Pew Research Center
Washington, D.C.

The Pew Forum on Religion & Public Life and the Council on Foreign Relations co-hosted a roundtable to explore geopolitical implications of the growth, distribution and migration of Muslim and Christian populations. Population momentum – rapid growth due to previously high fertility rates – is a critical issue in the Muslim areas of the Middle East, North Africa, and Central and Southeast Asia, yielding “youth bulges” that place substantial economic and political stress on those regions. In Europe, the Muslim diaspora has stimulated vigorous debates over cultural and political identity. At the same time, Christianity has literally “gone south,” exploding demographically in the developing world and augmenting ongoing sociopolitical turmoil in places such as West Africa. How might these demographic trends within the world’s two largest religions – Islam and Christianity – contribute to religious and political conflicts? Distinguished experts Todd Johnson, Brian Nichiporuk and Philip Jenkins discussed these issues and their significance for U.S. foreign policy interests.

Todd Johnson, Director, Center for the Study of Global Christianity, Gordon-Conwell Theological Seminary
Brian Nichiporuk, Political Scientist, RAND Corporation
Philip Jenkins, Distinguished Professor of History and Religious Studies, Pennsylvania State University

Timothy Shah, Senior Fellow, Pew Forum on Religion & Public Life

Download Global Christianity Fact Sheet (prepared by Dr. Todd Johnson for the Pew Forum)


The roundtable was part of a joint project on religion and U.S. foreign policy undertaken by the Pew Forum and the Council on Foreign Relations that is designed to help policymakers and analysts better understand religion’s role in world affairs and the possible policy implications. Although the roundtable was off-the-record, the speakers agreed to make their remarks available online:

Remarks by Todd Johnson:

The main mission of the Center for the Study of Global Christianity is to count Christians. We are fortunate to have so much good source material such as a complete listing of Roman Catholic dioceses published every year in Annuario Pontificio. In addition, churches in nearly every country conduct detailed surveys of their members. We also ask for help from a thousand informants and collaborators. Further documentation is provided by half of the governments in the world from a religion question in their censuses, now including the United Kingdom, which added a religion question in 2001 for the first time since 1851. If anyone here has clout, perhaps we can get the U.S. Census Bureau to add a religion question to the census.

We then publish religious demographic data in books, articles, annual tables and encyclopedias. Our two most recent large reference works include World Christian Encyclopedia, 2nd edition and World Christian Trends. Our methodology and sources are explained in these books. We have recently incorporated our data into the World Christian Database, now online at We integrate United Nations data, church and government censuses, social and health indicators, and data on a variety of Christian activities, such as Bible translation and evangelism. We then highlight the intersection between religion and demography by noting comparisons between religions such as “Christians take up twice as much space as tribal religionists” and “Buddhists have 40 times more wealth per capita than Hindus.”

We are helping the International Forum of Bible Agencies determine whether they are distributing enough Bibles to meet their stated goals. One goal is to give a Bible to every literate Christian family in the world. In our pilot evaluation in Ethiopia we first determine population, household size, number of households, Christian percentage, Christian households, adult literacy, and finally we can begin to evaluate the goal. How long will a Bible last? Seven years? If so, 243,436 Bibles should be distributed every year to meet the goal. It turns out that these agencies had distributed about 145,000 Bibles per year, and that puts them at about 67 percent of the goal. We hope to extend our evaluation to every language and people in the world over the next year and a half.

The global percentage of Christians has been about 33 percent for the last one hundred years. At the same time, the global ethnic background of Christians has changed significantly, transforming the profile of the typical Christian. Eighty-one percent of Christians were white Europeans in 1900. Today this is less than 45 percent and continues to decline. While Europeans constitute an increasingly smaller percentage of Christians, Asians, Africans and Latin Americans comprise an increasingly greater percentage. When we look at the Christian percentage of the six United Nations regions between 1900 and 2005, we see that some remain fairly level, while others are growing rapidly, such as Africa and Asia. Of course, Latin America has been solidly Roman Catholic the entire time.

