Washington, DC

In a noon conference call for journalists, Luis Lugo, director of the Pew Forum on Religion & Public Life, together with fellows John Green and Greg Smith, released the first of three reports on the Forum’s path-breaking U.S. Religious Landscape Survey, launched the interactive website accompanying the project and answered questions from reporters. This report, which focuses on religious affiliation, reveals an extremely fluid and competitive American religious marketplace and provides an in-depth look at individual religious groups and denominations.

The survey, principally authored by Green and Smith, involved 20-minute interviews with 35,000 American adults and is therefore one of the largest and most comprehensive studies of its kind. Because of the richness of the data, the Forum divided the analysis into multiple reports; analyses of beliefs and practices as well as social and political attitudes are planned for the spring of 2008. Go to the survey »

Luis Lugo, Director, Pew Forum on Religion & Public Life
John C. Green, Senior Fellow in Religion and American Politics, Pew Forum on Religion & Public Life
Gregory A. Smith, Research Fellow, Pew Forum on Religion & Public Life

Navigate this transcript:
Catholics and religious change
Will the unaffiliated return?
Atheism rates remain stable
Protestants becoming a minority
Evangelicals increase and diversify
Immigration and “new” faiths
Defining evangelicals

LUIS LUGO:Thank you for joining us for this first release of the U.S. Religious Landscape Survey. I am Luis Lugo, director of the Pew Forum on Religion & Public Life. Joining me are John Green, senior fellow, and Greg Smith, research fellow, who can be considered two of the principal authors of this report, both in the analyses and in the writing of this report.

We believe the unique combination of depth and breadth makes the U.S. Religious Landscape Survey a very valuable resource for understanding the changes taking place on the American religious scene, and how those changes relate to social and political values. In terms of breadth, we are talking here about a survey with more than 35,000 interviews, which allow us to analyze religious groups or subgroups as small as 0.3% of the U.S. population.

In terms of depth, there are over 40 religion questions on this survey, then a series of demographic questions and questions on social and political attitudes. Because we go so deep on religion, it means we not only are able to determine who is a member of an evangelical church, for instance, but whether it’s a Baptist church and not just whether it’s a Baptist church, but whether it is Southern Baptist, Free Will Baptist, Independent Baptist, et cetera. So it’s the breadth and the depth that we think makes this survey so unique.

This first report focuses on religious affiliation. Later in the spring, we will talk about religious beliefs and practices and social and political attitudes. I’m going to ask John to discuss the key overall findings of this first report, and then Greg will talk about some of the demographic groups, and what we found about them, as well as touch briefly on methodology.

In conjunction with the release of this report, we also rolled out a special feature on our website that presents the survey’s findings through interactive maps and dynamic charts and other tools that will help you analyze the findings on your own. Of course, we also have a PDF version of the report there. So we invite you to visit www.pewresearch.org/religion to do your own exploration of these very rich findings.

John, if I could ask you to talk first about the key findings.

JOHN GREEN:Thank you very much, Luis. My name is John Green, and I’m a senior fellow here with the Pew Forum on Religion & Public Life. As Luis indicated, this is just the first of several reports from the U.S. Religious Landscape Survey. This one focuses on religious affiliation, that is, the religious communities to which respondents say they belong. We believe religious affiliation is the best place to begin the analysis of this extensive survey because it provides a useful baseline against which we can then look at the demography of religious communities, their religious beliefs, their behaviors, and, ultimately, their political and social values.

This report shows religious affiliation in the United States is extraordinarily diverse and dynamic. The report, furthermore, shows this diversity and dynamism in unprecedented detail both because of the large number of people interviewed in the survey and because of the large number of religious questions employed. This report’s findings cover three major areas. First, it reports on the incidence of religious communities in the adult population with a great deal of precision. Second, it assesses changes in the respondents’ religious affiliation from their childhood. Third, it provides detailed demographic information on these religious affiliations including groups generally too small to be studied in even large surveys.

In terms of the incidence of religious affiliation in America, even the largest religious groups are relatively small. For example, this report finds that members of evangelical Protestant churches are the largest religious tradition in the country. But they account for only a little more than one-quarter of the adult population. Roman Catholics are the second-largest religious tradition, with a little less than one-quarter of the adult population. Filling out the top five religious traditions in the United States are members of the historic mainline Protestant churches, at a little less than one-fifth; people unaffiliated with organized religion, at about one-sixth of the adult population; and, finally, members of the historically black Protestant churches, at about one-twentieth of the adult population.

All other religious groups in the United States-and there are many other religious groups besides the top five-are much smaller. For example, this report finds that Buddhists make up about 0.7% of the population and Hindus about 0.4%. However, even the large religious traditions are characterized by a great deal of internal diversity. The Protestant traditions are made up of dozens and dozens of denominations, and the Roman Catholic community has considerable ethnic diversity. Even the unaffiliated population is highly diverse, and our survey documents that in great detail. The unaffiliated contain not just atheists and agnostics and people to whom religion is not important at all, but also people not associated with organized religion who, nonetheless, take religious seriously in respects other than affiliation.

This report documents a high level of change in religious affiliation over the lifetime of adult Americans. More than two-fifths of Americans have changed their affiliation in some respect since childhood and more than one-quarter have moved across the major religious traditions or into the ranks of the unaffiliated. This dynamism reveals the competitiveness of the religious marketplace in the United States.

One particularly good example of this dynamism is the experience of Roman Catholics. If everyone who reported having been raised Catholic had remained in the church from their childhood until now, then Catholics would account for almost one-third of the adult population, instead of a little less than one-quarter. Thus, Catholics have failed to retain many of their childhood adherents, and this survey shows roughly one-tenth of adult Americans are former Catholics.

Catholics, however, have been able to maintain their overall share of the religious marketplace because of extensive waves of immigration that continue to this day. The ranks of the unaffiliated have grown substantially as well, roughly doubling the proportion of the population raised unaffiliated. However, the unaffiliated also show a very high rate of change, with more than half of those raised outside of religion now reporting a religious affiliation.

Finally, our extensive demographic questions reveal that demography adds another level of diversity to religious affiliation. I’d like to highlight a few of those findings before turning our comments over the Greg Smith, who will talk more about these matters as well as methodology. Our survey found nearly two-fifths of married respondents have a spouse with a different religious affiliation. Mormons and Muslims have the largest numbers of children living at home. Age is also a very important factor when it comes to religious affiliation. Members of the historic mainline Protestant churches and members of the Jewish community are older than many other religious groups. The unaffiliated count among their ranks a larger number of young people.

Among racial and ethnic groups, African Americans and Latinos are the most likely to report a religious affiliation. And, finally, a finding appropriate for an election year: The Midwest region of the United States most closely resembles the religious diversity of the U.S. as a whole. Southerners tend to be more evangelical, Northeasterners more Catholic, and people in the West are more likely to be unaffiliated. Greg?

