The truth about us

IF YOU BLINKED, you might have missed it: the week or two between the 2006 midterm elections and the start of the 2008 presidential season, a brief hiatus when poll data didn’t lead the national news.

But it’s over now. In the weeks and months ahead, rest assured that pollsters will measure every twist and turn, whether major or minuscule, in the upcoming race. We’ll be buffeted by percentages comparing Clinton to Obama, McCain to Giuliani, red states to blue states, “values” voters to the “pocketbook” variety, and so on. And accompanying these polls, of course, will be all manner of survey results purporting to describe the citizenry to itself -- from our collective “consumer confidence” to our views on the so-called war on terror, from our habits of worship to our preferred television programs.

How did we arrive at this strange state of affairs, in which we look first to polling data to figure out who we are and where we stand? In which a flood of quantitative reportage drowns out other kinds of information and analysis? When -- and why -- did we become a survey-obsessed society?

Population surveys, whether for the purpose of levying taxes or raising militaries, date back as far as William the Conqueror’s Domesday Book of 1086. Modern nation-states have for centuries collected census data in order to track everything from public health to population growth to economic progress.

In the 19th century, American newspapers began running straw polls of readers during election seasons. Around the same time, early insurance and marketing agencies found profits in statistical tabulations of life spans and buying patterns. Reformers and philanthropists, motivated by the power of empirical data to clarify social problems, canvassed immigrant neighborhoods and factory laborers.

It was not until after World War I, however, that popular polls and surveys began to infiltrate ordinary Americans’ lives in earnest. By 1948, a reporter would remark: “This is the great age of confession. We are required now to tell everything.... We tell Dr. Gallup how we are going to vote.... Our psychiatrist delves into our sex dreams and Dr. Kinsey into our actual performance along those lines.” Mused another a few years later, “Today, unless you can say ‘According to the Poop-A-Doop survey, umpty-ump percent of the people chew gum while they read Hot Shot News!’ you fail to make an impression.”

A number of developments ushered in this new era. One was the rise of the professional social sciences -- sociology, economics and political science -- which firmly established themselves in the first decades of the 20th century through claims to special expertise and objectivity in investigating social life. Another was the invention and refinement of new statistical techniques, including, most notably, scientific sampling, a mathematical tool that allows researchers to gauge the attitudes of the entire society by querying as few as 1,000 people.

Equally important were the actions of the surveyors themselves, those who perceived a demand, and sometimes a market, for statistics about “ourselves.” Unlike their 19th century predecessors -- who had focused on social problems and those they considered degenerates, delinquents and defectives -- 20th century public opinion pollsters, commercial researchers and sex surveyors turned to investigating (and one might argue, creating) new social entities: “average” or “typical” American habits, attitudes and beliefs.

A torrent of social data conveyed through charts, graphs and statistics began to inundate newspaper and broadcasting networks. By 1940, for example, an estimated 8 million people were receiving tri-weekly reports of “What America Thinks,” George Gallup’s syndicated opinion polls.

The audience for these statistics was ready made. In an era of rapid urbanization and industrialization, and the seeming breakdown of older mores and communities, Americans were eager to know what bound a diverse and contentious population together. Surveyors were happy to oblige, measuring everything from what citizens bought to what they believed to what they did in the privacy of their homes.

One of the earliest surveys to gain national attention was “Middletown,” Robert and Helen Merrell Lynd’s 1929 investigation of an anonymous “representative community” (Muncie, Ind.). The study tabulated seemingly mundane trends, such as the movie-watching habits, house sizes and religious beliefs of “typical” citizens. The result was the first-ever sociological bestseller, surprising its publisher, booksellers and the surveyors themselves. What was it that fascinated readers in the Lynds’ lengthy, empirical study? Many regarded it as a mirror of modern America, the first scientific account to reveal, in one journalist’s words, “the truth about ourselves.”

Others would soon join the Lynds in what amounted to an exhaustive cataloging of American life. Among the most influential were “scientific pollers” such as Gallup and Elmo Roper. Rather than survey a single community, they questioned a small national cross-section of respondents to derive a representative sample of the population.

The pollsters had great ambitions for their new instrument. Gallup himself announced that it could “provide a continuous chart of the opinions of the man in the street.” Polls on topics from war plans to working women, tax policy to term limits, swiftly became public knowledge, prompting one writer in 1941 to claim that “America has had a firsthand opportunity to become acquainted with its own mind.”

It was the pollsters’ efforts to locate the majority, a fixed data point amid all the chaos of viewpoints and variability in the population, that made their techniques so appealing to politicians and citizens alike. As with the Lynds’ survey, opinion polls were in the business of forging, not simply investigating, “the American public.”

In the case of both “Middletown” and the Gallup Poll, some citizens wondered whether such quantitative knowledge about the nation -- and the way it was obtained -- was beneficial. One resident of Muncie complained, for example, that decent people would not “permit this peeping into the deepest recesses of their lives.” Many Americans worried that continuous access to poll results would distort political discourse or create a population overly attuned to (and swayed by) the statistical majority.

These fears were thrown in sharp relief by perhaps the most controversial social scientific study of the century, Alfred Kinsey’s “Sexual Behavior” surveys of 1948 and 1953. Kinsey, through in-depth interviews, claimed to uncover how average citizens really behaved in their most private lives. The surveyor’s shocking findings -- that 37% of American men experienced some kind of “homosexual contact” and that nearly half of American women engaged in premarital sex, to give just two examples -- spread like wildfire across the country.

Some commentators praised Kinsey’s findings as “the greatest mass survey of normal sex activity in our society.” But legions of others challenged the accuracy of his figures, the truthfulness of his subjects and the very notion of transforming a complex human activity into statistics about who was doing what with whom and how often.

More intriguingly, some citizens bewailed the release of facts about masturbation and extramarital affairs into the public sphere, worrying that individuals would begin to measure themselves against scientifically generated norms. They knew that, in a society constantly seeking information about itself, data about “typical Americans” could act as a powerful social force, shaping citizens’ views about normality, morality and even their own actions.

What is striking about the early history of modern polls and surveys is how regularly their consumers challenged and engaged with them. Can we say the same today? In the thoroughly surveyed culture of the 21st century, we don’t usually stop to think about the role that bell curves and pie charts play in forming our perceptions of that elusive entity, the “American public,” or the matter-of-fact way that statistics have come to color our understandings of social reality.

When a poorly conducted exit poll from the 2004 election trumpeted the finding that 22% of Americans cast their votes based on “moral values” rather than the war in Iraq or the state of the economy, we ought to have been suspicious. How was the question asked? What do “moral values” mean in the context of politics anyway? Instead, that single number fed simplistic media narratives about red states and blue states, a polarized nation and an ongoing culture war.

There is no going back to a world without polls and surveys. Statistical claims about “who we are” are clearly here to stay. But we might learn something from those Americans who first encountered the sex surveyor and the opinion pollster: They recognized that quantitative data created a new -- but only a partial -- way of seeing the nation, and they were skeptical that surveys could tell us the whole story about ourselves.