What Makes a Terrorist: Economics and the Roots of Terrorism

• Introduction
• Remarks
• Questions and Answers

Introduction

JOANNE MYERS:Good afternoon. I'm Joanne Myers, Director of Public Affairs Programs. On behalf of the Carnegie Council, I would like to thank you all for joining us this afternoon.

Our speaker today is Alan Krueger. He will be discussing his book, What Makes a Terrorist: Economics and the Roots of Terrorism.

Ever since 9/11, there have been many theories as to what drives people to commit acts of terrorism. Answers are varied. They range from such simplistic explanations as, "Terrorists hate us because we have freedoms which they don't have," to more complex accounts which view their hatred as a continuation of longstanding religious and cultural conflicts. Still, the argument that is most frequently raised contends that terrorists come from backgrounds of extreme poverty and that they lack an education.

In What Makes a Terrorist , Professor Krueger brings a new perspective to this challenging question. His book is based on a series of three lectures that he delivered as part of the Lionel Robbins Memorial Lecture Series at the London School of Economics and Political Science in 2006. In short, he disagrees with those who make the case that poverty and lack of education are the root causes of terrorism. He argues that this is merely an assumption, lacking any real empirical evidence to back it—and, after all, when you are impoverished and illiterate, you are not really in a position to do much more than survive.

Professor Krueger draws upon data from public opinion polls of the Pew Global Attitudes Project and from the Palestinian Center for Policy and Survey Research to support his thesis. He closely examines the factors that motivate individuals to participate in terrorism. Contrary to popularly held views, he found that if you are educated and have grievances, you are more likely to become radically involved and resort to extreme measures to pursue your complaints. This seems to be especially true if you live in a country which lacks legal and civil recourse to address political problems.

Though it may be easier to base the fight against terror on anecdote and speculation, Professor Krueger believes that if we are to correctly assess the root causes of terrorism and successfully address the threat, we must think more like economists do. To take us into this mindset, please join me in welcoming one of the most respected economists working on this issue, our speaker today, Alan Krueger.

Thank you for joining us.

Remarks

ALAN KRUEGER: Thank you. It's a privilege for me to be here and discuss this work with you.

It's a bit ironic, actually. A lot of the work for the book was done just down the street. In 2004, I visited the Russell Sage Foundation, which is on 64th between Park and Lexington, and I lived for that year even closer to here, in the Royale apartment on the corner of 64th and Third. So I feel like I'm coming home, in a way.

As Joanne mentioned, the lectures that are the basis for the book are from the Robbins Lectures at LSE. Just this morning, I received a very nice letter from Richard Robbins, who is Lionel Robbins's son. Lionel Robbins, for those of you who don't know, was a very distinguished British economist in the interwar period. He was actually John Maynard Keynes's boss during World War II, although I think it was hard to be Keynes's boss. Robbins is probably most remembered for the definition of economics: Economics is the field that studies the allocation of scarce resources and endless wants.

He did a number of other things, and I will probably return to him towards the end of this presentation.

Joanne advised me not to cover everything in the book, so that you will still want to read it. I think one way I can avoid that problem is that I am going to talk a little bit about some work that is not in the book as well, some work I have done since the book was published, looking at domestic terrorist groups. That is something which I think is closely connected to the book, but I didn't do the work before the book came out.

Why don't I begin with the book? One question I am often asked is, why would an economist work on this topic? What does this have to do with economics? Apart from the observation that economists work on almost every important issue, what seems to me to be particularly relevant is that economics has made progress in understanding people's occupational choices, and one can think of choosing to become a terrorist as being something along the lines of making an occupational choice.

That is kind of an ex post rationalization. I became interested in studying terrorism because I had done work on hate crimes after German reunification. I, together with a graduate student from Princeton, who is now at the London School of Economics, had done work looking at whether economic conditions were related to the outbreak of hate crimes, mainly against Turks in Germany.

We found, surprisingly, very little relationship. The unemployment rate in an area, stagnant job growth, the low education—those factors were unrelated to the number of hate crimes that were occurring. What seemed to matter most was being furthest from the West, furthest from traditional law enforcement.

The more I looked into the literature on hate crimes, the more I realized that our finding was not an anomaly. In fact, the earliest work that was done on hate crimes was done on lynchings in the American South. That work, I think, had the misfortune of ending just before the Great Depression. The number of lynchings had been declining prior to 1929 and the economic conditions were improving in the South. The price of cotton was rising. If you extend the analysis through 1936, the correlation goes away. Lynchings actually continued to decline during the Great Depression.

In other work that has been done on hate crimes more recently, looking at hate crimes in New York City, for example—antigay or anti-Semitic crimes, anti-black violence in New York—they seem to be unrelated to the unemployment rate or other measures we have of the state of the economy.

At a certain level, I thought of hate crimes as getting a more pure measure of the supply curve, if you will, the willingness of people to supply their services to do these kinds of heinous acts. Hate crimes are different from terrorism, although I think they are closely related. Hate crimes tend to be more spontaneous, not necessarily directed by a terrorist organization, although I think one could say that the activities of the KKK [Ku Klux Klan] could be described as a terrorist organization.