GREG SMITH:Thank you. I am Greg Smith, research fellow at the Pew Forum on Religion & Public Life. I’m going to talk briefly about a couple of the methodological features of this survey and then touch on just a couple of the details our survey reveals. To reiterate what Luis and John have already said, there are really two methodological features to this project that provide the unique opportunity to learn new things. The first is its very large sample size, with over 35,000 interviews conducted among a representative sample of U.S. adults. The second is its over-samples of three important groups: Buddhists, Hindus, and Orthodox Christians.

These features allow us to take a very precise look at the religious makeup of the U.S. The overall margin of error for results based on our full sample, for instance, is less than 1%. These features also make it possible to take a detailed look at many smaller religious groups, down to groups that account for roughly 0.3% of the U.S. population. This cannot be done with most surveys, which are based on smaller numbers of interviews.

These features and methodological strengths reveal some fascinating details about the demographics of a variety of religious groups. We find, for instance, that Hindus stand out compared with other religious groups for their extraordinarily high levels of educational attainment. Nearly half of Hindus in the U.S.-48%-have obtained post-graduate education over and above earning a college degree. That means more than four times as many Hindus have reached this level as compared with the public overall. We find that, in contrast to Islam and Hinduism, Buddhism in the U.S. is primarily made up of native-born adherents, whites, and converts. Only one-in-three American Buddhists describe their race as Asian while nearly three in four Buddhists say they are converts to Buddhism.

The landscape survey also makes it possible to look in detail at many individual Protestant denominations. Here again, doing so reveals many interesting things. The survey indicates, for instance, that more than one-in-10 members of the Church of God in Christ, the largest of the Pentecostal denominations within the historically black church tradition, are white and another 13% are Latino. We document revealing age patterns across denominations as well. Members of evangelical and charismatic non-denominational churches, for instance, tend to be young compared with the public overall.

Members of several mainline denominations, as John mentioned, by contrast, tend to be relatively older. Roughly six-in-10 members of the Episcopal Church in the USA, for instance, are over the age of 50, compared with only four-in-10 Americans overall who are over 50. These and many other findings are documented in detailed tables at the end of our report where we provide extensive demographic information on the 14 largest religious traditions in the U.S., 12 large Protestant denominational families, and 25 individual Protestant denominations. There you will also find estimates of the religious makeup of most of the states in the continental U.S., which can also be accessed via the new interactive mapping tools available on our website.

We hope these details will help fill in some of the blanks in our knowledge and understanding of religion in the United States.

LUGO:Thank you, Greg and John. I know there are a lot of you waiting in line so let’s get to your questions.

JEFFREY WEISS, DALLASMORNING NEWS:A bookkeeping question first: How many adult Americans are there in the United States, based on your sample?

LUGO:Two hundred and twenty-five million.

WEISS:I want to drill down on the unaffiliated. Other surveys have shown young people tend to be less affiliated, and then they affiliate as they get older. Do you have any sense as to whether you are looking at a cohort and, as these people get older, they are going to become more affiliated, or whether America really is becoming less affiliated?

GREEN:That’s a very good question. There’s no doubt the unaffiliated disproportionately have young people among their ranks. That pattern has been common for quite some time. There is some indication in our survey, as well as in some other surveys, that the number may be somewhat larger than in past generations, and there is some suggestion this generation may not return to religious affiliation at the same rate as the baby boomers and other previous generations.

We cannot address the future, of course, with this survey, but it does suggest the question of the potential return of younger people who are currently unaffiliated to the ranks of the affiliated is something we’ll have to pay a great deal of attention to. Given the large percentage of young people among the unaffiliated, this could have a profound effect on the character of American religion.

LUGO:Just to underscore what we say in the report, however, remember the ranks of the unaffiliated-and this is true for those under 30-include a very significant percentage who may not be affiliated with a particular religious group, but who nevertheless tell us religion is somewhat or very important in their lives. So you see a double pattern here: both a growth in the ranks of the unaffiliated and a growth in the ranks of the religious who are not affiliated.

WEISS:One quick follow-up. Other surveys have indicated that of those who answer “none of the above,” a substantial percentage also say they believe in God. Now, maybe I’m jumping your next report, but do you have any sense as to what percentage of the unaffiliated not only say religion is important or not important, but actually have something you might consider religious belief?

GREEN:You are jumping to our next report, Jeff. But, based on other surveys, there are likely to be a substantial number of people in this group of the religious unaffiliated who believe in God in one form or another.

SMITH: One of the indications we do have is a question where we asked people how important they consider religion to be in their own lives. We find more than a third of the unaffiliated population-about 6% of the U.S. population overall-consists of people who say they have no particular religion but nevertheless find religion to be at least somewhat important in their own lives.

LUGO:Among the ranks of the unaffiliated, only about 1.6% of the public as a whole describe themselves as atheists. If you add the agnostics to that-another 2.4%-that’s 4%of the population as a whole. So clearly the 16%that is unaffiliated contains a large number of people who have some relationship to religion, some more than others. In the next report, when we talk about religious beliefs and practice, we will of course tease that out extensively.

ADELLE BANKS, RELIGION NEWS SERVICE:I was wondering about a figure Greg mentioned: that more than one-in-10 of the Church of God in Christ are white and another 13% are Latino. I’m just wondering if that was the most striking example, and if there are another couple of examples like that, as far as whites in traditionally black denominations or the other way around.

SMITH:I’m not sure about other particular findings that jump out at me, but one of the things-

BANKS: Can you tell me if I heard that one right, though?

SMITH:You did hear that one right, that more than one-in-10 members of the Church of God in Christ are white and a similar number are Latino. You can find similar information on the racial and ethnic composition, as well as the age and family size and geographic distribution of members of a couple of dozen of Protestant denominations, in our detailed tables.

LUGO:But the broader point is when we talk about members of historically black churches, we’re not just talking about African-American members. Those historically black churches have a significant number of whites and Latinos as part of their membership.

DAN HARRIS, ABC NEWS:First of all, when is this next report coming out?

LUGO:We don’t want to upstage the Pope’s visit in mid-April so probably late April or early May, Dan.

HARRIS: And it’s based on the information you’ve gathered from this polling?

LUGO:Yes. We were going to do one report, but we sat down with this, it was just overwhelming, so we said, “Maybe two reports.” It continued to be so overwhelming we said, “In order for us to digest and analyze it, and then give it to you folks for you to digest and analyze, we really have to go to three reports.” There’s simply so much data. Keep in mind what John said: A good-sized survey is one-twentieth the size of what we’ve got here. So this is an awful lot of data, with many, many nuggets and angles. To be honest with you, even in terms of affiliation, we can just hit the highlights. There’s a lot of interesting stuff we haven’t even been able to surface yet.

