Think fast: How many Americans went to church last Sunday? Forty percent? Twenty percent? Somewhere in between?
Whatever the guess, it might be right.
For years, some polling groups and sociologists who specialize in tracking the religious behavior of Americans have been at odds trying to get a handle on how many of us go to worship services on a regular basis.
Two of the best-known names in religion polling _ the Gallup Organization and Barna Research Group _ say about four in 10 Americans go to church weekly, a figure that has remained fairly consistent since 1939, when Gallup first began collecting such data.
But sociologists C. Kirk Hadaway, Penny Long Marler and Mark Chaves say in the February 1998 issue of the scholarly journal American Sociological Review that if actual heads were counted each Sunday just 20 percent of Americans would be found in church.
In "Overreporting Church Attendance in America: Evidence That Demands the Same Verdict," the authors blame overreporting of church attendance by survey respondents as the primary reason for a major "gap" between their findings and the poll-based estimates.
"Most sociologists tended to think there was a little bit of a gap between what people said and what they did but we didn't see how large the gap was until we began (counting attendance) ourselves in 1992," said Hadaway, a United Church of Christ official who has tracked church membership and attendance trends for 20 years.
"Our initial study, which was based on attendance counts of Protestants in one Ohio county and Catholics in 18 dioceses, indicated a much lower rate of religious participation than the polls report," he said. "Instead of 40 percent of Protestants attending church, we found 20 percent. Instead of 50 percent of Catholics attending church, we found 28 percent."
The sociologists counted people in pews and cars in church parking lots, as well as studied attendance reports provided by pastors in Ashtabula County, Ohio, a Midwest region generally indicative of church-going trends for the nation, Hadaway said.
Despite these findings, David Kinnaman remains confident in the 43 percent church-attendance rate the Barna Research Group came up with for 1998.
"We totally stand by our numbers, they are arrived at with ongoing research," said Kinnaman, research director at Barna.
"Are there ways to come up with the information differently? Are there other approaches to study the problem and do those approaches and methodologies yield differing percentages? Yes," he said. "But those are not inherently contradictory. We are simply looking at an issue from a different angle."
But Hadaway and his colleagues believe some data-gathering methods are more reliable than others.
"We believe that too much trust has been placed in survey data and not enough attention given to membership records, patterns of giving, and even the incredulity of local church pastors when they hear that 40 percent of Americans attend church during an average week," said Hadaway.
Robert Woodberry, a sociologist at the University of North Carolina, Chapel Hill, who has studied church-going trends for three years, has serious reservations about the methodology used by Hadaway and his colleagues.
In counting Catholic attendance, for example, Hadaway, Marler and Chaves did not include Eastern Rite Catholics, or those who attended Mass at a venue other than a parish church, such as at shrines, hospital chapels or on college campuses, he said.
Woodberry also believes Hadaway overestimated the Catholic population in Ashtabula County.
"My research shows that when you adjust to correct for the people who aren't included in head counts and use more accurate estimates of the Catholic and Protestant populations attendance figures come up substantially," said Woodberry, who estimates the national church-going rate at about 29 percent.
"We don't buy Woodberry's argument that we overestimated the Catholic population," Hadaway said. "He's got a fairly convoluted way of suggesting that we do based on some surveys that have found slightly lower Catholic populations" in Ashtabula County.
"But we are dealing with the entire Christian population, not just the Catholic population," he added. "It really doesn't matter whether you slighly overestimate or underestimate the population of a single group."
At the same time, Woodberry is critical of trying to determine who goes to church by asking questions over the telephone, as Gallup and Barna do.
"Most polls oversample regular church attenders and so overrepresent their opinions," he said, adding that regular churchgoers are generally easier to contact at home, are more cooperative in answering surveys and don't mind telling pollsters about their religious activities.
Kinnaman rejected Woodberry's claim, saying he has never seen reliable data to suggest regular church-goers are oversampled.
Even with its inherent complexities, Hadaway prefers the head-counting method over polling because he questions whether Americans reliably report their activities to pollsters.
"How much trust should we put in what people say they do? Perhaps not as much as we have up to this point," said Hadaway. "Numerous studies show that Americans misreport voting, gifts to charities and illegal drug use. And their misreporting is in the expected direction: higher for voting and charity giving; lower for drug use."
But, he said, that doesn't mean survey respondents intentionally lead pollsters astray.
"People do not tend to engage in mental enumeration as much as they report behavioral averages or personal rules for their behavior," he said. For example, if someone attends church regularly and misses the Sunday before being polled because of illness or vacation, they might still say they went to church because they see themselves as someone who indeed does go to church every week.
Experts call this socially desirable exaggeration, and Michael Hout of the University of California, Berkeley, and the Rev. Andrew M. Greeley of the University of Chicago and a columnist for RNS, acknowledge its presence in polling, but not to the degree Hadaway and his colleagues claim.
"We doubt that people exaggerate so much that reported attendance is actually twice as high as their actual attendance," Hout and Greeley write in a rebuttal to Hadaway, Marler and Chaves appearing in the same issue of American Sociological Review.
"Lacking a study that links head counts with survey responses, can we say anything about how much socially desirable misreporting could bias survey estimates of church attendance?" they ask.
One thing is certain: Analyzing mounds of data mixed with a seemingly infinite quantity of unknowns makes calculating the exact number of Americans who attend church weekly an arduous assignment at best.
"With so many variables to consider it is very difficult to say this is precisely what it is," Woodberry admitted about his 29 percent figure.
"It is a formidable task," agreed Hadaway. "But by putting together all the information, you can make some summary judgment," he said, standing by his research.
"In terms of the real level of American church attendance, based on our total counts in these places . . . I would say that American church attendance is somewhere around 21 to 22 percent at most."
Said Kinnaman: "People try to discount an approach simply because the numbers don't line up with what they experience or expect and that's not appropriate. The real struggle is for truth and accuracy."