So that was the background of where I started.

After 9/11, a number of world leaders drew a connection between poverty and terrorism. So that you know that I am not picking on straw men, President Bush, after resisting for a while drawing a connection between poverty and terrorism, said in a speech that he gave in Monterrey, Mexico, “We fight poverty because hope is the answer to terror.” His wife said that she supported education because education teaches the values that defeat terrorism. Tony Blair spoke quite eloquently about how “the dragon's teeth of terrorism” are planted firmly in impoverishment. Gordon Brown made similar statements when he was Chancellor of the Exchequer.

It's bipartisan. Al Gore and Bill Clinton drew a connection between poverty and terrorism.

I have to say, coming from the background of this research on hate crimes, I was skeptical when I started, and the more that I have worked in this area, the more skeptical I have become.

I should highlight that one of my coauthors, who has worked with me on much of this work, Jitka Maleckova, is here, from the Russell Sage Foundation, just a block away.

How does one go about gathering evidence to assess the hypothesis that poverty or lack of education is an important cause of terrorism? Let me describe the kinds of research that I have done in the book.

The first line of research is to look at public opinion polls, recognizing that terrorism doesn't occur in a vacuum, that the prevailing views in society probably influence who becomes a terrorist. In the book I draw on two main sources here. The first, as Joanne mentioned, is the Pew Global Attitudes surveys, and the second is data from the Palestinian Center for Survey Research. Both are first-rate survey research organizations. Pew has done surveys in a number of countries. I was able to look at data from Morocco, Jordan, Turkey, and Pakistan. The handout that was distributed—I won't walk you through all of these charts, don't worry—the handout presents some of the key findings.

In the survey, one of the questions I looked at in those four primarily Islamic countries was, “What about suicide bombing carried out against Americans and other Westerners in Iraq? Do you personally believe that this is justifiable or not justifiable?”

A couple of points jump out at you when you look at this graph, broken down by education level. First, in countries like Morocco and Jordan, the level of agreement that suicide bombings carried out against Americans and other Westerners is justified is remarkably high—distressingly high. Second, it's particularly high among those who have secondary education or university education. For Turkey and Pakistan, the level of support was lower, but it was particularly high, again, among those with secondary education and, in Turkey, university education.

One of the things that one sees in these kinds of polls is that the less-educated or the illiterate are a lot more likely to say, "No opinion." I suspect that that is in part because they haven't really had the luxury to think about those kinds of issues. They are more concerned, as Joanne said, with day-to-day survival.

In one of the earlier studies that Jitka Maleckova and I did, we looked at data from the West Bank and the Gaza Strip. The questions were somewhat similar, but focused more on Israel. You can see, broken down in the handout, answers to the question concerning armed attacks against Israeli targets, "I support, strongly support, oppose, or strongly oppose." Again, the illiterate are the least supportive, although, still, 72 percent were supportive; for high-school graduates, 86 percent said they supported the attacks; those with more than high-school education, 82 percent.

Then, if you look by occupation, the unemployed were the least likely to say that they supported or strongly supported the attacks. Merchants and professionals were actually the most radicalized—well, after students, merchants and professionals were the most radicalized.

It was really quite surprising to me, I have to say, that the unemployed were different than laborers. You might think, the way the labor market works there, that the unemployed and laborers are pretty much a similar pool; just on one day, some of the laborers are unemployed. But this suggests that the prospect of being unemployed might have some impact on willingness to at least say in a public opinion poll that you support these types of attacks.

There is a lot of discussion that I won't go into in the book about the interpretation of terrorism by the people who were responding to these surveys, which is quite different than the interpretation of terrorism, I think, in the West.

But one point I would make is that this finding that the relatively well-off, the better-educated, tend to be more likely to hold extreme views is not new. In 1958, Daniel Lerner wrote a book called The Passing of Traditional Society. In that book he looked at attitude surveys which he conducted, and what he found there was a similar result: The more highly educated, the more advantaged, were those who were more likely to be radicalized. He wrote, "Poverty prevails only among the apolitical masses."

There is a long way, potentially, between what one says in a public opinion poll and what one actually does in practice. While I would say it's surprising and kind of an eye-opener to see that the more advantaged in these societies seem to be more supportive of terrorist activities—activities that I would describe as terrorist activities—that doesn't necessarily mean that that is who becomes involved.

The next part of the book looked at characteristics of those who become involved in terrorism and compares them to the population as a whole. Here I draw on some work that Jitka Maleckova and I did together, some work by one of my graduate students at Princeton, and other work as well. In the handout, I just highlighted two examples.

The first is from work from my graduate student, Claude Berrebi, who is now at the RAND Institute in California. Claude collected information from the biographies of suicide bombers and other terrorists—here I just show you results from the suicide bombers—from Hamas and the Palestinian Islamic Jihad. He compared them to the population in the West Bank and Gaza Strip, from labor-force surveys. What he found was that the suicide bombers were much less likely to come from impoverished families. About 12 percent of the suicide bombers grew up in families that were below the poverty line, compared to 32 percent of the population in the West Bank and Gaza Strip.

