Less than two weeks before Election Day, Americans are drowning in opinion polls. New surveys are released at a rate of about a dozen per day. We are, by turns, heartened and tormented by the respected Quinnipiac, the troubling Times/Siena, the right-leaning Rasmussen. People who are not paid to attend to such matters can be heard using phrases like “sample size,” “margin of error” and even “cross tabs.” We are a nation obsessed.

For this obsession we can thank, in part, George Gallup, who took public-opinion polling from the private-sector world of market research and applied it to American politics in the mid-1930s. He saw a problem with the informal and unscientific way polls were being conducted at the time. In 1936, Literary Digest went about surveying the public by scouring the phone book and car registration data and then sending millions of people a postcard asking whom they planned to vote for in the presidential election: Democratic incumbent Franklin D. Roosevelt or Republican challenger Alf Landon. Most respondents opted for Landon. The magazine predicted he would pick up 370 Electoral College votes. Landon wound up with eight. The magazine closed not long thereafter.

George Gallup, by contrast, knew that poorer voters who tended to support Roosevelt were less likely to have phones or cars. A former professor of journalism at Iowa State University, Gallup sought to craft a sample of voters that better reflected the voting public. By doing so, he predicted Roosevelt’s landslide in the 1936 election. On the strength of that success, he built a polling operation based in Princeton, New Jersey, which began publishing surveys of the American electorate’s opinions about both candidates and policies.

In early 1940, The Nation published an article about Gallup and the brand-new world of opinion polling, written by a 24-year-old James Wechsler, later the influential editor of the New York Post. Wechsler noted that, partly thanks to Gallup’s predictions in Roosevelt’s favor, most criticism of opinion polling was coming from conservative opponents of the New Deal, who argued that “polls are intrinsically a menace to the republic: they reveal only the sum total of popular ignorance; they foster the heresy that Mr. Milquetoast has something to say and a right to be heard even between election days; they thereby imperil the structure of ‘representative government.’”

While Wechsler didn’t share such antidemocratic concerns, he warned Nation readers to be skeptical of the profusion of opinion polls. After witnessing employees of Gallup’s American Institute of Public Opinion conduct interviews in three cities, he concluded that the potential for “manipulation of the polls by the conservative interests which ultimately pay for them” was a legitimate concern. (Gallup’s operation was funded by newspapers that subscribed to receive his data and reports.) “There is ample chance for sabotage along this assembly line,” Wechsler wrote.

Wechsler depicted Gallup as a nonpartisan expert “conscientiously groping for a vantage point above the battle where questions can be formulated in a spirit of peace and neutrality.”

Nearly five years later, Nation readers had a chance to hear from the man himself in a December 1944 article headlined, “I Don’t Take Sides.” He was responding to a critical essay the magazine had published two weeks earlier by a New Deal economist named Benjamin Ginzburg, who accused Gallup of refusing to share how he conducted his polling operations on the eve of that year’s presidential election. Whereas four years earlier conservatives had criticized Gallup, now it was Democrats who accused the pollster of putting out cooked survey figures.

In his Nation piece, Gallup dismissed the criticism as “unintelligent” and accused Ginzburg of “careless reporting,” while insisting that he welcomed any investigation of his polls. He also noted that, since both parties were accusing him of bias, he must be doing something right. “You can‘t make much of a case of political bias out of this record, unless perchance you regard us as chameleons who change political colors every two years,” Gallup wrote. “In this case at least one would have to admit that we distribute our political favors evenly.”