Iowans: None of the Above

For 60 years, the Iowa Poll has been sounding out residents of this heartland state on the critical issues of the day.

Is the president managing the economy well?

Should the governor allow more casinos to be built?

And -- tell the truth now, please -- how many of you really like corndogs?

Founded in the days when getting a representative sample meant interviewing people with different types of cars in their driveways, the Iowa Poll is the longest-running scientific state survey in the nation.

It offers much well-respected political analysis. But it also boasts a venerable tradition of quizzing residents on every oddball topic imaginable, from their lawn ornaments (mostly plastic deer) to their favorite vegetable (lettuce) to their faith (they put their trust in God, the armed forces and the Iowa State Patrol, in that order).

Such questions offer offbeat insight into the state that holds the nation’s first presidential contest. And they give Iowans a chance to consider the values, customs and quirks of sensibility that unite them.

“There are just a handful of states that have a real strong state identity: Texas does, and California, and Iowa. Think about the others: Is there such a thing as a Michigander? Not really,” said J. Ann Selzer, the Iowa Poll’s director.

“Part of what has given Iowa such a strong identity is that over the years, the Iowa Poll has been able to hold up a mirror. It gives us a chance to take a look at ourselves. And to not take ourselves too seriously.”

Last fall, for instance, with the Democratic caucus just a few months away, the poll tossed aside all the pundit patter and asked likely voters to define themselves, by picking one or more labels from a long list. The results: 39% said they were “do-gooders.” Feminists and flag-wavers each drew 22%. And 11% confessed to being a “couch potato.” (Raising the question of who tunes into the political debates that crowd the calendar all fall, only 10% of likely voters described themselves as “policy wonks.”)

It’s hard to imagine in this era of nonstop polling, but when the Iowa Poll was launched in December 1943, it was a novelty.

Newspaper reporters had been conducting informal straw polls since the 1880s, stopping people on the street for quick interviews or inviting readers to send in their thoughts on the issues of the day. But the first polls with a claim to being scientific did not emerge until 1936, when George Gallup bet that he could predict the presidential election better than the Literary Digest, a national magazine that routinely surveyed its readers.

The Literary Digest, with a heavily Republican readership, announced that Alfred M. Landon would win. Gallup, who sent interviewers door-to-door in rich and poor neighborhoods alike, placed his money on President Franklin Roosevelt. FDR won reelection, and Gallup had himself a career.

Newspapers around the country scrambled to buy Gallup’s syndicated poll stories, which were packaged as full-page spreads on political races and policy debates. The concept intrigued Gardner “Mike” Cowles Jr., then the president of the Des Moines Register, Iowa’s largest paper.

On Dec. 12, 1943, he announced that the paper would hire lawyers, housewives, college professors and others to ask Iowans a list of questions each week. Expecting readers to be skeptical about the newfangled science, Cowles created a 14-member committee to assure them that the survey was on the up-and-up. Among the Poll’s first advisors: A bishop, a rabbi, a bank president, a union organizer and the president of the Iowa Farm Bureau.

Credibility established, Cowles promoted his survey’s results nationwide, touting the Register’s “forward-looking innovation.” (A Texas editor founded a state poll three years earlier, but took a hiatus during World War II, giving the Iowa Poll a claim on the title of oldest continuous state survey.)

Then, as now, the poll was used not only to make news and influence policy but also to generate buzz about the Des Moines paper. Other polls throw in screwball questions now and then, but none has made them such a staple.

Register readers seem fascinated by the resulting stories, which have informed them over the years that 11% of their neighbors have had a supernatural vision and 31% have mastered the hokeypokey. In 1990, they learned that -- stereotypes aside -- only one in four Iowans owned a pair of bib overalls. And how about this baffling factoid, from 1989: 40% of tropical-fish owners wished that their pets were more affectionate.

State lawmakers, in impassioned floor speeches, may cite the poll’s findings on such sober issues as the need for parking meters in downtown Des Moines, or support for abortion restrictions. But poll writers know what the people like to read: Over the years, they have asked more questions about margarine than about nuclear weapons.

“God knows, we need to laugh at ourselves now and then,” said Richard Tapscott, the Register’s managing editor.

Preserved in ragged, paste-blotched scrapbooks in a corner of the Register’s old mailroom, the first polls delved into issues that still resonate today, from farm subsidies to universal health insurance.

Those early polls uncovered a strong internationalist strain in Iowa -- one that still persists, especially among farmers hoping to sell their crops abroad. In 1944, for instance, 81% of Iowans supported postwar food rationing in the U.S. so that more of America’s bounty could be sent to Europe. National support for that idea was 67%, according to a Gallup survey.

On other issues, an austere Midwestern sense of morality seemed to tug Iowans out of step with the rest of the nation. In 1951, 46% backed the death penalty for “dope peddling.” In the nation as a whole, just 14% took such a harsh view.