HARRIS:Let me ask you a couple of questions about affiliation, and you can go as far as you want. One of the things our internal polling people have said to me all along is, “Notwithstanding the very highly publicized success of atheist manifestos on the bestseller list, the number of those who describe themselves as either atheist or agnostic hasn’t moved much.” It seems your polling would support that assertion, correct?

GREEN:Yes, it would. The numbers of atheists and agnostics has been very stable for quite some time. Our survey, of course, is able to look at them in more detail because the sample is so large. What we see is a lot of interest in books written by atheists and about atheism, but not necessarily a dramatic jump in the number of people who identify that way.

HARRIS:Do you have a sense of why young people are increasingly unaffiliated?

GREEN:There are two hypotheses, both of which have a good bit of truth in them, though of course the answers vary from individual to individual. One hypothesis is when young adults leave home they are busy figuring out a lot of things about their lives. They are trying to figure out their careers, their family situation, where they want to live and so forth. In that situation, many people-even people who are quite religious at the level of belief-stray away from organized religion. So part of this is a lifecycle effect, that younger people are in that transition part of their lives, and that transition includes religion.

However, there may be features of the current generation that have led to a higher number of unaffiliated, including the stress that has been put on major religious denominations as part of that change, and also the greater number of options available to young people given the great diversity of American society. So, in addition to lifecycle effects, there also may be cohort effects. There may be something peculiar to the last decade or so that has made it more likely for younger people to be unaffiliated.

CATHY LYNN GROSSMAN, USATODAY: I’m going to throw in three questions, some of which are very quick. First, why is there no data on Alaska and Hawaii in the report?

SMITH:Unfortunately, we did not conduct any interviews in Alaska or Hawaii; like most surveys we restricted ourselves to the continental United States. So we’re not able to speak directly to the religious characteristics of those states.

GROSSMAN:I find your statistics pretty much in keeping with the statistics about people who believe in God. Usually about 90 to 92% of Americans say they believe in God or a higher power. In a recent LifeWay study, it was down to 72% saying God or some kind of higher power exists.

Apart from the atheists and the agnostics, as I understand it, you found 6.3%saying not only do they have no affiliation, but religion is unimportant to them; they would be in the “this is just not something they think about” category. So that adds up to about 10% of people who are unaware of or uninterested in or have decided against God. I don’t think that necessarily contradicts other studies that show 90% of Americans believe in God, or does it? Am I missing something there?

GREEN:No, this doesn’t necessarily contradict other studies. Of course, what we’re measuring here is affiliation, not beliefs. There may be a few people who have affiliations but don’t believe, just like there are some people who lack affiliations and probably do believe. There’s nothing inconsistent between these overall patterns and what other people have found, but we will be able to look at them in much more detail.

GROSSMAN:You were talking about young people earlier. Other people who study Catholic data are finding, contrary to the old idea that people grow up and “grow into” church, people are not growing into church. If they do grow in, they wash back out again as they age out-after their kids have been confirmed. Neither are they seeing older people suddenly being overwhelmed by spiritual thoughts as mortality looms. What can we look at in your study to help us trace whether there is a lifecycle effect under way?

GREEN:Our survey, like all other surveys of this sort, is really a snapshot or portrait of a particular moment in time. So we can’t look at lifecycles occurring. We can only look at it cross-sectionally. It’s much easier to look at the earlier part of the lifecycle because we can look at young people.

We do know the dynamism in religious affiliation we are reporting overall masks a great deal of other movement we can’t get at. It’s not just that people change from their childhood religious affiliation. In fact, they may change affiliations several times and, unfortunately, our questions were unable to get at those particular items. Greg, do you have the age distribution in front of you there?

SMITH:One thing our report shows that speaks perhaps indirectly to this notion of change going on across the lifecycle is we find young people and older people are more or less equally likely to have experienced a change in religious affiliation. However, there are some important differences about the nature of that religious change across age groups.

If you look at Americans over the age of 70, for instance, 40% of them have changed their religious affiliation since childhood. But that 40% is about evenly divided between people who have made a change across traditions, that is to say from Protestantism to Catholicism or from Catholicism to being unaffiliated, while the other half of them who have changed religion have changed within a tradition. They have gone from being members of one Protestant denominational family to belonging to another Protestant denominational family. Perhaps they were Baptist, and now they’re Methodist, for instance.

If you look at people under the age of 30, about the same number have experienced a religious change, but they are much more likely to have changed across traditions. By a three-to-one margin, in fact, young people who have changed religious affiliation have changed across traditions rather than within a tradition.

LUGO: As we say in the report, we hope to go back in the field every several years to administer this survey again, at which point we can begin to make these various snapshots into a movie; we can begin to see change over time.

On this question of people switching and converting and what’s driving all of that: We had to keep the survey interviews to no more than 20 minutes in order to maximize the response rate. But we are inclined, with the support of our funders, the Pew Charitable Trusts, of perhaps going back into the field at some point when all of the dust settles and re-contact some of these folks and delve a bit deeper into the phenomenon of this high rate of change. So perhaps we can find some answers to those questions, including those about young people, before the next eight years when we do another similar survey.

Let me just underscore again: One of the significant findings on the unaffiliated is the high percentage for whom religion is still important. It’s about 6% of the U.S. public as a whole. One question we’ll want to look at going forward is whether we’re seeing in the United States a phenomenon quite common in Europe, which British sociologist Grace Davie has labeled, “believing without belonging.” Many folks in Europe, where religion has become deinstitutionalized, still have certain religious beliefs. One key question going forward is whether the ranks of those who believe but do not belong will continue to expand. Again, we’ll need to wait for the movie version on that one.

JOHN PARKER, THE ECONOMIST: To a foreigner, the thing that leaps out is not so much the unaffiliated, but the finding that the Protestant tradition in America is coming down toward a minority, or 51%, which I think would be the first time in 200 years. I have two questions about this. Can you tell from the other findings, what is the source of the decline of the Protestant traditions? Is it mostly the mainline traditions? Which particular one? Are certain people leaving in their middle age? Is it that young people who are joining a church are not joining the Protestant churches? Can you give us any more details about this decline of Protestantism in America?

Secondly, I know this is not so much the subject of the report, but can you speculate about what this would mean in the grand scheme of things? As I said, this would be the first time America had become a minority Protestant place ever. That seems very striking to me.

GREEN: Most of the decline among Protestants has been among mainline Protestants. Correct me if I’m wrong, Greg, but it seems to be pretty uniform across the different elements of mainline Protestantism. Some portions of the Protestant community are actually growing. There are evangelical churches, for instance, that are gaining adherents. So, there is a mixed pattern of decline and increase that averages out to a decline overall.

