JUDGING from American network television and movies, almost no one in the United States attends church or believes in God.
When was the last time you saw a character go to church, turn to a pastor, or pray to God? Indeed, when ``L.A. Law'' this season added a fundamentalist Christian lawyer to the cast, it was so unusual that several newspapers and magazines took note.
Now, once again, social science has stepped in to show that, when it comes to religion, the American media have got it wrong - seriously wrong.
The United States is a unique society, manifesting high levels of economic development, education, and religious belief. This conclusion, reported in detail by such researchers as George Gallup Jr. and the Rev. Andrew Greeley, is reaffirmed by the National Survey of Religious Identification (NSRI). The survey, conducted in 1990 by the Graduate School of the City University of New York, is discussed in a new book, ``One Nation Under God,'' by Barry Kosmin and Seymour Lachman (Harmony Books, 312 pp., $25).
The telephone survey polled a sample of 113,000 people over a 13-month period across the continental US. According to the authors, it is ``the largest and most comprehensive poll ever on religious loyalties, and the most accurate and detailed as to geographical distribution.''
In answer to the National Survey of Religious Identification question ``What is your religion?,'' 86 percent of respondents said they were Christians: The two largest groups were Roman Catholics, at 26 percent, and Baptists, at 19 percent. Jews represent 2 percent of the population, with Muslims, Buddhists, and Hindus numbering no more than 0.5 percent each. Of those polled, 7.5 percent said they had no religion, and 2.3 percent refused to answer.
The survey found that many more Americans claim to be Catholic, Baptist, Methodist, Episcopalian, Lutheran, Presbyterian, and Unitarian-Universalist than membership figures of those denominations show. This indicates that many Americans who are not members of those churches, and who may not even attend, still identify themselves in those traditions. On the other hand, the numbers of Mormons, Jews, Muslims, and Eastern Orthodox Christians were found to be less than the claimed membership of those bodies.
One can understand little about the US without understanding its religious diversity and how that plays out in the national makeup and political decisionmaking. For example, membership statistics on a county-by-county basis compiled by the Glenmary Research Center in Atlanta showed that there are geographical regions in which one religion predominates. The NSRI likewise finds ``heartlands'' for four religious families: Baptists in the South, Lutherans in the Upper Midwest, Roman Catholics in the Northeast, and Mormons in Utah and other Rocky Mountain states (see maps, Page 11).
Methodism (including the mostly white United Methodist Church, the three historically black Methodist denominations, and several smaller bodies) is a more ``national'' religion, but Methodists are more likely to live east of the Mississippi River. ``The Methodists, and the other religious denominations that were dominant in the Colonial period, tend to be well distributed now across most of the states,'' Kosmin and Lachman write.
Among other groups:
r People who say they are Jewish by religion are mostly found in southern New England, the Middle Atlantic states, Florida, and California.
r The largest percentage of Pentecostals are found primarily in Arkansas and the states surrounding it.
r A higher percentage of nonreligious people is found in the West (see map, Page 11), but the West is also the most religiously diverse region of the country.
Ethnicity and religion
Ethnic and religious stereotypes abound in the US, but are increasingly off the mark. The survey finds that most Americans who claim Irish or French descent are not Catholic, and most Arab-Americans and Asian-Americans are Christians. Only 66 percent of those who claim Jewish origin are Jewish by religion; 12 percent are Christians.
Kosmin and Lachman conclude that as white Americans of different ethnic groups continue to intermarry and switch religions, ``religious identification will increasingly come to be viewed much more as a personal than as a family or group characteristic. Such individualism, for good or ill, contributes to the Protestantization of all religion in America.''
Previous surveys have shown that black Americans have higher church membership and attendance than whites. The NSRI shows that 50 percent of blacks are Baptists (see chart, right). Only 9 percent claim to be Methodists, well below historical figures, but the same percentage say they are Catholic, a larger number than previously documented. While about 40 percent of US Muslims are black, only about 2 percent of the black population is Muslim.
Among Hispanics, 66 percent are Catholic, while 25 percent say they are Protestant (see chart, right). Kosmin and Lachman cite other studies to indicate that Protestants have made inroads mostly among Hispanics of Puerto Rican and Central American origin. American Indians are 46 percent Protestant and 21 percent Catholic. About 47,000 out of 1.8 million native Americans are practicing traditional Indian religion.
Religion and politics
The links between religion and voting patterns have changed from traditional patterns. George Bush was elected president in 1988, according to CBS and ABC exit polls cited by Kosmin and Lachman, because he won 80 percent of the born-again, white-Protestant votes and 60 percent of mainline Protestant votes. But he got about 8 to 10 percent less of the Catholic vote than his predecessor, Ronald Reagan, and won less than 30 percent of the Jewish vote, even though Republican candidates had won 30 to 40 percent of that vote since 1972.
In 1992, Bill Clinton beat Mr. Bush in the popular vote 43 to 38 percent, with Ross Perot receiving 19 percent. While Bush got 52 percent of the Catholic vote in 1988, he got only 36 percent of it in 1992, with 44 percent going to Mr. Clinton and 20 percent to Mr. Perot. Bush's percentage of the Jewish vote fell to 12 percent, with 10 percent going to Perot and 78 percent to Clinton.