If you look at the educational distribution, you see a remarkable imbalance. The suicide bombers were overwhelmingly likely to have graduated from high school or to have enrolled in college. Over half of the suicide bombers had some education beyond high school, compared to about 13 percent of the population in the West Bank and the Gaza Strip.

The suicide bombers tend to be young. If you compare them to younger people in the population as a whole, you still see this tremendous imbalance.

This also, I think, is similar to work that Jitka Maleckova and I did looking at members of Hezbollah. We obtained data in a similar way, looking at the biographies of the shahids [martyrs] from Hezbollah. Based on the family situation—there were often some descriptions about the parents' occupation, whether they had traveled, where they lived, and so on—we made an estimate about whether the individuals grew up in poverty and could compare them to the Lebanese population.

By the way, I never thought it would be easier to get data on members of Hezbollah than it is to get it on the Lebanese population. But it took quite a bit of work to get the data from the Lebanese census.

If we look at the education levels, members of Hezbollah were also more likely to have a secondary education or higher, particularly when we compare them to the population in southern Lebanon and Beirut, which is probably the relevant pool.

This work is described in the book. More rigorous econometrics have been used to analyze these results. I won't bore you with those details. The results seem to hold up when you do multivariate logistic analysis.

In the book I also talk about biographies of members of Gush Emunim, which was an Israeli extremist group. I think it has been known to intelligence analysts and people who work in terrorism for a long time that individuals who join terrorist groups tend to be relatively more advantaged. That was the case, for example, with a number of the European terrorist groups. Even in Latin America, that finding seems to hold.

I think the new addition here is to more systematically compare the backgrounds of the terrorists to the population as a whole. But maybe that step wasn't so necessary. I think, in a way, it is kind of rediscovering findings that you could see out there. In fact, there was a report prepared by the Library of Congress for an advisory committee to the CIA, which I cite in the book, which reached pretty much the same conclusion. What that report said is that the only exception that they came across was possibly Northern Ireland. I would emphasize "possibly," because the evidence for Northern Ireland, I think, is somewhat conflicting. It's interesting to consider why Northern Ireland might be an exception.

But by and large, I think this pattern tends to hold. I think the support is strong enough there to think about explanations for why. Why is it that the more advantaged, better-educated are more likely to join extremist groups? As an economist, I tend to think about the supply side and the demand side for this market. I call it "the market for martyrs."

On the supply side, I think the knee-jerk reaction is to think about the supply side of people who commit crime. What drives people to commit crime? There, economists have a pretty well-worked-out model on the economics of crime. We tend to think that people who have few legitimate opportunities—a low opportunity cost of their time—are the ones who turn to crime. That model works pretty well when it comes to property crime. It doesn't work very well when it comes to violent crime, which tends, in my view, to be more emotional.

That model also works pretty well when it comes to explaining who commits suicide. Suicides tend to be more prevalent among people of lower means. There have been some economists who have worked on the economics of suicide and who argue that the alternative opportunity costs are important.

When it comes to terrorism, however, I think opportunity costs are outweighed by other factors. Most importantly, I think they are outweighed by belief in a cause. After doing this work, I think a better analogy for thinking about terrorism than crime is the analogy to voting. For terrorism, which I haven't defined, the definition I have in mind is a premeditated violent act against people or property for the purpose of influencing an audience, making a political statement. If you think about who votes, it is not the people who have the lowest cost of voting. The people who have the lowest cost of voting are people who are unemployed. They don't have to miss work. They have a lot of leisure time, so leisure, presumably, is less valuable to them than it is to a busy executive. Yet the unemployed, less-educated are less likely to vote.

What I come away with from this evidence is that the forming of political views requires some understanding of the issues, requires some investment in understanding the issues. Making that investment is less costly for the better-educated. I also think better-educated people tend to be more self-confident. They are more willing to believe their opinions. They might be more opinionated. That might be why the less-educated are less likely to express an opinion in these polls.

While, as an educator, I might like to think of education as building consensus around society, I think it also can lead some people to extremist positions. There is a wealth of evidence from psychology. If you take people who are on opposite sides of the debate about the death penalty and you present them with the research that has been done on the death penalty and you ask them afterwards where they stand, you might think, having looked at the same evidence, that that brings them together. In fact, what it does is it pushes them apart. People are very good at looking through this evidence for the parts of evidence that support their preconceived views.

So I think presenting them with evidence, trying to educate them in an area, doesn't necessarily bring about consensus. I don't think people are necessarily wired to work that way.

That's the supply side. I think on the supply side, people who are more willing to join the terrorist groups tend to be those who are better educated and more advantaged. In fact, the way I often put it is, if you want to think about who becomes a terrorist, don't think about who is so desperately poor that he or she has nothing to live for; think about who believes so strongly in a cause that they are willing to die for it. That would probably lead you to think about people who have passionate views, fanatical views, and are willing to act on them—probably those who are more advantaged and better educated.

I think the demand side points in the same direction. The terrorist organizations find failure very costly, because they might be caught and killed as a result. So they want to succeed. They want to choose people who are most likely to succeed.