A typical Northern Christian such as me has to face up to this changing face of Christianity. To illustrate this I want to show you a few portraits of Jesus. I grew up with Warner Sallman’s 1950 portrait of Jesus on the wall of my Lutheran church in Minnesota (over one billion of these hang in churches and homes around the world). In the 1970s I met the Southern California Jesus–“surf’s up,” sand in the hair Jesus–very different than that of my youth. Today I have to try to understand the Sri Lankan Jesus. In this portrait the eyes are half closed–a sign of humility in Sri Lanka as opposed to the eyes wide open in my Southern California Jesus. For all Western Christians it is essential to recognize the changing “face” of Christianity.

A list of the countries with the largest Christian populations in the world in 2005, 2025 and 2050 reveals Christianity as less European and Northern and increasingly more Southern, except for the U.S., which remains at the top of the list across these 45 years. Between 1900 and 2050 Western countries gradually fall off of the list while countries in Latin America, Asia and Africa are added to the list. When you compare these Christian country lists with countries with the most Muslims, India and Nigeria occur in significant positions on both lists–as two of the top Christian and Muslim countries in the world.

A map that I published in April 2004 in the International Review of Mission illustrates the trajectory of the statistical center of gravity of Christianity across its entire history. This statistical center was in Asia Minor for the first thousand years, then in Europe for approximately a thousand years, and since 1900 it shifts decidedly to the Southern hemisphere. As it turns out, the statistical center of Christianity today is in Timbuktu, Mali. In raw percentages, shown on the accompanying graph, Christians were equally divided between North and South for the first 1000 years. However, by 1500, 92 percent of Christians were Europeans. Today 60 percent of all Christians are non-Europeans and Americans.

Recently, I published a paper on the future of religion in the British journal Futures, in which I presented four different scenarios for the year 2200. A purely demographic scenario, where everyone stays in the religion that they are in right now, would yield 32 percent Christians, 26 percent Muslims and 9 percent nonreligious. If we adjust for expected conversions, then we see 39 percent Christians, 27 percent Muslims and 11 percent nonreligious. If there is a Muslim renaissance, we project 38 percent Muslims, 31 percent Christians and 10 percent nonreligious, and if there is a nonreligious resurgence, we could see 34 percent nonreligious, 25 percent Christians and 18 percent Muslims. These four scenarios illustrate how the importance of both religious and demographic trends.

We could imagine several other scenarios for the future of Christianity and Islam:

(1) Large numbers of Christians converting to Islam.
(2) A return to the Crusades. I don’t think the economic and political situation in the world is necessarily right for this.
(3) Non-religious people have played a significant role in so-called religious wars. In our estimate of the number of Christians who were killed in the 20th century, 77 percent were killed by atheists–not by Muslims or Christians or anybody else. What does this say about potential religious wars in the 21st century?
(4) Another trend is highly contextualized Christianity within Muslim culture, such as in Indonesia, Turkey and Pakistan. Today there are many Christian movements that are growing in an Islamic religio-cultural context.
(5) There are also unexpected developments in Southern Christianity. For example, a few weeks ago I visited seven Coptic Orthodox churches in caves alongside a mountaintop in Cairo, including one that seats 3,000 people and an outdoor one that seat 10,000 people. These are completely full throughout the week because there have been credible reports of people healed and four people raised from the dead.
(6) There is a new wave of Southern missionaries traveling around the world, such as the Chinese “Back to Jerusalem” movement, seven million Filipinos in diaspora, as well as large numbers of Nigerian, Korean and Brazilian missionaries.
(7) One should not underestimate the role played by positive contact between ordinary Christians and Muslims.
(8) Finally, with this increased positive contact there is the distinct possibility of some kind of return to what was called the “ornament of the world,” the city of Cordoba in medieval Spain, where Christians, Jews and Muslims studied and lived together in harmony.

So with that I’ll turn it over to Brian.

