Counting religious believers is a tricky business

God and the demographers

Sometime in the next few months, the United Jewish Communities will release the data from its 2000 census of the American Jewish population–for the second time.

After a flurry of controversy over the phrasing of questions, the use of weighting techniques, the loss of data, and the very definition of who is a Jew, the UJC retracted its findings last October and went back to the drawing board. The survey had raised an alarm over a decline in Jewish numbers, and critics wondered if the statistics were withdrawn because they were demoralizing. Defenders countered that the $6 million survey had no methodological problems that couldn’t be fixed with a little effort. All this controversy over a faith with only some 6 million adherents in the United States.

The American press is awash with estimates about the size of various religious groups, from Roman Catholics (63.4 million) and Southern Baptists (19.9 million) to Mormons (5.3 million), Buddhists (1.1 million), Wiccans (307,000) and Baha’is (84,000). The statistics are often contradictory, sometimes wildly so. Recent counts of American Muslims range all the way from 1.1 million to 7 million.

Though it isn’t part of the US Census, counting religious folks is big business. Church leaders want to know how well their message is playing. Politicians and marketers want to see whether beliefs translate into voting and economic behavior. And whether you’re an evangelical Christian or a Muslim immigrant, the best way to get your voice heard is to demonstrate there are a lot of you. Even the atheists are getting into the act. In a recent New York Times Op-Ed, the philosopher Daniel C. Dennett urged his 27 million fellow atheists, agnostics, and religiously unaffiliated Americans (to cite the results of a 2002 poll) to come out of the closet and assert their nonbelief. He even proposed a new name, “brights,” for this fledgling demographic.

One of history’s first recorded censuses, explains Leonard Saxe of the Cohen Center for Modern Jewish Studies at Brandeis, was taken by the Jews as they were leaving Egypt. But today, the field of religious demography is fraught with controversy–over polling methods as well as over the complicated matter of just how to categorize people religiously in the first place.

Though many different definitions are used, most academics follow the United Nations in defining a believer simply as someone who identifies themselves as belonging to a particularly religion. So explains Preston Hunter, a computer programmer who compiles data on “over 4,200 religions, churches, denominations, religious bodies, faith groups, tribes, cultures, movements, ultimate concerns, etc.” on his independent website Adherents.com.

Of course, the simplest way for churches to count their members is just by noting who shows up. The Roman Catholic Church, for example, conducts an annual “October Count,” in which ushers go up and down the aisles tallying attendance for several weeks in a row. (Some priests just submit a list of Catholics living within parish boundaries instead.)

Protestant churches often base their figures on the number of people who have filed letters of membership at a church. Frank Newport, editor-in-chief of the Gallup Poll, argues that this can lead to overcounting, because churches–he cites Baptist churches in particular–fail to purge their rolls as members move.

But according to Bill Merrell, the vice president for convention relations of the Southern Baptist Convention Executive Committee, his organization’s count is “very specific” and very accurate. The 16,247,736 number from last year’s census, he says, was arrived at using “20 to 30 different factors,” including the number of people in the pews, Sunday school attendance, baptisms, and so on.

Of course, counting believers in a religion that doesn’t require group worship or public rituals can be especially tricky. This may partly explain the controversy over the 1.5 million figure for American Muslims published in September 2000 by the Glenmary Research Center, a Roman Catholic organization. The Council on American-Islamic Relations, which later released its own figure of 7 million, said the count didn’t adequately include those–especially women–who practiced the faith but didn’t attend services, or who simply came from a Muslim background.

The telephone surveys favored by independent polling organizations have their own problems. According to Brandeis’s Leonard Saxe, the proliferation of blocking devices, cell phones, and do-not-call registries means fewer people are talking to pollsters. Frank Newport of Gallup notes that his organization’s telephone surveys will not pick up people who don’t speak English. And in the case of the smallest religious groups, it’s hard to get a big enough sample size.

Whatever their accuracy, the numbers are put to many practical uses. The Center for Applied Research in the Apostolate at Georgetown University. Catholics, she explains, “tend to be well-educated relative to the general population and relatively affluent, too.”

Political pollsters are also paying attention. “Religion is the most important factor in the political world,” says Newport. A Pew survey just after the 2000 election, for example, revealed that in states that went for Bush, 70 percent of the population said they believed in the statement, “We will all be called before God on Judgment Day to answer for our sins.” In states that went for Gore, only 50 percent said they believed it.

Of course, as we have seen, counting voters may be only slightly less complicated than counting believers.

Naomi Schaefer, an adjunct fellow at the Ethics and Public Policy Center, is writing a book on religious colleges.

Comments are closed.