Why have mainline Protestant churches declined? There are all kinds of theories and explanations for that, some of which we can’t address in this survey. But there are two findings we can address. Part of it appears to have been a decline in birth rates over time, and another seems to have been an inability to keep people who were born into those churches in those faiths, with a large number of people who were mainline Protestant reporting moving to the ranks of the unaffiliated. You might think of this as family problems, both at the level of having children but also of raising children.

Of course, it could very well be that many people in mainline churches are proud of those patterns because it speaks to a certain freedom and tolerance and less concern with maintaining the cultural dominance that mainline Protestants once held. I think the continuing decline of the size of Protestantism is very important for the American culture and politics. So many values and institutions in American public life came out of Protestantism, particularly mainline Protestantism. As those groups are replaced, we’re likely to see a change in those institutions and in the cultures that support them. That doesn’t necessarily mean the change will be bad-it might in fact be good in some respects-but it is certainly going to be very different than it has been in the past.

But it is important to not focus too much on Protestantism as a whole because there are enormous differences within Protestantism, between the mainline and the evangelical churches, between the African-American churches and the white churches and now, of course, the growing portion of Protestantism that is Latino. So Protestantism is not just losing influence as a whole; it’s also losing influence because of its internal divisions.

LUGO:Let me just add to that because, John, you mentioned Latinos. I think immigration is a very important aspect of this shift in the balance. Our survey documents the vast majority of immigrants to the United States-over 70%-are Christian. In that sense, the pattern is somewhat different from the immigration pattern to Europe, let’s say.


But the balance within that Christian community is very different among immigrants than it is among the native-born. Among the native-born, Protestants outnumber Catholics by more than two-to-one. Among immigrants, Catholics outnumber Protestants two-to-one. So clearly, even though immigration is by and large confirming the essentially Christian social nature of the American people, it is helping to tilt the balance toward Catholicism.

Again, a large number of those are Latinos, so Latinos now comprise one-third of the U.S. Roman Catholic Church. Among Catholics under 40, nearly 50% are Latino. So we’re seeing the Latinization of the American Catholic Church, which in some sense presages-if you look at demographic projections-what the United States as a whole is going to be. One of our sister projects released a great report a couple of weeks ago looking at census figures and doing demographic projections. It’s not only that the United States will become increasingly less Protestant; it’s that the United States will become increasingly less white. The big pattern there is among Latinos. Today, they comprise 14% of the U.S. public. According to the demographers’ projections here at the Pew Research Center, that number will go to 29% by 2050. Asians will also grow to about 10% during that time, and African Americans will remain basically steady at 13%. So it’s not just religious change: It is ethnic change closely related to religious dynamics.

Greg, did you want to add anything to that?

SMITH:If you look at people who were raised in largely evangelical denominational families, and who have left their childhood denomination for another denomination, most of them are leaving for other evangelical denominations. But if you look at people who were raised in largely mainline Protestant-denominational families and have left for other Protestant denominations, they’re more or less evenly divided between people who have become evangelical and people who have remained in the mainline. That speaks to the relative decline of mainline Protestantism compared with evangelicalism.

CONNIE KANG, L.A.TIMES:Following up on the decrease of Protestants in the United States: What was the highest point? Do you have figures on the time when the United States had the highest percentage of Protestants? As I understand it from the report, the decline took place in the ’90s. Is that correct?

GREEN:It’s difficult to know for sure what the high point was for American Protestantism because we don’t have the same measures going all the way back. We only have good survey data for about the last 60 years so it’s difficult to know. But as late as the 1980s, Protestants were two-thirds of the American population in most surveys. So a lot of the decline we’ve seen has been fairly recent. Now, it may very well be that the two-thirds in the 1980s was substantially lower than some higher figure in the 1890s; there may have been a long and steady decline of Protestantism in the United States, but for most of that period, the decline would have been very small on a year-to-year basis. It’s just in the last couple of decades we’ve seen this really sharp decline.

LUGO:The last time the U.S. Census Bureau asked a religion question, which is in 1957, it showed also, John, the Protestant population at about two-thirds. So clearly from the mid-50s to about the early ’80s, it remained fairly constant at two-thirds, according to most work that has been done on this. But since then, particularly in the last decade or two, we have seen a significant decline.

MICHAEL PAULSON, THE BOSTON GLOBE:What do you know about where the Catholics who left went, this 10% of Americans who are ex-Catholics? Is there some trend on where they are now?

LUGO:I can tell you generally-and Greg can give you the exact percentage-they split roughly in half, with a little bit more tilting toward Protestant denominations, and more specifically evangelical Protestant denominations. The other half, or actually slightly less than half, go to the unaffiliated. So in sheer numbers, former Catholics are populating both the ranks of the unaffiliated and the ranks of Protestant churches. One-out-of-10 members of evangelical Protestant churches, for example, are former Catholics.

We’re also seeing this trend, incidentally, among Latinos, where a good percentage, close to 20%, leave the Roman Catholic Church. That’s at a lower rate than among non-Latino Catholics, but it’s still a significant number. We know the majority of them tend to end up-or at least a plurality end up-in evangelical churches. Greg, you have some specific numbers?

SMITH:We find 68% of everyone who is raised Catholic is still Catholic, which means 32% of people raised Catholic are now something else: 15% are Protestant, including 9% who are evangelical; 5% are members of mainline churches; and 1% are members of historically black churches. Four percent of people who were raised Catholic have changed to some other faith. This would include Judaism and Mormonism: basically any other non-Protestant faith. Fourteen percent of everyone who was raised Catholic is now unaffiliated with any religion.

LUGO:To put that retention rate in context-you mentioned a third of those who were raised Catholic told us they were no longer Catholic-That rate, aside from some of the smaller, largely immigrant religious groups, is by no means on the high end of churches losing their membership. In fact, the group that has lost the highest percentage is a group that, quite frankly, surprised many of us. It’s the Jehovah’s Witnesses. Close to two-thirds of those raised as Jehovah’s Witnesses told us they’re no longer Jehovah’s Witnesses, which may give you an indication of perhaps why they’re so zealous in their recruiting effort: On the retention side, they are suffering more than anybody else.

Incidentally, getting back to the unaffiliated, they don’t do very well either at retaining “members.” Over half of those we interviewed who told us they were unaffiliated growing up have now joined a religion formally. So they don’t tend to stay in that category either. Everybody in this country is losing members; everybody is gaining members. As we commented, there are some net winners and net losers, but it’s a very competitive marketplace, and if you rest on your laurels, you’re going to be history.

PAULSON: Luis, what differentiates this study from other studies? Is it fair to say this is the largest since the Census got out of the business? Is there something that makes this study more reliable than other research we’ve seen over the last half century?