Remarks by Brian Nichiporuk:

This afternoon I am going to talk about the demography of the Muslim world and its politico-military or security implications. I should note that my background in the demography of the Muslim world has mostly been looking at individual regions like the Middle East, Central Asia and Southeast Asia, and the traditional balance of power dynamics in those regions. This is my first attempt to look at the Muslim world crosscutting as a unit and to piece together lessons from the different regions.

As far as the outline of my presentation this afternoon, I would like to start with a very straightforward overview of high-level demographic trends in the Muslim world, focusing mostly on fertility rates. Then I would like to talk about the demographic transition occurring in much of the Muslim world, and then go into some security implications. The security implications that I’ll look at fall into three categories. First, how are the demographic trends in the Muslim world possibly affecting the nature of conflict? Secondly, how might they affect the nature of military power in the Muslim world? And third, how might they affect sources of conflict? I will just offer one theme for each of those different categories.

I will start with some basics of the global Muslim population. There are about 1.2 billion Muslims worldwide. The annual growth rate is a little over 2 percent–about 2.1 percent. From what I can tell from looking at various sources–and I have just done a very superficial look at this–most of the growth right now, over 80 percent, seems to be due to high fertility rather than conversions. If you forecast out this growth rate, the Muslim population would grow from about 19 percent of the global population today or around the turn of the century to about 22 or 23 percent by about 2025 and then maybe a quarter of the world’s population by about 2050. Todd gave a variety of scenarios that could change that, such as increased conversion, but if you straight-line current trends, that is essentially where we are going.

The largest Muslim countries in the world are non-Arab. They are outside of the Middle East, which many people realize but I think the general public may not. Indonesia, Pakistan and Bangladesh are currently the top three most populous Muslim majority countries in the world. If one examines a list of the top ten most populous countries in the world today, one sees that Indonesia, Pakistan and Bangladesh are all on that top ten list. Nigeria is in that top ten list as well, although many would not consider it a Muslim country because its population is so evenly split between Christians and Muslims. If one defines Nigeria as a Muslim country then there are four Muslim countries among the top ten most populous in the world. Those four stay in the top ten, and I don’t think any other Muslim countries will make it into the top ten by 2050 because of the growth of places like the Democratic Republic of the Congo and India that keep them in the top ten. Current projections estimate that Nigeria will have a larger population than either Pakistan or Bangladesh in 2050.

Muslim regions tend to have significantly higher fertility rates than many other parts of the world. Using data from the Population Reference Bureau, we see that the core of the Muslim world–the Middle East and the Persian Gulf–has a Total Fertility Rate of about 3.7 children per woman, which is reasonably high, and 35 percent of the population in this region is under the age of 15. In North Africa, the numbers are roughly the same; in South Central Asia, the fertility rate is a little lower but not by much. By way of comparison, Western Europe has a fertility rate of 1.6, which is below replacement, and only 17 percent of the population there is under 15. The fertility rate in Latin America is a little higher than in Western Europe but still significantly lower than in the heavily Muslim regions.

If we drill a little deeper though, the story is more complicated, and I use the example of the Middle East as one key region. Muslim demographic trends country to country are not monolithic. There are probably three tiers that we could talk about in terms of demographic profiles. You have a set of countries with very high fertility–say about five children per woman–like Saudi Arabia and Yemen. In a May 2000 analysis, I found that Saudi Arabia’s fertility rate was 6.4; it has dropped a little since then. Then you have some moderate fertility states in the three-to-five Total Fertility Rate range, such as Egypt. And then, interestingly enough, you have Muslim states that are heading in the direction of reasonably low fertility. Iran and Turkey are two examples that are at or below 3.0 children per woman. Turkey is at about 2.6; the U.S., by comparison, is 2.1.

So basically you have three tiers, but I should indicate that even the countries with lower fertility rates, like Turkey, are still growing rapidly. As you know, Turkey is heading toward 88 million people by 2025 because there are so many people in the child-bearing cohort, ages 18 to 30, as a result of previous high fertility rates. Even if they only have two children per woman, the absolute number of people still entering the world in that country is quite large. That is the phenomenon that is called population momentum; it’s sort of a lag, and most Muslim countries are still adding fair numbers of people even though their fertility rates in some cases have dropped.

