about Made by History at TIME here. Opinions expressed do not necessarily reflect the views of TIME editors.