By the way—I should have mentioned this earlier—most of the research I do in my life is on education. It is on the value of education in the labor market. Education, I think, is extremely valuable for individuals and for society. I think education does raise individuals' productivity. It's one of the reasons why people with higher education are paid more.

I think education also probably does make people more committed to whatever it is they are doing, which might be, in this case, carrying out a terrorist act. It might make them more skilled at committing these crimes. And I think there is more skill involved in these kinds of acts than meets the eye, especially when it comes to international terrorism.

So on the demand side, the organizations want to choose those who are most likely to succeed, who are probably those who are more committed and better educated.

There is a very interesting study by Claude Berrebi and Efraim Benmelech, who is at Harvard, looking at suicide bombing attacks that took place, as well as failed attacks, which were either intercepted or the individuals backed out. They were able to compare the characteristics of those attacks that succeeded and those that failed. What they found was that the less-educated were more likely to fail. They also found that the better-educated were sent to more important targets. They were sent at more important times in terms of the political cycle. In this sense, I think the deployment of the suicide bombers has this element of rationality, where the terrorist organizations are choosing people for missions where they think they are more appropriate for their skill set.

Instead of going through the next chapter of my book, I want to say a little bit about the new work that I have done on domestic groups. Here I should say, even though I have been emphasizing in my comments a lot of Islamic terrorist groups, I argue in the book, and I present a lot of evidence, that virtually all major religious groups have had experiences with terrorism; terrorism is not confined to Islamic groups. I think one of the problems that we have now is that we are so focused on Islamic terrorism that we can miss instances like Timothy McVeigh or Ted Kaczynski or more run-of-the-mill homicides. I tried to argue that I think we need to have a broader perspective.

Having said that, I am going to say a little bit about some of the work I have done more recently on domestic Islamic terrorist groups.

I have gone through a similar kind of exercise. The Pew Research Center has done one survey of American Muslims, which was a real challenge to do. It was a remarkable survey. It was conducted between January and April of 2007. They paid people $50.00 for participating. It was a 30-minute interview. The interviewing was done in four different languages. They ended up interviewing 1,050 Muslims in the United States, age 18 and older.

I use this data set for two purposes:

• • One, to look at how extremist views vary by people's education and by their income.

• Second, to compare the U.S. domestic Muslim population to those who have been caught up, or at least accused of, terrorist activities in the United States.

You have a few slides here in the handout.

I would say the results pretty much fit the pattern that I have shown you, although I would back off a little bit. The question that I show you here is, "Some people think that suicide bombing and other forms of violence against civilian targets are justified in order to defend Islam from its enemies. Other people believe that, no matter what the reason, this kind of violence is never justified. Do you personally feel that this kind of violence is often justified, sometimes justified, rarely, or never?"

You can see the results broken down by whether the individuals are native-born or foreign-born. About 35 percent of the Muslims in the United States are native-born; 65 percent are foreign-born. There is not a very strong pattern by education. The high-school dropouts are a little bit more likely to say, "Often justified" than those with more than a high-school degree, but it's really not much of a pattern.

By income, there was no pattern at all. If you look at the native-born with more than $40,000 of income a year, 3 percent said, "Often justified," versus 1 percent of native-born with less than $40,000 of income. The income pattern was not statistically significant if you do a formal test.

So public opportunity for domestic Muslims, I would say, doesn't bear a strong relationship to individuals' socioeconomic status.

The second analysis that I have done is more interesting: The Homeland Security Department provided me with an unofficial list—it said right on it, "Unofficial list"—of Islamic terrorist cells in the United States, and then—you will like this—with links to Wikipedia for each of the groups. This was all based on unclassified information. [Laughter]

Then I had my research assistant try to track down as much as we could about the 59 individuals who are part of these homegrown terrorist cells. You can see a list. “Homegrown” and “terrorist”—both of those words should be in quotes, probably. Some of these were kind of hapless groups who got wrapped up in things.

When I mentioned this to my source at Homeland Security, he pointed out that he had access to classified information which made him more concerned about them. I only know what I read in The Washington Post , which made me think some of them were pretty hapless groups.

What is interesting is that they had certainly formed some opinions or some stereotypes from the biographies that they had seen of these individuals. Then, when you look at them more formally, I think it changes your views.

We have data on 59 individuals from what I call alleged homegrown terrorist cells. For all of them, we know their country of birth. For 38 of the 59, we know their education. We know their age, roughly, when they became involved. We have citizen status for all of them.

Let me summarize some of the results.

The homegrown groups are a little bit more likely to come from the Arab region, as defined by the United Nations, than are U.S. Muslims in general. They are quite a bit less likely to come from Iran.

Educational attainment is really interesting. This is the chart in the middle of page 5. Those who have been charged with being involved in terrorism are more likely to come from the middle education ranks—some college, college graduates—and less likely to come from high-school dropouts or from post-college graduates than are other U.S. Muslims. The average level of education is about the same, which is really remarkable. You read about the Miami Seven or the Buffalo Six, and it seems like these were people who didn't have much going for them. But the average education for these 59 people was 14.2 years, which is not very different than the overall average for the United States and not different than the overall average for Muslims in the United States.