I want to show two quick examples of the extreme bracketing of fertility patterns for countries in the Muslim world. At one extreme you have the Gaza Strip, which along with Saudi Arabia, Yemen, Oman and similar countries, has extremely high fertility. When we studied the Gaza Strip recently, it had a fertility rate of around seven children per woman, with the pattern of a very high fertility, rapidly growing area. Each youth cohort is growing steadily larger in absolute and relative terms. This type of a population profile, even if fertility rates were to drop very quickly, would still be probably about two generations away from population stability like we see in Europe and much of the West.

Indonesia is a counter-example of a lower fertility Muslim state. It is either the largest or second-largest Muslim state in the world, and it has a very different profile. Its population pyramid does not taper out at the edges, which shows that each successive youth cohort is not much larger than the one above it. Indonesia has a fertility rate of about 2.7, which means that you start to see straight-line edges on the population pyramid as opposed to the cutting out you see in the Gaza Strip. Indonesia is maybe about a generation away from some level of population stability.

Now, you really can’t talk about trends in Muslim demographics without looking at the whole diaspora issue. In addition to high fertility within core Muslim regions like the Middle East, Central Asia, Southeast Asia and North Africa, you also have diasporas abroad, and the best example is Europe. There are significant Muslim populations in Europe, ranging from four to five million in France to the high 200,000s in Spain and Sweden. These populations come from various parts of the world: North Africa, Turkey, and, in the case of the United Kingdom, Pakistan and Bangladesh. There are some politico-military implications to that which I will address.

I’ll just note that there is a demographic transition going on in the Muslim world. We are starting to see across-the-board decreases in fertility rates, even in the high fertility countries. The demographic transition starts to occur when a developing state moves from high fertility and low death rates to decreasing fertility and low death rates. The death rates stay low because in most of these countries public health has advanced; you have immunization and basic child care, so that the mortality rates are reasonably low even in a place like the Gaza Strip, which has very high population growth.

Right now in terms of the demographic transition, maritime Southeast Asia–Indonesia, Malaysia, Singapore, the Philippines, etc.–is relatively far along this path, with fertility rates below three, especially in Indonesia. Central Asia is following behind, and South Asia and the Middle East, which hosts Pakistan, Bangladesh and some of the fast-growing Middle Eastern countries like Saudi Arabia and Yemen, are only really in the early phase of the transition. But I should note that it’s beginning. For example, even Yemen, which used to have fertility rates in the low sevens, has seen some significant fertility drops recently. I don’t remember what the figure is today but it is down to well below the peak. So there are signs that this transition is starting even in the traditionally very high fertility parts of the Arab world. The Central Asian Republics offer another example. In almost every country, with the exception of Kazakhstan, there is a steady decline in fertility rates. Since Kazakhstan’s fertility rate is already extremely low, its lack of decline in the next 25 years is probably not extremely significant.

As far as the security implications, I’ll talk about three. First, the relationship between diasporas and the nature of conflict. We have had diasporas for a long time, for example, the Chinese, Jewish and Armenian diasporas. However, recent advances in transportation and communications now allow people to communicate more effectively with their home countries and to go back and forth more quickly. In the case of the Muslim diasporas in Europe and elsewhere, we have to keep in mind that most of the citizens of these diasporas are perfectly law-abiding, productive citizens, in excess of probably 90 percent across Europe. But there is a small fraction, an activist or a radicalized fraction, that can take advantage of these new technologies and the diaspora settlement patterns to form what one could consider a transnational radical Muslim identity, sort of a Salafist-Wahhabist transnational identity that supercedes state boundaries. This creates ripe recruiting for organizations like al Qaeda and Hizb ut-Tahrir, which is the transnational organization based in London that is now recruiting very actively in Central Asia.