LUGO: I think it’s the combination of the breadth-over 35,000 interviews-that places this study in a very small circle, but it’s also the depth. If you look at the Census, and the way it asked about religion: In fact for most of the 19th century, they didn’t even ask individuals about their religion. They asked about denominations, and how many ministers and all of that. It was valuable information, to be sure, but it was not the affiliation of Americans. So when they did ask about religion they didn’t go very deep; it was just a handful of questions. We ask more than 45 questions on religion here.

So we believe it is the combination of the breadth and the depth of this survey that makes it unique. John?

GREEN:This is one of the largest surveys of religion ever done, but it is not the largest. In 2000, the American Religious Identification Study, which was done at CUNY, had 50,000 respondents. We owe a great debt to other researchers who have done other similar studies, both in terms of the size of their samples and the depth of their questions. As Luis pointed out, we are trying to fill a niche in between, where we have large sample size but also a significant number of religious questions. The tendency is to go one way or the other: either to interview just a few people and ask them lots of questions or to interview a lot of people and just ask them a few questions. So our survey is unique but it does build upon the strengths of other surveys.

PAULSON: Greg, you mentioned earlier the high age of Episcopalians. Are there any other quirky findings that would differentiate the mainline Protestant denominations?

SMITH:The one that jumped out at me when I was perusing it this morning was the Anglicans, which in this survey we’re able to differentiate from Episcopalians. Anglicans also tend to be older than the population as a whole. But that pattern is true across mainline denominations. Mainline Protestants, overall, tend to be older than the general population, and that’s particularly true of Anglicans and Episcopalians.

The other group that stood out in our analysis in terms of being older than the general population and older than other religious groups are Jews. They have about the same age distribution as mainline Protestants.

ED STODDARD, REUTERS:I’m interested in the seeming increase of evangelicals. The National Council of Churches says there are more Catholics than evangelical Protestants in America by their count. Your survey shows evangelicals outnumbering Catholics. I was wondering: Is this a first? And for John Green: Is the evangelical movement still growing, and what would be the political implications of that?

GREEN:There are different ways to measure religion, and this is a survey-based measure. Our findings, while much more detailed than other surveys, are not inconsistent with what other people have found. The Handbook of U.S. and Canadian Churches, a very valuable resource put out by the National Council of Churches, bases their findings on official denominational figures, which are valuable as far as they go but are very different from survey-based measures. Lots of small evangelical denominations, for instance, are not picked up in these surveys of denominations, maybe because some of those denominations don’t keep any records or don’t want to participate.

All these measures are useful, and, when put together, they give us a good picture. But what survey-based measures can tell us is what people think they are, which has a lot of influence on how they behave.

It does appear to us the evangelical Protestant tradition is still growing. But part of what’s happening as it’s growing is it’s becoming more diverse. The growth may give evangelicals increasing clout in politics, but because of their diversity, that clout may not be used in exactly the same way. What we’ve read this year about the different perspectives of evangelical leaders may be the wave of the future. Evangelicals may not be as united politically as they have been in the recent past, precisely because of their success in attracting new members.

LUGO:Many of the trends we pick up more broadly are also at play within evangelicalism. You might have noted in the report a growth segment of the evangelical community are those who belong to non-denominational churches. We also had a significant number of evangelicals who did not identify with a particular denomination. They might have said, “I’m just a Baptist.” So the loosening, as it were, of denominational identity and perhaps denominational loyalties are at play within evangelicalism just like they’re at play in the larger American religious scene.

BILL TAMMEUS, KANSAS CITY STAR: John, you mentioned earlier you can’t use this survey to predict the future. But I wonder if the three of you, based on other surveys and your own knowledge, can look ahead a little bit, particularly on this question as to whether by the end of this century, most Americans will not be Christian. My second question is: Can you determine what is the fastest growing religion in America today?

GREEN:It is always hazardous to look toward the future. But it seems we do see in these data a great capacity for change. The diversity and dynamism of American religion may feed upon itself. Because there are more groups to move to and from, we may see a lot more motion; because of the continuing importance of immigration, we may see changes in the ethnic character of American religion. Perhaps because of the growing size of the unaffiliated, we may see a decline in the number of Americans who are religious in some broader sense of the term.

I think we can safely say American religion is likely to be even more diverse in the future than it is now. Where exactly we come down is hard to say. One can certainly make the case that America will be a less Protestant and a less Christian nation a century from now. But how much less Protestant and how much less Christian is hard to gauge because of the dynamism. Certainly, we see some strong indications of forces that will depress the number of Protestants and Christians, but we also see other forces that, with just a slight change, could in fact increase the number of Protestants and Christians. That might be a very different kind of Protestantism orChristianity than we see today. So it is difficult to predict where America will end up 100 years from now, but I think we can say it will be more diverse, and diverse perhaps in ways that we can’t anticipate.

LUGO:One could quote Casey Stengel for our motto on how we approach these things at the Pew Forum: We don’t make predictions, especially about the future.

SMITH:The second question was which is the fastest-growing religion or religious group. Our survey is able to speak to that by looking at religious change. The religiously unaffiliated-people not affiliated with any particular religion-are the group that has grown the most as a result of religious change. There are more than twice as many people who are unaffiliated with any particular religion today than there are people who say they were unaffiliated with any religion as a child.

Now, those numbers don’t reveal that the unaffiliated have one of the poorer retention rates of all the religious groups. More than half of people raised unaffiliated have now become something else. But they’re the single biggest gainer via religious change because the people becoming unaffiliated outnumber the people leaving the ranks of the unaffiliated by more than three-to-one.

GREEN:The data Greg just talked about refers to the change over the lifetime of the respondent, which we think is a useful way to look at religious change. But it isn’t a point estimate. In other words, we’re not saying the unaffiliated are a bigger group in 2007 compared to 1967. Our survey is not designed to have that kind of point estimate. One reason for that isour very detailed religion questions were not asked in 1967, so it would be very difficult to get a point estimate. But if you just look at the change over people’s lifetime, which is a powerful way to look at it, you do see this dramatic increase in the unaffiliated.

TAMMEUS:But setting aside the unaffiliated for one second, are the Buddhists, for instance, the fastest-growing group in the country, or is it some other particular denomination or religious group?

SMITH:After the unaffiliated, the group that has seen the next-largest net increase is non-denominational Protestants. One-and-a-half percent of the population say they were raised in a non-denominational church. That compares to 4.5% who are non-denominational Protestants today.

LUGO:Keep in mind conversion is not the only thing that drives these numbers. Conversion is a very important dimension of it in terms of projecting growth going forward, and, indeed, some of these smaller groups do pretty well in terms of their efforts to attract converts. Muslims are quite effective in doing that, so are Buddhists; Greg mentioned a high percentage of American Buddhists are converts.

But it’s also immigration rates. Adherents to these traditions are much more highly represented among immigrants than they are among the native-born population. Then you add the number of children, and you begin to see the complexity of making projections going forward. But every indication is that adherents of these other world religions such as Islam, Buddhism, Hinduism, et cetera, will continue to grow as a percentage of the U.S. population. It is now at 5%, which is not insignificant. When the Census Bureau took its own numbers back in the mid-50s, all these groups were virtually a rounding error. So clearly, they are growing.