But Bush's biggest problem, according to research cited by the authors, was with white Protestants, who make up 60 percent of the voting population: He fell from 66 percent in 1988 to 46 percent in 1992, with 33 percent going to Clinton and 21 percent to Perot. The combination of these percentages cost Bush Ohio, Michigan, Pennsylvania, Illinois, and Missouri, Kosmin and Lachman write. Certainly the first all-Baptist ticket of Clinton and Al Gore Jr. had an effect on the results.
The NSRI finds that traditional Democrat-Catholic, Republican-Protestant affiliations are blurring. Protestants were still 9 percent more likely to vote Republican in 1988 and 1992, but ``there are growing differences between the denominations.'' (See table, left.)
``The historic and overwhelming support for Democrats has largely disappeared from the nation's Catholic population,'' Kosmin and Lachman write. Catholics are 7 percent more likely to be Democrats than Republicans, but ``white Catholics now favor the Republicans.... Among whites, the only remaining loyal bastion for the Democrats is Irish Americans....'' This last point will come as no surprise to residents of Massachusetts or New York.
Blacks are the exception to the close Republican-Protestant correlation. The Church of God in Christ, a black denomination in the Holiness tradition, is the most partisan Democratic denomination (see table, left). (Holiness churches teach the possibility of a second ``sanctifying'' religious experience after conversion that allows the individual to live a sinless life.) Black Baptists, Catholics, Methodists, Pentecostals, and Seventh-Day Adventists are all overwhelmingly Democratic, whereas their white coreligionists tilt towards the Republicans.
Gender, age, and culture
The survey confirms that women tend to be more religious than men. The NSRI showed that 87 percent of American men said they are religious believers; 83 percent are Christians and 56 percent are Protestants. Among women, 92 percent identify with a religion; 89 percent are Christians, and 61 percent are Protestants.
The gap is far larger among African-Americans: 52 percent of black Catholics are women, 56 percent of Episcopalians, 58 percent of Baptists, 60 percent of Methodists, all the way up to 69 percent among Holiness denominations. Among non-Christians the reverse is true: A slight majority of Jews are men; men with no religion outnumber women by a 2-to-1 margin.
``There is little evidence that Americans have abandoned religion during the course of the 20th century,'' Kosmin and Lachman write. But they did find significant denominational shifts among generations. Based on median ages, the ``oldest'' denominations are the mainline Presbyterian, Methodists, Episcopalians, and Lutherans. The youngest are Pentecostals, Catholics, and Mormons.
Mainline churches appear to be losing ground among younger age groups, as do the Eastern Orthodox, Seventh-Day Adventists, Jews, and Unitarians. Baptists, Pentecostals, and Jehovah's Witnesses appear to be gaining, as are Islam and all Eastern and alternative faiths.
``Catholicism is more than twice as prevalent among the young than the very old,'' the authors write. But this is due as much to defections among older Catholics as to the immigration of young Catholics to the US. The NSRI data confirm other research that finds that Catholic Americans are the most likely to stray from their church.
Examining the phenomenon of interfaith marriages and religious ``switching,'' Kosmin and Lachman write that ``switching occurs most often between denominations similar in social class composition.'' They cite a 1992 Gallup poll showing that the ``winners'' in this process in recent years have been small Protestant denominations, while one of the biggest ``losers'' has been the Catholic church, which attracted only 1 in 11 switchers.
The survey showed that 78 percent of Americans live in religiously homogenous households. The most homogenous of all are those of Eastern religious immigrant groups and conservative Protestants. Comparing the findings with those of a 1957 survey, the researchers found that the religious homogeneity of Baptists was the same, while that of mainline Protestants fell between 4 and 7 percent. The percentage of Catholics living in one-religion households, however, has fallen by 9 percent, while that of Jews has fallen 25 percent, showing the effects of assimilation and what the authors term ``Protestantization.'' In interfaith households, ``Jews and Catholics are more likely to abstain from any religious identification than they are to adopt another religious position,'' Kosmin and Lachman say.
What it all means
The data, while useful and interesting, can only tell us so much. They indicate significant demographic changes that politicians, for example, will ignore at their peril. And they demonstrate that the image of America that the media often depict is deeply flawed. But there is no way that a survey can measure the depth of religious commitment or the expression of religious conviction in people's daily lives. As Kosmin and Lachman point out, ``public religious display does not necessarily reflect inner devotion among contemporary Americans.''
Indeed, although the United States is characterized by an overwhelming identification with religion, the authors write, ``it exists within a secular framework, an outer shell of secular values. For what we have witnessed in the latter part of the 20th century is the growing secularization of a self-described religious people.''
``There appears to be a collective schizophrenia whereby the public says one thing, with apparent sincerity, in answer to public-opinion polls and acts quite differently,'' Kosmin and Lachman write. Yet they find the secular and the religious in American society affecting and changing each other in ways that do not occur elsewhere.
Despite strong forces undermining the family, violence and sexuality depicted in the media, youthful challenge to authority, and the unwillingness of modern society to confront the phenomenon of death, Kosmin and Lachman also see ``a growing minority searching beyond the civil religion and beyond a loose identification with religion for a spiritual base, and questioning the direction of a basically secular society.''
Thus they say that religious interest will increase during the 1990s ``as we approach a new millennium with all its hopes and fears for the future.''