With new technologies, these diasporas can also find new ways to operate in innovative networks, which can involve NGOs, charities and medical organizations. They can lay the groundwork in Western Europe for insurgencies or terrorist acts in the Middle East or in other parts of the world by working with certain charities and NGOs. This raises the prospect of new types of conflict and new types of terrorist organizations based in these diasporas, using modern technology–again, based on a relatively small minority of these diaspora populations.

As far as demographics and the sources of military power in the Muslim world, what we see in the Middle East probably applies to other regions as well. One of the things you find is that, unlike in the industrial age, burgeoning populations are becoming relatively harder to convert into effective conventional military power. The revolution in military affairs that is going on is placing a premium on technology for conventional warfare, substituting capital for manpower.

At the same time, most Middle East states face lots of pressure based on the youth bulges we saw earlier to maintain food subsidies, a social infrastructure, education systems and large internal security forces. The result is that you do not have a lot of desire in the Middle East to translate population growth directly into huge armies. If you look at the size of Middle Eastern armies, they have not grown as their populations have grown. Most of these countries have tended towards a pattern where they focus on a small number of elite conventional forces, like the Republican Guards in Iraq, the Revolutionary Guards in Iran and certain elite units in Pakistan, and ballistic missiles and weapons of mass destruction for the higher end of military capability. At the low end, they focus on large, irregular forces that can be used to harass an invader or pursue an asymmetric strategy. So, unlike what happened with the Franco-Prussian conflict in the 1870s, in today’s Muslim world demography is not being translated into huge force sizes. Instead, these states are moving toward a niche strategy with small elite forces, missiles and weapons of mass destruction all overlaid on a framework of asymmetric warfare.

Finally, there is the relationship between demography and sources of conflict in the Muslim world. I think we may hear a little more about this later, but there are a number of strategically positioned divided states where the Muslim and Christian worlds bump up against each other, such as Nigeria, Lebanon, Ivory Coast and Uganda. In previous work we have done on ethnic conflict and demography at RAND, we found that the combination of heavily intermixed settlement patterns and differential growth rates does not have to create ethnic conflict. However, it can create conflict when combined with a variety of other factors, such as a weak national government, as in Bosnia in the early ’90s, and nationalist mythmaking, as we saw a lot of in the Balkans in the ’90s, where Serbian, Bosnian and Croat demagogues propagated all sorts of nationalist myths.

Unequal economic and income prospects are other important factors. Going back to the Balkans again, at that time you saw the Serbs thinking that they didn’t have a future in a Bosnian state, in the bureaucracy, and they thought that they were losing some of their land. When all of these things come together in a framework where there is a perception that one side has a higher growth rate than another and the populations are heavily intermixed, that can lead to explosions. Northern Ireland in the ’70s is a case; Lebanon, as Tim mentioned, was another case in the ’70s; and of course we had the Balkans in the ’80s and ’90s.

I’m not sure about the extent to which the Muslim-Christian patterns in Nigeria and Uganda are intermixed, whether they have their own cantons or whether they are living next to each other in some parts of Nigeria; I think they are. I haven’t looked into that deeply, but to the extent that you do have this formula of factors, this is definitely something for the intelligence community to watch.

These ethnic and religious conflicts can be very vicious. Historically, if you look at the number of cases, most of them have ended either in some sort of foreign intervention or partition. They are not easy to stop because there is so much enmity and antagonism, and so much of a blood-feud quality that occurs when these things start to escalate.

Finally, in terms of U.S. policy options, we have talked about a lot of policy options for demography and conflict. I think many of them are rather straightforward and obvious, but one thing we have looked at is improving awareness of demographic developments and incorporating demography more into Indications and Warnings measures. We have also said that it would be helpful to promote family planning programs in nations that want to reduce their fertility. For those that don’t want to reduce their fertility, we can offer foreign aid, development assistance and forgiveness of IMF debts to help them deal with the consequences of high fertility and maintain an ability to fund their infrastructure while they are dealing with the medical and educational needs of large-growing populations.

