This essay is adapted from Sid Bedingfield's remarks delivered at the ICW's first symposium held in November, 2024.
The first results from the Minnesota Poll appeared in the Minneapolis Sunday Tribune on March 19, 1944, three months before allied troops stormed the beaches of France in the D-Day invasion. Earlier that year, the federal government’s Office of Price Administration had asked American citizens to reduce their use of gasoline or risk harming the war effort. The Minnesota Poll asked state residents two questions: Could their community cut back on gasoline, and could they, personally, use less gas? The answers were not surprising: About half the respondents said, “Yes,” their town could get by on less, but an overwhelming majority—75 percent—said they could not spare a drop. And so the Minnesota Poll was off and running.
The poll has its roots in Iowa, where the Cowles newspaper company got its start, and in Murphy Hall, home of the Hubbard School of Journalism and Mass Communication. One of the pioneers in the scientific study of public opinion, George Gallup, had attended graduate school at the University of Iowa, and until 1932 he taught journalism at Drake University in Des Moines. Gardner Cowles, patriarch of the Cowles family and publisher of the Des Moines Register,, got to know Gallup and followed his rise to prominence in the 1930s. In 1943, Cowles launched the Iowa Poll in the Des Moines Register; the next year, the Cowles papers in Minneapolis followed suit. They became the first two newspapers to sponsor their own polls.
To develop the poll’s methodology, John Cowles, owner of morning Minneapolis Tribune and the evening Star and Journal, engaged the director of the School of Journalism, Dr. Ralph D. Casey, and the school’s head of research, Dr. Ralph O. Nafzinger. Casey and Nafzinger modeled the Minnesota Poll after Gallup’s work at the American Institute of Public Opinion at Princeton.
The Minnesota Poll appeared just eight years after Gallup and Elmo Roper had used systematic random sampling to humiliate the famous Literary Digest straw poll by correctly forecasting FDR’s landslide reelection in 1936. We are so inundated with polls, polling averages, and forecasting models today that it is hard to imagine what it was like when systematic polling was a bright and shiny new object.
Emergence of Surveys of Mass Opinion
Mass surveys first emerged in the 1920s, after decades of economic and social upheaval in the United States. Since the 1880s, an industrializing economy had triggered waves of immigration and rapid urbanization as workers flocked from the farm to the factory—and from Europe and Asia to North America. For many, the country seemed to be changing before their eyes.
As historian Sarah E. Igo has noted, the term “mass society” came into vogue. It was used to describe the country’s transition from a collection of small towns and villages into a national community. The emergence of mass media had played a role in this process. The telegraph had spawned wires services that delivered national news to newspapers. The motion picture industry emerged in early 1900s and broadcast radio followed in the 1920s, both delivering content aimed at a large, diverse, and anonymous mass audience. In his book, The Search for Order, historian Robert Wiebe described the transition to a mass society this way: “It seemed that the age could only be comprehended in bulk… So people everywhere weighed, counted, and measured it.”
The Arrival of the Minnesota Poll
By the late 1930s, systematic public opinion polling was often described in utopian terms, as a kind of magic tool that could revitalize democracy in this new mass society. In launching the Minnesota Poll, the editors of the Sunday Tribune contributed some utopian rhetoric of their own. Democracy depended on the consent of the governed, the paper noted, but the public’s views were often hard to ascertain. In the Tribune’s view, public opinion polling would remedy this problem. The Minnesota Poll will help us “better understand and analyze our beliefs and attitudes” and show us “what information we lack, which can lead us to the right answers.”
The newspaper took laudable steps to root the poll in the community and enhance its credibility. The paper established an advisory board of 18 prominent citizens to recommend issues and review the wording of questions. They included the presidents of the University of Minnesota and of Carleton College; the chairman of the Pillsbury—John S. Pillsbury; labor leaders from the Minnesota Federation of Labor and the Congress of Industrial Organizations; and the president of the state’s Federation of Women’s Clubs.
To conduct the poll, a staff of 65 surveyors from across the state interviewed residents in their homes. Four out of five of the surveyors were women, mostly housewives, the newspaper reported, and many had to drive long distances into the country to interview farmers. The Minnesota Poll did not turn to telephone surveying until 1976.
In that first installment of the poll, the newspaper described what it called “the principle of public opinion sampling”—the need to capture an accurate cross-section of society. “You must construct in miniature a sample of the population which has all the significant characteristics of the entire population,” the paper said.
Early findings of the Minnesota Poll
In that first year, 1944, 78 percent of Minnesotans predicted Roosevelt would defeat Thomas Dewey and win an unprecedented fourth term in office. But a majority also favored a constitutional amendment prohibiting future presidents from serving more than two terms.
Sixty-three percent said they wanted the United States to belong to any “new national council or union of nations that comes out of the war.” The Tribune pointed out that this contradicted the view that the Midwest was overwhelmingly isolationist.
In April 1945, after the collapse of Germany, most Minnesotans—57 percent—said they did not believe Adolf Hitler was dead. (Conspiracy theories are nothing new.)
At the same time, a significant number of Minnesotans expressed pessimism about the war in the Pacific. One quarter said the war would not end before 1947.
After the United States dropped atom bombs on the Japanese cities of Hiroshima and Nagasaki, more than nine out of ten state residents said the nuclear attacks were “justified” and that they helped “shorten the war.”
Another 45 percent said they thought the existence of atomic bombs “makes war less likely in the future,” but 31 percent said they believed there was a “very real danger” that A-bombs might be used against the United States in the next quarter century.
And the questions keep coming. Most of them about serious public policy issues at the state, national, and global level. But in the 1970s, the newspaper shifted the poll’s emphasis, asking fewer questions about public affairs and more about how Minnesotans lived their lives. An internal history written by a member of the Star Tribune staff said the impetus for the change came from “outside evaluators,” presumably members of the advisory board.
Changing Demands on Polls and Surveys
The poll still surveyed political opinions, but increasingly it spent more time encouraging Minnesotans to describe themselves. For example, one question asked state residents to “describe who you REALLY are?” (With REALLY in all capital letters.) The interviewers said most Minnesotans were “shocked by the question” and at first were reluctant to answer. But once they did, an overwhelming majority were—in the internal history’s words—“positive, almost glowing about who they thought they were.” Two-thirds said their jobs “are really interesting and enjoyable,” while only five percent said their work was “dull and boring.”
In 1975, the Minnesota Poll asked a more provocative question: Do you think you are a “winner” or a “loser”?
Three quarters of respondents declared themselves winners. And who were these self-identified “losers”? Sadly, seventeen percent were women, more than three times the number of men who said they were “losers.”
In 1969, a quarter century after the poll began, Minneapolis Tribune editor Bower Hawthorne wrote something that remains true today. “What people are thinking today is news,” but “what people were thinking and doing 25 years ago is history.” Hawthorne concluded that, “The opportunity to view history with accurate insight on what people are thinking gives future historians an incomparable advantage over their predecessors.”
Thanks to the efforts of Dr. Benjamin Toff, director of the Minnesota Journalism Center, and his counterparts at the Roper Center at Cornell University, a searchable database of the Minnesota Poll now puts 80 years of public opinion just a few clicks away. It is a treasure trove for researchers. As the Tribune editor said, the digital database should bring more “accurate insight” to the study of public opinion and its history in Minnesota.