A nasty dispute recently broke out about the number of Muslims in the United States. One study (available online in pdf format) claims six to seven million. Funded by four Muslim organization and directed by Ishan Bagby of Shaw University, it counted Muslims by contacting mosques and using various assumptions. The American Religious Identification Survey conducted by Barry Kosmin and Egon Mayer through the City University of New York Graduate Center found at most three million Muslims using a telephone survey. About one-quarter of all Muslims make New York State their home, according to Kosmin and Mayer Kosmin is one of the leading students of religious demography, and is co-author of "One Nation Under God," based upon a similar study done 10 years ago. The American Jewish Committee commissioned Tom Smith to assess the available estimates and make one of his own. His independent estimate largely agrees with that of the CUNY Graduate Center team.
Who is right? Why is it so difficult to count the number of Muslims? First, the Census Bureau is actually prohibited from counting Muslims. Islam is a religion rather than an ethnicity or a national origin, and the bureau is barred from asking about religion. Muslims and supporters of civil liberties might be especially thankful for this at this particular moment in American history. In a notorious chapter of their history, the Census Bureau assisted the US government in rounding up the Japanese-Americans at the beginning of World War II by using the agency's very detailed block-by-block tabulations. At a recent conference, a Bureau official said that they had received over 150 requests from other governmental bodies for information about Arabs (which are considered an ancestry) and immigrants from Muslim countries.
Since Muslims, whatever their number, are a small part of the United States population they are hard to count. They are also hard to define. Considering just adults, what does one need to do to be counted as a Muslim? Go to a mosque? Pray several times per day? Observe Ramadan? Have made a pilgrimage to Mecca? Eat Halal foods? Those trying to count Jews face similar problems. One definition of Jewish, for instance, is "Feeling you should go to High Holiday services." By such a definition apparently, it is enough to either go to the services or to feel guilty about not going.
When Bagby released his study in April, many reporters noted that if his figures were right then Muslims outnumbered Jews in the United States. This relatively unimportant fact in April that might add to Muslim bragging rights over Jews took on added significance with the September 11th attack. The American Jewish Committee commissioned Smith, and Egon Mayer is head of CUNY's Center for Jewish Studies, so some suspicion has greeted the release of their reports.
However, when one compares Bagby's methods and assumptions with those of Smith and Kosmin, it is plain that Bagby's is severely and fatally flawed. He enumerates all American mosques (1209 he says) and then interviews mosque representatives from a sample of 631. His response rate is a respectable two-thirds. But plainly big well-organized mosques with a well-staffed office would be much more likely to respond than smaller less well-funded mosques with part-time office staffs. He finds that on the average more than 1,600 people are involved with each Mosque, and then multiples that figure by all 1209 Mosques, claiming that about 2 million are involved. He then asserts that two thirds of Muslims are missing from the mosques, so he multiplies by three and announces that there are between six and seven Muslims in the United States.
Kosmin and Mayer by contrast base their estimate on the 219 Muslims contacted, while they were phoning about 50,000 people. Weighting their estimate appropriately they find 1.1 million Muslim adults, which is double what they found in 1991. Including children this would make the Muslim population about 1.5 million. Since there is sampling error they admit that the total could be as much as three million. Smith by ransacking available data comes up with an estimate of 1.4 million, 1.9 million including children. Looking at standard errors, his upper bound is very similar to that of Kosmin and Mayer. An analysis that my associates and I did for the New York Times of immigrants or children of immigrants from Muslim countries comes up with similar numbers. (It is true, of course that some immigrants from Muslim countries are not Muslim, and other Muslim immigrants come from non-Muslim counties. Other Muslims are native born.)
It seems that Bagby, along with many "ethnic group researchers" is willing to bend every effort to increase the count of his group . Kosmin and Smith, both renowned and careful social scientists, do as much as possible to base their counts on established methods. But even they give the benefit of the doubt to their highest possible count.
No matter how many Muslim there are, all researchers agree that the Muslim community, as with the Arab community (discussed here in September) are doing quite well in America. They are highly educated, many are quite prosperous, and they occupy a wide variety jobs. But it is also the case that they are Americanizing. As one of my students, an American citizen and the son of the Bangladeshi Islamic religious leader, told me: "My brothers and I are very observant, but my nieces and nephews seem much more American. I don't know if they will all continue to be Muslim." So like other groups, being successful in the United States may increase the group's influence and prosperity, while ultimately diminishing its size.
Andrew A. Beveridge has taught sociology at Queens College since 1981, done demographic analyses for the New York Times since 1993, and provides expert testimony on a range of cases, including housing discrimination. The opinions expressed are his alone.