In terms of democratic institution building, which is something that we support as national policy in the Middle East now, I think there is something to be said for being careful about pursuing this in countries with youth bulges and for doing this in a very slow and steadfast manner. Robert Kaplan talks a lot about this. I have heard him speak about the importance of building up middle classes that have a stake in the global economy and also building up strong judicial systems before going headlong into free elections and all of the headline aspects of democracy building. In countries with youth bulges that have tremendous pressures on their infrastructures, such as what we see in much of the Muslim world, we would be well served to be very cautious in building democracies and to start looking at the economic and judicial aspects of these countries. Thank you.

Remarks by Philip Jenkins:

The only words I can begin with are “where do I start.” I suppose I would start with a few lines in terms of “if I was going there maybe I wouldn’t start from here.”

A couple of issues. Firstly, in years to come, we may look back on this day and think, you know, there was a time we used to talk about religious conflict worldwide and we used to look at Christians and Muslims, and we didn’t even think of Hindus and Buddhists back then. Please remember that we are moving into a four-player religious world. If you look in the very long run, it’s interesting. Through most of the last 2000 years, Buddhists represented 20 percent of the world’s population consistently like the 33 percent of Christians. Over the last hundred years, they have been down to five percent; they will be back. Remember we are dealing with a multi-player world.

Since about 1800, the proportion of the world’s population that is Christian and/or Muslim has grown from around 40 percent combined to around 60 percent. That is a very dramatic change. Todd mentioned Christian growth in Africa during the 20th century. Just an observation on the scale of that growth: between 1900 and 2000 the number of Christians in Africa went from 10 million to around 360 million. I believe that is quantitatively the largest religious change of any kind that has ever occurred. We are dealing with quite dramatic things.

I want to emphasize one thing. It might be that the labels that people choose–Christian, Muslim, whatever–might be less important than the style of that religion. Also, I cannot overestimate that youth bulge idea. Think of a religious community–a church or a mosque–with an average age of 60, and you wouldn’t have to go too far away to find one like that, and then think of a typical African church where the average age of the congregation is more likely to be 19 and where the pastor is an elderly graybeard of 25 or 26. So much of the difference that we see around the world that is ostensibly Christian versus Muslim isn’t; it is the Christianity with which we may be familiar in Euro-America versus a young Islam. Many of the differences can be accounted for in that way.

Let’s work through this idea of the youth bulge and religion. What purpose does religion serve here? A couple of comments. Firstly, religion provides ways of managing thwarted opportunity. Combine the youth bulge with a lack of career opportunities, and we face a truly perilous situation, especially when all those young and active people are unable to fulfill the lifestyles that are held out on television, Internet, whatever. Religion provides different ways to channel this resentment. One is an apocalyptic view, to bring down Babylon; the other is different forms of world rejection. But commonly the way in which people deal with the gap between the opportunities that are held out to them and what they can really fulfill is expressed in religious terms.

The second one is war. Brian mentioned the idea that states do not need vast armies, and of course he is right. But on the other hand, if you are running a state with a very high youth population, maybe you don’t need them militarily but you tend to get them into the armed forces for different forms of social discipline, social control. And maybe in cases like Iran in the 1980s, it is a highly efficient way of winnowing a young adult male population.

I recently read a book that is very popular in Germany right now. Götz Aly wrote a book titled Hitlers Volksstaat and it is something you never thought you would see, a new view of looking at Nazi Germany. His argument is that Hitler’s Germany was a robber state that exists by robbing other countries in order to provide generous welfare benefits at home, and welfare is the way that you buy off a turbulent domestic population. It is like the classic Marxist argument about Bonapartism: if you have a troublesome population at home, channel people’s views into religious and/or nationalist sentiment. It is a very interesting study.