Let me take an extra five minutes to just wrap up.

There are some other—you can read it at your leisure—really interesting results here. For example, 22 percent of the homegrown terrorists were enrolled in college at the time that they were arrested, which seems very high. It exactly matches the percent of U.S. Muslims who are currently enrolled in college. The terrorists were a bit younger.

It seems like you hear a lot about converts who are involved in Islamic terrorism. But those individuals in our sample are actually less likely to be converts to Islam than was the general population of Muslims in the United States.

Interestingly, 78 percent of those whom we have charged with terrorism were in the United States legally; 22 percent were in the United States illegally.

Let me just very quickly summarize the rest of the book, which will meet Joanne's warning of not telling you too much about the book.

The second chapter looks at the countries that terrorists come from and the countries that they target. What I find there is that GDP is not a very good predictor of the countries that terrorists come from. On the other hand, terrorists do tend to target wealthier countries.

I should say, this is looking at international terrorist incidents. The vast majority of terrorist attacks are purely domestic and carried out for purely domestic reasons. International terrorism is quite rare.

But when I do look at information I culled from the State Department, from the Patterns of Global Terrorism reports—the book has an interesting story about some of the problems with those reports—what I find is that the terrorists are tending to come from countries that are more repressive when it comes to civil liberties and political rights. They also tend to come from nearby. So even when you look at international terrorist attacks, which are attacks carried out by people from one country against people in another country, they tend to be fairly local, not traveling large distances. In fact, in, I think, 88 percent of the international terrorist incidents, the perpetrators were operating in the country where they were from.

September 11 was a really rare event, where people from one country came to another country and commited an act of terrorism. The typical terrorist attack is an attack on foreigners who happen to be in that individual's country.

I kind of extended this analysis to look at the insurgency in Iraq. I was able to look at data on the national origin of foreign fighters whom we captured in Iraq. To a remarkably similar extent, I found that the same variables mattered. So the insurgents were coming from countries that were nearby, from countries that were more repressive when it came to civil liberties and political rights. GDP didn't matter. In fact, to the extent that GDP mattered, they were coming from wealthier countries.

I have to say, I think of this as a little bit of a victory lap for the negative binomial regression analysis in the book, in that the information I have, which is by no means perfect, on the national origin of foreign insurgents captured in Iraq showed that only 10 percent of those whom we captured were from Saudi Arabia. In the model I estimated, Saudi Arabia was the biggest outlier. I predicted that Saudi Arabia should have accounted for 44 percent. The data I was provided were indicating that Saudi Arabia was a small proportion. Yet the model suggested 44 percent.

Some of you may recall that about a month ago there was an article in The New York Times , which cited American sources in Iraq, claiming that Saudi Arabians accounted for 40 percent of the insurgents who had been captured in Iraq.

So in this case, I think the model was actually more correct than the figures that the military had released at the time.

I should also emphasize that, from everything I have seen and from the work I have done, I think that the insurgency in Iraq is mostly local. The State Department had a little-noticed report where they said foreigners probably make up 4-to-10 percent of the insurgency. Most of terrorism is local. We have been focused a lot on outside influences, but I think that might be getting more attention than it deserves.

The last chapter of the book, which I will touch on very, very lightly—I would be happy to address some questions—addresses the economic, psychological, and political consequences of terrorism. It's a big area.

I think the economic impacts are probably small. The reason I say that is that the types of terrorist attacks that the world has seen so far have a relatively small effect in terms of destroying capital, human capital and physical capital, which is what is most important for modern economies. I think it's only if people overreact to terrorism, which is certainly a real possibility—if irrational fear takes over and governments make bad decisions—that terrorism can have a large impact on the economy.

One of the goals of writing this book was to try to put the threat of terrorism in perspective. I present calculations. Others have done this. I think John Mueller has done this in a book called Overblown , comparing the risk of dying in a terrorist attack to the risk of dying in an airplane crash—which, I read today, is actually lower than what it was in the data that I used—to the risk of dying from other causes.

To put this in perspective, someone is over 500 times more likely to die from their own suicide than they are to die from a terrorist attack.

I think we do need to be concerned about terrorism, but I think we need to keep the risk in perspective. There are lots of risks that people face. I think over time people come to adjust and cope with those risks. The risk of dying in a commercial plane crash is substantially higher than the risk of dying from a terrorist attack, and people certainly learn to cope with their fears of flying.

What I have tried to do in this book—and I hope I have indicated this in my remarks—is to show how we can use data in the war on terrorism. In fact, I thought about calling the book Enlisting Data in the War on Terrorism , and my publisher thought What Makes a Terrorist was a much better title. One of the reviews, which was otherwise quite flattering, said that the book does a better job of saying what doesn't make a terrorist than what does make a terrorist, which was a fair point.