You don’t need a high percentage of folks who are perceived as new or different to generate a lot of conversation, not least in politics: How do we accommodate the increasing diversity? It’s controversies about Muslim cab drivers in Minneapolis and a Muslim member of Congress swearing on the Koran rather than the Bible. Even 5% of folks from these other world religions, which to most Americans are very new and very strange, generates significant challenges for the country to begin the process of accommodating beyond the categories we’re most comfortable with: Protestant, Catholic, and Jew.

NEELA BANERJEE, NEW YORK TIMES:I just wanted to confirm: Non-denominationals are largely evangelical, is that correct?

LUGO:Actually, non-denominationalism is a characteristic that goes beyond evangelicalism. But it is true they do tend to be predominantly evangelical. What are the numbers there, Greg?

SMITH:We find that among non-denominational Protestants, three-quarters are within the evangelical tradition, 20% are within the mainline tradition, and 5% fall under the historically black Protestant tradition. If you look at it from the opposite direction, though, non-denominational Protestants are not an insignificant part of mainline Protestantism: 5% of all mainline Protestants are members of non-denominational churches. Now that’s less than half of what you see among evangelicals, but nevertheless, it’s worth noting.

LUGO:One of the things we’ll be analyzing for the next iteration of this report is the extent to which this phenomenon is associated with the mega-church phenomenon, which we know is quite prevalent in evangelical church circles, but is not limited to them. It will be interesting to see how those numbers come out.

CHRISTOPHER QUINN, ATLANTA JOURNAL-CONSTITUTION:Do you have any indication that Americans overall are getting more religious as a percentage of the population, or is the growth in those churches coming simply from the conversion of Catholics and from other churches? The second part of that question has to do with the Catholic immigration. Do you see that one-third percentage of Catholics migrating to another faith continuing, so that maybe in 10 years when you do this again, you’ll see the percentage of Protestants is going back up again?

GREEN:I’m not sure the growth of evangelical churches is an indication of an increase in the overall religiosity of American society because, as we have documented, there are other trends going in the opposite direction. That simply highlights American religious diversity. What’s happening in one sector of the religious landscape doesn’t necessarily apply to other sectors, and so we have to be cautious about generalizing from one religious group’s numerical success to the entire society.

Most general measures of religiosity-things like worship attendance-show an enormous amount of stability over time. Our survey is taken just at one point in time, but our numbers fit with that overall pattern of stability. It may very well be, for instance, that the growth of evangelical churches is producing more regular church attenders while the decline of the mainline and the rise of the unaffiliated are reducing the number of regular church attenders, so that it stays pretty much even year in and year out.

SMITH:I do think Latinos may be slightly less likely than non-Latinos raised in the Catholic Church to leave the Catholic Church, but I don’t think there is a huge amount of difference there. Whether or not this will continue to restore the proportion of the population that is Protestant is an open question, because you have to keep in mind there is lots of religious change in the other direction. You see people leaving Protestantism, some for Catholicism, and lots who become unaffiliated. But the other thing you have going on is continuing immigration. As we point out in the report, immigrants are much more likely to be Catholic as compared with the native-born population. If that continues to be true, that will continue to prop up the Catholic numbers as a proportion of the overall population.

MICHAEL HILL, BALTIMORE SUN:I’ve got one question on response rates. This is a 20-minute survey, and I know you tried to limit that, but it’s still a pretty big chunk of time. As a result, despite your huge sample, does your survey underreport those who just don’t care enough about religion to talk to someone about it for 20 minutes? My second question is where do you draw the line between mainline and evangelical?

SMITH:Our response rate for this survey overall is about 24%, and that’s a complicated number, but it’s very comparable to lots of other surveys, including most surveys done here at the Pew Research Center. So we’re confident we obtained a good response rate. In terms of appealing particularly to people who happen to be religious, I don’t think that’s the case here, and one of the reasons is how we structured the questionnaire. We didn’t start off by asking about religion; rather, we started by asking introductory questions about people’s satisfaction with their lives and their communities, and we had lots of questions as well at the beginning of the survey about people’s social and political attitudes. It was really only much later in the survey when we started to get into these explicitly religious topics. I think that should cut down on any concerns about appealing to religious people more than people who aren’t religious.

GREEN:This is, as we indicated, a report based on religious affiliation. In the context of this report, when we speak of evangelical churches and mainline Protestant churches, this is not based on people’s beliefs or behaviors, but rather on their affiliations. We divided up Protestants according to a historical understanding of the development and doctrine of those churches. The mainline churches are actually relatively easy to do because there are fewer of them, and they are well-established institutions for the most part.

With evangelicals it was challenging because there is enormous diversity in the evangelical community. But over the years, scholars, sociologists, historians, and political scientists have developed a good deal of knowledge about these different affiliations. Generally speaking, evangelical churches are those that, either by current statements of doctrine or by historical development, hold certain distinctive evangelical beliefs-not the individual members of the church, but the churches as a whole. Mainline Protestants adhere to the tenets of liberal Protestantism. So that’s how we came up with these distinctions based on affiliation.

JOHN DART, CHRISTIAN CENTURY: The ARIS survey asked people who were born or raised in the Mormon Church how they identified themselves [as adults.] One of their findings was the number of self-identified Mormons was about half of what the church gives out for membership. Other sociologists have done studies, particularly overseas, indicating there is a big dropout rate.

The other question is: Does it look like you will have repeat surveys every four years or so?

LUGO:We hope to do it on a regular basis, probably not every four years. We’ll probably want to do it more than that. And we’re definitely not going to do it during another presidential election year because John has been working 150% of his time on that, and so is everybody else around here. We do hope to do it on a fairly regular basis but, frankly, we haven’t determined just how frequently that’s going to be.

On the Mormon question, in terms of retention and so forth, Greg, did you want to address that?

SMITH:We find 70% of people who say they were raised as Mormons are still Mormons today. So that’s a retention rate for Mormons of 70%. That’s comparable to what we see for Catholics and Orthodox Christians. It’s a little bit lower than what we see for Jews and for Hindus. But it’s substantially higher than what we see for other groups, including Buddhists, Jehovah’s Witnesses, and the religiously unaffiliated.

DART: So do you have an estimate for the Mormon population in the 48 states now?

SMITH:Yes, we estimate 1.7% of the U.S. adult population is Mormon.

LUGO:Let me underscore: that’s adult population. The survey was administered to those 18 years and up. Although we have a table on “number of children,” we don’t make any determinations there because, as you know, in many of these religious traditions, children could be counted as members or not counted as members. We don’t go there. So just take this as a good estimate of the percentage of adults who are members of these various traditions.