One word that I don’t think has been mentioned yet and which I don’t think we can exaggerate is welfare. If there is a key to looking at the growth of radical Islam worldwide, it is not people handing out leaflets saying “become a suicide bomber today”; it is the social and charitable networks, which, yes, do act as fronts for recruiting subversives, but also act as forms of maintaining and promoting community where they are the only forms of promoting and maintaining community. If you want to understand why Christianity is growing in urban Africa, and why Islam is growing in many parts of the Middle East, look at the welfare facilities they provide and look at the increasing proportion of the world’s population that survives only because it is dependent on religious-based voluntary charities. The best example of this is undoubtedly Hamas. If you want to understand the basis of political religion, look at welfare. Look at the idea of corruption and hold out the idea to somebody in the Gaza Strip or Lagos, or wherever, that maybe there are institutions that can help you and aren’t in it primarily to help themselves; that is such a radical idea. I think we have to understand the social-political basis of the radical appeal.

I absolutely do not want to paint an overly grim picture of Muslim-Christian crusades. My book, “The Next Christendom,” went to press on the last day of the old world, September 10th, 2001. In it I used the phrase, “The next crusade will be fought with nuclear weapons and anthrax,” which a lot of people picked up. I would point out the importance in the last few years of reconciliation movements, especially in Nigeria. Vast numbers are dead: 53,000 killed in one Nigerian province since 2001. Think of that number; that is almost equivalent to the number of Americans who died in Vietnam. But now the growing trend is more in reconciliation movements, democratization movements and human rights movements. So I am certainly not trying to suggest that, for instance, global Christianity or global Islam are necessarily scheduled for mutual collision.

The most important thing I would like to add to the discussion has to do with troubled areas of the world. I think we might be missing the most important potentially troubled area of the world, Europe, because what you have there is a classic Lebanese situation of a youth bulge population colliding with a youth collapse situation. If you want to understand just how dramatic the youth collapse is in Europe, think about this: why has Europe in the last 10 or 20 years not produced mass anti-immigrant movements, mass paramilitary forces or violent militias directed against immigration? The answer is that the sort of teenagers and young adults who would be expected to form those groups–the skinhead militias–aren’t there.

So European politics increasingly represent this clash between, as I say, a youth bulge population and a youth collapse population and I would suggest that the potential for conflict, if anywhere, is there. One of my hobbyhorses where I would underline the greatest single danger of recruitment to the radical Islamist cause is the prisons, the most dangerous sources of recruitment and propaganda.

One final comment. How does this affect the United States? I think that the trends we are talking about represent the most significant split between Europe and North America; the single most important issue. When we look at the 1960s, that greatly over-hyped decade, I would argue that the key legal measure is the 1965 Immigration Act, which Senator Edward Kennedy assured would not mean a major shift of immigration in the U.S. towards people from Africa, Asia and Latin America, and did of course lead to an influx from Africa, Asia and Latin America. The U.S. absolutely unconsciously adopted in 1965 a Christian immigration policy. The overwhelming majority of the people who came into the United States were Christian; the overwhelming majority of people who came into Europe were Muslim. And if you want to understand future attitudes over the Middle East, look at that fact.

By the way, there are interesting surveys of attitudes to religion, which show Africa over here as taking religion most seriously. Africa is, to quote John Mbiti, “notoriously religious.” Europe is over at the other extreme: “God? What is that?” The United States is floating somewhere in the Mediterranean, halfway between Africa and Europe. I think that the religious gap between Europe and America represents potentially the sharpest divide now in what used to be the West and it is the key thing to watch for.

I have an image in my mind from Lebanon. I don’t know how many of you remember some of the images of that war but some of them were so much from a science-fiction fantasy. The key battle in Beirut in 1976 was the battle of the Holiday Inn, and you have the battles of the hotels, when Shiite militias finally put enough cannon in the Ramada to take out the Holiday Inn. Sometimes I wonder if something like that might be a face of civil conflict in Europe. However, I would be most alarmed not where you necessarily have a growing population or a shrinking population, but where a growing poor population meets a shrinking rich population. That to me would seem to be a Lebanon in the making. I wanted to end on a non-controversial note, but I’ll have to leave it here.