But what I take away from this analysis that I have described and from this general pattern of results is that people are drawn to terrorism for a number of different reasons. People are drawn to terrorism because they have extreme grievances. There are lots of people in the world who have extreme grievances, and there are lots of reasons. Some are religious, some are nationalistic, some are political—any number of reasons. Ted Kaczynski was worried, I guess, about technology. You have this very large supply of people who might be willing to commit extreme acts, if a terrorist group were to help channel them.

One of the reasons why I think it's very hard to find evidence that poverty or education matters is that people are drawn to terrorism for diverse reasons. There is not kind of a standard profile.

This suggests to me that working on the supply side is not going to be a very effective policy. I think there are many reasons to reduce poverty around the world and to improve education around the world. I don't think reducing terrorism is one of those. But I think focusing on the terrorist groups, which are kind of the finite resource which channels the activities of people who are willing to perform these acts, is probably a more effective route.

Why don't I stop there?

Questions and Answers

QUESTION: You referred regularly to education, but you didn't distinguish between a Wahhabi-financed madrassa in Pakistan and Roberts College, or something like that.

ALAN KRUEGER: That was in part because of the limitations of the data. One of the points that I emphasize in the book is that, to the extent that we try to use education, we should focus on the content of education.

Jitka and I did look at Hezbollah and whether individuals were in Hezbollah-supported schools. Most of them were not.

There has been some other work more recently that has been done, looking at people's majors. The most common major is engineering, followed by doctors, and then I think it might be lawyers. [Laughter]

I did a little bit here looking at the occupations of the domestic Islamic terrorists who have been apprehended. They are pretty much kind of broad middle-income occupations.

My own sense is that I think it would be useful to have education which is focused more on useful skills. I suspect, however, that that is not going to have that much of an impact on reducing terrorism.

QUESTION: John Schindler, who is now at the Naval War College and formerly of the National Security Agency, has a new book called Unholy Terror, dealing with al-Qaeda in Bosnia. To a large extent, he says, most of the people who write on terrorism really don't know what they are talking about. He would, in this case, agree with your findings. That is precisely what he is saying.

On the other hand, I think where he would part company with you comes at the end of your presentation, when you say that the odds of dying in a terrorist attack are different from your own suicide. There you have to take a look at the situation itself. The odds of dying in a suicide attack in Israel or some other place where attacks are concentrated are going to increase dramatically. There you have to look at, in some cases, what is really going on here. Is it a consolidated attempt by a terrorist group to launch some sort of attack, as they did on America or in Spain or something? There they are looking to kill people.

Maybe you don't want people to get too bent out of shape overall, but, on the other hand, you have to be more concerned.

ALAN KRUEGER: Certainly I think we need to be concerned about the risk of dying from anything. But I think we want to allocate resources in an efficient way.

I was at a program at Princeton where a former analyst talked about what would happen to the U.S. economy if a bomb went off in the Mall of the Americas, and that it would kind of close down commerce. If it does, it's completely irrational. The risk of going to any mall in America, even after such an event—the risk of dying from driving to the mall would exceed the risk of what might happen to you at the mall.

When I talked about the economic impacts in the book, where I argue that terrorism could possibly have an effect on the economy is when it is sustained and more frequent, like in the Basque region. I think the one-off terrorist attacks—and I think the evidence supports this—tend to be fairly limited in terms of their economic impact, fortunately.

I have gone through calculations worldwide, looking at the risk of dying in terrorist attacks, not just in the United States. I don't know if we do part company, because I think the audience that I really had in mind is in the United States or in the United Kingdom, where I gave the lectures. It would be interesting to do that calculation for Israel. But we don't face that risk here, and I suspect that we won't.

QUESTION: I want to thank you for addressing the question of probabilities in this whole debate, because there is far too much emphasis, I think, on intentions out there, which provokes a lot of speculation about what might happen as opposed to the probabilities of what might occur.

My question deals with the whole idea of counter-radicalization. You have put together a book that tells us what makes a terrorist. Do you have any thoughts on government counter-radicalization strategies and how we prevent people from going down that path?

ALAN KRUEGER: It might be too late for this, given the path that we have gone down. After 9/11, I gave a talk at USAID on some of this work. One of the things which comes through in my results and other people's work is that international terrorists are less likely to come from countries that support civil liberties and political rights. They are more likely to come from countries like Saudi Arabia. Fifteen of the 19 on September 11 came from Saudi Arabia.

What does the United States stand for in the world? What we stood for was an example of democracy, protection of individual civil liberties. I would have thought that if Madison Avenue wanted to do something about 9/11, we would have thought about how to have a campaign to try to spread democracy and civil liberties around the world. I don't think you can do that militarily. I think we are learning that.

I think that is one step that we could make, to try to promote alternative means of protest, to promote developing a tradition of having nonviolent means of protest.

What is interesting to me is, I think people who come from a society which doesn't have a tradition of openness, of freedom of assembly, freedom of expression, even when they move to a society that does have such freedoms—it might be that the heritage that they grew up with affects their judgments or the way they think they can effect change. People often say—which is a good question—what do civil liberties have to do with the attacks that the United Kingdom has experienced?