ELECTA DRAPER, DENVER POST:On evangelicals, you’re saying you base [your determination] primarily on doctrinal beliefs. I’m assuming these are fundamentalist people for whom evangelism is the main point, and this includes a lot of these non-denominational Christians. But could you just flesh out a little bit more for me who you’re calling evangelical? You said it was about a quarter of the adults of the American population, so I guess I could get the numerical estimate for that, but if you had a specific percentage, I’d take that.

On the unaffiliated, I just want to make sure I understand that group. That’s someone who’s not even vaguely Christian or into Asian [religions] but is non-denominational. Is this someone who absolutely is not identified with any tradition whatsoever? Is that correct?

GREEN:Let me try to answer your question about evangelical denominations as we measure them in this survey, and then talk a little bit about the unaffiliated. The exact number is 26.3% of the American adult population belongs to evangelical churches or denominations. When I mentioned doctrines, these are the doctrines of the churches, not the doctrines of the individuals. These would include doctrines such as the idea that the only way to salvation is through belief in Jesus Christ, that the Bible is the inerrant word of God, that believers have an obligation to proselytize or evangelize, as evangelicals would say, and also that one has to come to salvation through a personal acceptance of Jesus Christ.

Now, within those basic beliefs, there is a lot of variation. Pentecostals and Southern Baptists both agree with those four things, but with some important nuances. In the back of the report, we list all the denominations we classified this way, so you can see exactly who’s there. But basically these large categories of evangelical denominations would include the Southern Baptists and other doctrinally conservative Baptist churches. It would include the Pentecostal churches, headlined by the Assemblies of God, and it would include a lot of the non-denominational evangelical churches, including many of the mega-churches, and then a large number of other smaller denominations such as the Evangelical Presbyterian Church and the Christian Reformed Church, which fit within this doctrinal definition.

Regarding the unaffiliated, one of the things we wanted to do was measure affiliation accurately, but that meant measuring non-affiliation accurately. So we asked people questions that allowed them to feel comfortable telling us they didn’t have a particular affiliation. So when we say “unaffiliated,” we mean people who have told us they do not consider themselves to be members of any organized religion. Some of those people identified themselves as atheists or agnostics. Some of them said, “I’m just nothing in particular.”

Then we were able to divide that nothing-in-particular group up into people who are clearly indifferent to religion, though not necessarily hostile to it, and those who say religion is important to them; they are just not part of organized religion. So the unaffiliated group is a broad group. What they have in common is they told us they didn’t consider themselves to be part of a religion.

LUGO:That 6% of the American public we classify as religiously unaffiliated, again, are people who tell us religion is at least somewhat important in their lives. I bet when we do the analysis of church attendance in the next report, some of those people will go to church sometimes. It could be that they find themselves on some church [membership] roll somewhere; that’s entirely possible.

What we can tell you is, when presented with a clear, explicit choice between “are you this, are you that, are you the other,” these are folks who said, “We are nothing in particular, thank you,” but then tell us religion is somewhat important. If they have some relationship with institutional religion, they don’t consider that to be their primary religious identity. Their primary identity, they tell us, is none of the above: “We’re not associated or affiliated in any significant way with those religious groups.”

DRAPER:(Off mike)-a generic Christian, you would put them in the non-denominational group?

SMITH:That’s a great question. There is a very clear distinction between the unaffiliated and the non-denominational Protestants or non-denominational Christians. Non-denominational Protestants are people who told us either they’re Protestant or Christian, but rather than being associated with a specific denomination, they belong to a non-denominational church. The unaffiliated population, in stark contrast, is made up of people who tell us they have no particular religious affiliation, and that includes the absence of an affiliation with a non-denominational church.

LUGO:This could cut in the other direction too, by the way. You might be tempted to say, “You say you’re nothing in particular but you also say religion is important, and you attend sometimes, so why don’t I put you under the-” We’re not going to do that, just as we’re not going to take those people who tell us they’re affiliated with a particular religion, and yet when we ask them other religion questions, it’s hard to tell they are in any meaningful sense associated or affiliated with it-We’re going to leave people where they tell us they are after we give them the explicit choices.

DRAPER:Did you also say this is the first time evangelicals have become the dominant tradition? Or is this something that has been brewing over the last-Could you characterize that evangelical primacy here for us?

GREEN:Our survey found evangelical Protestants to be more numerous than mainline Protestants. Other surveys have found this as well. Because of the great precision of the measure, we feel very confident our numbers are accurate. Some of the surveys I did in the past, had many, many fewer cases, and so there was some question as to whether the larger size of evangelicals was statistically valid.

LUGO:There is no question the demographic balance within American Protestantism has shifted in the last several decades decidedly in the direction of evangelical Protestant churches. So this protest and reform movement of the late-19th, early-20th century, which we have historically contrasted to mainline Protestantism, is, as a matter of demographic fact, soon becoming the mainline of American Protestantism. At some point, we may have to revisit even our labels as these numbers continue to shift in the direction of evangelical Protestant churches.

JESSICA RAVITZ, SALT LAKE TRIBUNE:Earlier in the call, you threw out a figure about the rates of interfaith marriages. I was wondering to what extent you looked at interfaith marriages, and if there are faiths that are most likely to intermarry, and the ones that are least likely, and if you looked at adult children of interfaith marriages and where they identify themselves, and in the interfaith marriages, how those couples might or might not identify.

SMITH:We do look at intermarriage patterns and rates of intermarriage across different religious groups. We find Hindus and Mormons are the groups in which members are most likely to be married to other people who share the same religion. So 90% of married Hindus are married to other Hindus, and the comparable figure is 83% for Mormons. The groups who have higher rates of intermarriage include the religiously unaffiliated: Six-in-10 married unaffiliated people are married to someone of a different religion, that is to say, someone affiliated with one religion or another.

LUGO: That makes for some interesting conversations around the dinner table.

SMITH:That’s exactly right. In terms of conversion of by adults whose parents were intermarried: Unfortunately we don’t know whether or not someone’s parents were a religiously mixed couple. That’s something maybe worth exploring down the road. We did include some questions on our survey about the activities people pursue in the religious realm with respect to their children. That’s something we will explore in subsequent reports.

LUGO:One interesting finding is on the relationship between intermarriage patterns and conversion. I think many people assume marriage is driving conversions, that as you fall in love with a member of another religious group, you will tend to adopt that other religion to keep peace in the family. What did we find on that particular relationship? It’s not an easy one to explain; it’s a little bit counterintuitive.

GREEN:I’m going to try to explain it and then Greg’s going to correct me when I get it wrong.


GREEN:If it were the case that people changed religions to match the religion of their spouse-if a Baptist married a Catholic and then became a Catholic-then we would expect married people to be much more likely to be married to someone of the same religious affiliation. In fact, we find the reverse. Greg, you better help me here.