Such findings were printed on the Register’s editorial page each Sunday. Because the Register served then and now as a statewide newspaper, the articles had remarkable influence.

“It sounds kind of corny now, but the whole purpose was to let Iowans know what their fellow Iowans were thinking,” said Glenn Roberts, who joined the poll as a researcher in 1945 and later directed it for some 25 years.

The science behind Iowa’s early polls was a technique pioneered by Gallup called quota sampling. Researchers studied the state’s demographics and then tried to replicate them in miniature, ordering each interviewer to question five men, four women, one college graduate, two farmers, three blue-collar laborers and the like.

Each interviewer was assigned a specific neighborhood, usually his or her own. For 65 cents an hour, it was up to the interviewer to figure out whom to approach.

In a chart of “factors useful in identifying socio-economic status,” an Iowa Poll manual from the 1940s recommended that interviewers scope out a potential subject’s car, refrigerator, telephone and furniture -- or lack thereof.

Polling in those early days was so unusual that respondents rarely slammed the door on interviewers. The man or woman on the doorstep bearing a clipboard full of questions was often treated as a guest, not a nuisance; Iowa Poll interviewers were given an official ID card, but the manual cheerfully noted, “It is doubtful whether you will have much need for this.”

On the contrary: “You’d be invited in the house and served coffee and cookies. You’d be trying to interview the husband and he’d be yelling, ‘Hey, Maude, come down and talk to this gentleman.’ You had trouble getting away,” said Merv Field, who started a state poll in California in 1946.

Quota sampling was discredited in 1948, when nearly every poll, including Iowa’s, called the presidential election for Thomas E. Dewey instead of Harry S. Truman. Meeting in Iowa City a few months later, top national pollsters agreed to alter their methods. Instead of letting interviewers pick their subjects, poll directors would randomly select households, then insist that the interviewer return repeatedly until he or she had made contact.

That system has been refined over the years, especially as telephone interviews replaced face-to-face surveys in the 1970s. Most scientific polls today use random digit dialing. Computers generate endless strings of numbers within a given area code; so every resident with a phone line has an equal chance of being called.

“To many people, it appears to be a kind of voodoo. We talk to 800 or 1,000 or 1,200 people and then project their views onto the whole population,” said Mark A. Schulman, a veteran pollster and head of SRBI, a research firm in New York. But if the interviews are truly random, they should accurately reflect the views of a much broader group, within a few percentage points, Schulman said.

In a small state like Iowa, with a population of 3 million, interviewing 800 people is often enough to say with authority not only how adults divide on an issue but also how various subgroups view the topic.

A recent Iowa Poll on sin, for instance, uncovered wide divides: While 27% of Republicans called looking at Playboy a “major sin,” only 13% of Democrats agreed. As for padding expense accounts? A whopping 72% of rural Iowans deemed that a major sin. Just 56% of city folk were as harsh.

By keeping its interviews to 12 minutes -- and by throwing in fun questions now and then -- the Iowa Poll attains “outstanding cooperation” from the public, said Selzer, its director. The response rate for a three-day poll is usually 50% to 60%, meaning interviewers reach at least half their randomly selected targets.

For the Iowa Poll, the quadrennial caucus surveys pose the toughest challenge.

Only the most committed activists vote, since participating in a caucus requires leaving the house on a cold January evening for a three-hour argument with the neighbors. This year’s Democratic caucus is expected to draw about 125,000 voters. Selzer’s interviewers can easily spend an hour on the phone before reaching a single voter who plans to attend.

And just how firm those plans are matters -- a lot. The most recent Iowa Poll, taken in early November, found Rep. Dick Gephardt of Missouri leading the field of Democrats by 7 percentage points. But when Selzer looked only at voters who said they would definitely, as opposed to probably, attend the caucus, the congressman’s lead shrank sharply. The “definite” group favored former Vermont Gov. Howard Dean. Subsequent polls by other organizations have found Dean and Gephardt locked in a race for first.

The caucus poll has other limitations as well. Selzer calls only registered voters; she doesn’t try to question anyone who might register on the night of the caucuses, as state law permits.

Also, none of her interviewers is fluent in Spanish or the other languages spoken by Iowa’s growing immigrant population.

Still, when the next Iowa Poll is published -- the Register won’t disclose the date -- candidates and their advisors will pore over cross-tabulations of voter demographics, comparing the data with their own internal surveys as they plot the final days of their campaigns.

They might also gain insight from the quirky factoids that the poll has made famous.

For instance: Candidates hoping to strike a chord with Iowa voters invariably pose with foot-long corndogs at the Iowa State Fair. But a poll taken in September revealed that a full 44% of Iowans found the breaded grease-on-a-stick distasteful.

The politicians -- and their arteries -- might be better served by choosing corn on the cob. That’s far and away Iowans’ favorite food.

No word yet on whether the latest fair treat, deep-fried Oreos, is a crowd pleaser. But a scientific analysis of Iowans’ views on battered desserts can’t be far off.