The other thing is—you highlighted the probabilities—the models that we estimate don't predict everything. I don't think we are going to be able to stop all—I don't think we will have a failsafe policy. But I think what you can do is lower the odds. To my mind, that would have been one step that would have been a step in the right direction.

QUESTION: If you look at the array of motivations that you mentioned, the three examples that you gave, is there a preponderance of, let's say, religious over irredentism, one over the other? That is one question.

The other thing is, of course, everybody is talking about and is extremely concerned about our government leaving Iraq. One of the arguments against any precipitous departure is the possible growth of terrorism, perhaps coming in from Iran or al-Qaeda, which is already there, finally. (The president thought they were there a long time ago, but they only recently arrived.)

So the expansion of those groups—is that something we should be concerned about?

ALAN KRUEGER: The first part of your question is actually easier, so why don't I start there.

What I found when I looked across countries was that religion was surprisingly unimportant for predicting the countries that terrorists came from—the predominant religion of the countries. We have a tendency, I think, when we see violence on the television set in a foreign country, to attribute it to religion. I wrote a piece in The Washington Post about the fact that all major religious groups have had experience with terrorism. You should see the hate mail I got. It was really remarkable.

I think we just assume that these attacks are being done for religious reasons, if we don't know very much about the situation.

I think it's very context-dependent. My own sense is that it's probably nationalism. If I were to choose one that was probably the most important, it's probably nationalism. That is the way I would describe the conflict between the Palestinians and the Israelis.

What happens if we withdraw from Iraq? We have an awful mess on our hands, regardless, I would say.

This is an area where I am not an expert, so I will politely decline to answer what I think we should do.

But one of the benefits, I think, of the analysis I did about where the foreign insurgents are coming from is that they are eventually going to go home, just like the group went back from the Bosnian conflict. I think this will give us a handle on where they are going back to and where we need to be more cognizant of activities of people who have training. Now they have been trained up in how to carry out these kinds of attacks.

QUESTION: This is more of an observation that, perhaps, you could address, rather than a question. If you look back historically, the leadership of revolutionary movements were all intellectuals. They were cerebral people. They were educated. They were idealistic. They were creating new worlds, in a sense, whether it was Marx or Stalin or whether it was in the French Revolution, and certainly our Founding Fathers were.

You call all of these "terrorists." I agree with you; indeed they are. But they don't call themselves "terrorists." Many of them may think of themselves as revolutionaries who are creating a new world. The ones who want to re-create the caliphate, although it has a religious dimension, may think of themselves as revolutionaries.

It seems consistent that a lot of these people would be better educated, because they have thought this through and this is something that they are emotionally and intellectually committed to.

Unfortunately, in some of the earlier revolutions, the second generation becomes horrifically distorted, if you look at Communism and you look at the Terror in France—not here. It became diluted after a while.

But do you anticipate that that might happen, that it will become more of the non-thinking people, who will do it more as acts of desperation and protest?

Anyhow, it's just an observation.

ALAN KRUEGER: I think the observation is absolutely right. If I were to work in this area again, I might avoid using the term "terrorism" altogether. It is a very loaded word. A more accurate description is "political violence." In fact, Jitka and I were thinking about writing a book called Randomly Targeted Acts of Violence , which would encompass not only terrorism, but also hate crimes.

The British called George Washington a terrorist. That is probably a violation of the Patriot Act today, to say that in the United States. [Laughter]

If you think of it that way, the model that I have in mind, if I may talk about it in terms of a model, is that the leadership, as you said—and this is exactly right—is people who are dedicated to the cause, and that is what is motivating them. If it spreads, then they have to get people who are motivated by something else, because there are only so many people who are motivated by the cause.

One of the reasons why terrorism differs from civil war, although terrorism can be used during a civil war—if it turns into a full-scale civil war, I think economic factors do matter. You tend to get a civil war when you have a large enough group of people who are willing to oppose the government, usually when the government itself is weak. Then I think getting mercenaries matters and economic factors matter, once you get beyond the top cadre.

I don't do enough in the book in drawing historical comparisons, but I think that they are perfectly appropriate.

QUESTION: I don't know the precise number—you probably know it much better than I—but I believe there are something in the neighborhood of 15 million Muslims in this country, give or take a couple of million. If you put the percentages together from your table, you still come up with something like a million or a million and a half of those who believe that violence is often justified. That is a very big number.

My question is whether anybody asked in your surveys, not just, "Do you think it's justified," but, "Would you do that? Would you sacrifice yourself for this kind of thing?" That seems to me a core question that is perhaps more revealing than whether you think some abstract principle is appropriate.

ALAN KRUEGER: I am not aware of that being asked: Would you join this cause? Would you carry this out?

First, factually, the 15 million number, I think, is way off. What Pew found was that 1 percent or 1.5 percent of Americans identified themselves as Muslims. That would be closer to 4.5 million. As I said, 65 percent of them were not U.S.-born.

The way that this poll got attention in the press—on Fox News I saw it described—the same question was asked in Europe; the same question was asked in the Middle East. Even though the United States came out on the low side compared to other places where it was asked, it was still shockingly high. I wouldn't deny that. But just compared to what you see in other countries, I would characterize it as lower.