SMITH: If marriage was driving religious conversion, then you would expect to find lower levels of intermarriage among people who have changed their religious affiliation. But in fact, we find just the opposite. People who have changed their religious affiliation are actually much less likely to be married to someone who shares their affiliation as compared with people who have not changed their religious affiliation.

LUGO:It’s an interesting finding. It’s just very difficult to explain. Even in Spanish I can’t explain it.

GREEN:What it does tell us, though, is there does seem to be a great deal more acceptance of interfaith marriages than there once was.

LUGO: That was my takeaway, too. Americans are not only comfortable with living with a high degree of religious diversity in society as a whole, but they’re also comfortable living with a high degree of religious diversity in their own homes, which, again, I think reflects the seeker nature of the American public. Americans are on the move, and they don’t seem to be bothered that even in the confines of their home they find considerable diversity.

ANN RODGERS, PITTSBURGH POST-GAZETTE:I’ll look at some of the statistics, but it seemed at first glance as if you had some answers on the particularly contentious issue of whether there are more Muslims than Jews in America. If I’m recalling correctly, it looked as if there were about twice as many Jews. Is that correct?

GREEN:According to our survey work, your understanding of the data is correct. There are more Jews in America than there are Muslims. We not only have data on that from this survey, but last year we had a special survey of American Muslims, which made some extraordinary efforts to measure the size of the Muslim population accurately. So we are confident, within the limits of survey research, in saying the Jewish population is significantly larger than the Muslim population.

Having said that, there are limits to survey research. There are margins of error, so it’s possible our numbers are not completely accurate, but that would be true of any other estimate of the Muslim population. Even if we do calculate that margin of error, it is very unlikely Muslims would be as numerous as Jews at this point in time.

LUGO:Let me again point out this survey was conducted with adults. Aside from the regular challenges of doing good survey work, particularly with groups composed to high degree of immigrants, one has the added complication of language. If we had not offered this survey in Spanish or done some bilingual surveys with the Pew Hispanic Center, this, frankly, would be a significant problem for us because, especially among Hispanics, about two-thirds are not proficient enough in English to be able to take a survey. Now, those numbers are much less for other groups aside from Hispanics, the majority of whom are proficient enough to take a survey in English.

In the Muslim-American survey to which John alluded, we not only gave it in English, but also we had an Arabic option, an Urdu option, and a Farsi option. We’re very transparent in this report that it turned up a higher number of Muslim adults, and we report on that in this survey.

One has to be careful in particular with immigrant groups. It’s important to underscore, as we did in the report, that these percentages should be viewed as minimum estimates of those communities. Again, English-language difficulties make it difficult for folks doing surveys to capture all of those people.

RODGERS:Do you have anything for regional breakdowns within a state? Western Pennsylvania, religiously, is quite different from Eastern Pennsylvania.

GREEN:The answer is no, unfortunately. Even with this large sample size, we can’t get within regions of states. We know what you’re talking about, but there are some states where we don’t even have enough people to look at the state. So unfortunately, no, we can’t look at those regional differences.

LUGO:We provide a good snapshot of the states as a whole. But you’re talking about, to quote James Carville, “The state of Pennsylvania, Philadelphia in the East, Pittsburgh in the West, and Alabama in between,” We cannot distinguish one Pennsylvania from the other on that one.

ANDREW HERRMANN, CHICAGO SUN-TIMES:Do you have some idea as to why so many people leave the affiliation of their childhood in general and Catholicism in particular?

GREEN:We don’t have anything in this particular survey that tells us that. One thing we will want to look at in the future is why that happens. But we do know from the work of sociologists of religion that there are a number of reasons why people do this. Frankly, some people leave for religious reasons. They don’t find their childhood faith particularly cogent when it comes to their own spiritual needs, and that’s one major reason. Another reason has to do with these lifecycle affects. Younger people are in transition. That often includes a transition in their religious commitments. Some people leave because of marriage, apparently many fewer than we originally thought, but maybe because of that. Some people become disaffected with a particular church. Remember, too, the United States has great geographic mobility, and someone who’s very involved in a religious community in Pennsylvania may move to Texas and just not get connected up again. So there are all kinds of different reasons. We have to be careful not to assume religious change has just a single cause.

HERRMANN:If you do man-in-the-pew interviews with ex-Catholics or former Catholics, you often hear the sexual scandals have driven them out of the church. Is there any evidence of that?

LUGO: No, but that question is a prime candidate for this re-contact survey. Again, this area of conversion and why people leave is where we’ll want to probe further. I can tell you on the Latino side, one slice of the Catholic community, when we did that combined survey with the Pew Hispanic Center last spring, by far the most popular reason people gave us when we asked them why they converted to another religion, more specifically evangelical or Protestantism, and even more specifically Pentecostal evangelicalism, it was the desire for a closer experience of God. No other reason was even close, including, by the way, intermarriage, which was really quite low.

We picked up very little evidence that it was disenchantment with the teachings of the Catholic Church or anything of that sort. At least among Latinos, it was more the pull of what they found within Pentecostal evangelicalism rather than a push from within the Roman Catholic Church. That may be different for non-Latino Catholics, and we’ll want to probe that further in the re-contact survey.

SUE FISHKOFF, JEWISH TELEGRAPHIC AGENCY: I’m interested in how you worded the overall question on affiliation because you’ll get a larger number of Jews who will say they identify Jewishly than if you asked, “Is Judaism your religion?” That will help me understand the “other” group in the breakdown in the Jewish category.

SMITH:All of the questions we asked in the survey are available at the end of your report, so I encourage everyone to take a look. The specific religious affiliation question we asked was as follows: “What is your present religion, if any? Are you Protestant; Roman Catholic; Mormon; Orthodox, such as Greek or Russian Orthodox; Jewish; Muslim; Buddhist; Hindu; atheist; agnostic; something else; or nothing in particular.” Then we asked a series of branching follow-up questions depending on people’s responses to that first question.

LUGO: In the next report, we’ll then be able to-remember, we can go down in the analysis to a group as small as 0.3%-go into the sample of American Jews and tell how closely they are connected with a synagogue, for instance, and how often they attend, and how often they pray and so forth. I think we’ll be able to sort out those who are, let us say, culturally Jewish from those who are more religiously Jewish-Or, as you friends in the media said during the 2000 election, describing Joe Lieberman as being “not only Jewish but seriously Jewish,” which meant he was a practicing religious Jew. We’ll be able to sort through all of that in the next report.

Thank you so much for your attention. Our apologies for those of you who are still on hold to ask questions. We’d be glad to chat with any of you. Our communication folks are standing by to schedule interviews; we don’t want any of your questions to go unanswered. Thank you very much.

This written transcript has been edited for clarity, spelling, grammar and accuracy by Andrea Useem.