QUESTION: I am somewhat puzzled by how you moved from your empirical observations of terrorists to the causal argument that poverty doesn't cause terrorism. These people clearly believe they are acting on behalf of people who are poor. If there was no causal relationship, these people would be terrorists absent poverty and oppression. I think that is absurd.

ALAN KRUEGER: That's a good question. The book spends a lot of time on that. That is Chapter 2, which I went over quickly.

I think what we could agree on is that the elites are the ones who are joining the terrorist groups, but maybe what is motivating them is the poverty of their countrymen.

I think there is very little evidence for that. That is a harder hypothesis to test. Why do I say there is very little evidence for it?

First, if you look at the countries where the terrorists are coming from, they are not disproportionately poor countries. That's what the State Department data show. That's what the NCTC [National Counterterrorism Center] data show. They are not, particularly, countries which have more inequality compared to other countries of the world.

One of the poorest places in the world is sub-Saharan Africa. That is not a major concern, for international terrorism at least.

That is one piece of evidence.

A second piece of evidence is, if you look at what they are saying, they say very little about poverty. I have some quotes in the book from the leader of the attack in London, in July of 2005. He said, "We're not motivated by tangible commodities that the world has to offer." I read a transcript of Osama bin Laden's most recent statements. It's a laundry list. I think the strategy that they have is to have a laundry list of grievances to draw people in and then they are going to direct them for their purposes.

If you do a textual analysis of the statements that the terrorist leaders are sending out, they are not using poverty. They use religion more frequently than they use poverty.

QUESTIONER: That may be the case for this particular group of Islamic terrorists. But I don't know if it is historically—if you go back through various terrorist groups, it certainly wouldn't be the case for Narodnaya Volya in the 19th century, the anarcho-syndicalists in Western Europe in the 19th century. It's not the case for the IRA [Irish Republican Army].

You may be correct here, but I don't think it's historically correct.

ALAN KRUEGER: For the IRA, I think you could say it was more nationalism.

QUESTION: Just a quick comment and a question. When you put up that list of these groups of homegrown U.S. Islamic terrorists, you referred to them—

ALAN KRUEGER: The unofficial list.

QUESTIONER: You said "alleged" and "classified." None of that is classified. Every one of these is in the press. That's number one.

Number two, "alleged"—of this group here, I can tell you, seven of them have gone to court and been convicted in federal court, and three of those groups are awaiting trial right now.

My question is this. You said only if people overreact to terrorism can that have an effect on terrorism [sic]. I have spent a lot of time in the Middle East. This summer I have been in Morocco, in Egypt. They look upon this somewhat differently. They overreact. Their idea is, "We have to be very tough on this." If you take Egypt, Sharm El Sheikh, Taba—the bombings in the Arab countries—they look upon that as a necessity, to be very hard and tough. Otherwise, their economies will duly be affected; Egypt, for example, 6.5 million tourists just last year. With one or two of these incidents, you can bring down an economy.Would you discuss that a little more, since you are an economist? But I do think that you cannot overreact when you are protecting the infrastructure of many of these countries, and it applies to us equally.

ALAN KRUEGER: I don't agree with the last part, that it applies to us equally.

On the first part, the standard I used for this list is that they needed to have been indicted. That's why all of them had been indicted or convicted of something. Obviously, everything I had access to was unclassified.

On overreacting, I will refer to Colin Powell. Read the interview Colin Powell gave in the last issue of GQ , where he said, can terrorists change your way of life? Terrorists can knock down buildings, they can kill people, and we should try to stop them from doing that, but only we can change our way of life, by making bad decisions, by ending our tradition of civil liberties.

If you believe these results—I think one could have valid doubts about some of the analysis I have done, but if you believe the results that suppressing civil liberties is one reason why terrorism arises—then I can make the argument that we are moving in the opposite direction, that in the long run we are going to give people like Timothy McVeigh more reasons to be paranoid. If you think about what motivated him, it was concern that the government had overreacted. My understanding of what he said motivated him was the government action at Waco. He also had concerns about having been spied on and so on.

But hold all that aside, because, hopefully, he is a unique case.

On the economic impacts, certainly some industries can be impacted. The travel industry was impacted in the United States. The United States is a $12 trillion economy. We are a very diverse economy. It's one of our great strengths. The comments that I had in mind were mainly for the U.S. economy. I think it's very unlikely that terrorism can have a noticeable effect on economic activity in a $12 trillion economy which is diversified.

Look at what happened after September 11 in New York City. It bounced back very quickly, and it was, in part, because there were so many substitution possibilities. So even though a lot of the business space was destroyed downtown, people moved into hotels, which became vacant because there was less travel. They discovered empty floors that they had in New Jersey.

Yes, certainly, some industries can be affected. But what you learn from economics is to look at the system as a whole. For the system as a whole, the experience with hurricanes, which cause more damage, has generally been that they have very little impact on economic growth in the long run or on overall economic activity.

JOANNE MYERS: Alan, I think you have given us many good reasons why we need to continue the conversation. I invite you all to join us.

I want to thank you very much.