Publications
Publications
china leadership monitor
education next
hoover digest
defining ideas
policy review
March 28, 2008

Religion and Economic Development

The advantage of moderation


Reading the historian Arnold Toynbee’s lectures on the British Industrial Revolution, it is quickly apparent that conditions in England prior to 1760 were in many respects similar to those in developing countries today: Poor infrastructure and communication, lack of technological innovation, no division of labor, a focus on local commerce, and a weak banking system.1 Surprisingly, the modern study of religion and economics begins with Adam Smith’s An Inquiry into the Nature and Causes of the Wealth of Nations (1776), an examination of conditions leading to the Industrial Revolution. In his book, Smith applies his innovative laissez-faire philosophy to several aspects of religion. However, Smith’s fundamental contribution to the modern study of religion was that religious beliefs and activities are rational choices. As in commercial activity, people respond to religious costs and benefits in a predictable, observable manner. People choose a religion and the degree to which they participate and believe (if at all).

Smith’s contribution to the study of religion is not simply theoretical. He held substantive views, for example, on the relationship between organized religion and the state. Smith argued strongly for a disassociation between church and state. Such a separation, he said, allows for competition, thereby creating a plurality of religious faiths in society.2 By showing no preference for one religion over others, but rather permitting any and all religions to be practiced, the lack of state intervention (short of violence, coercion, and repression) creates an open market in which religious groups engage in rational discussion about religious beliefs. This setting creates an atmosphere of “good temper and moderation.” Where there is a state monopoly on religion or an oligopoly among religions, one will find zealousness and the imposition of ideas on the public. Where there is an open market for religion and freedom of speech, one will find moderation and reason.

A contemporary of Smith (though they were not acquainted) and a public intellectual during the British Industrial Revolution, John Wesley had much to say about the relationship between religion and economic development, though his perspective differed radically from Smith’s. Wesley (1703–1791), a theologian and the founder of Methodism and the Holiness Movement, championed the two-way causation between religion and economic growth, preaching in 1744, “Gain all you can, Save all you can, Give all you can.” Later, in his famous sermon of 1760, “The Use of Money,” Wesley expounded upon these three points, emphasizing hard work, self-reliance, and mutual aid. Finally, just two years before his death (he lived to be 88), Wesley berated his congregants from the pulpit for their comfortable lifestyle and urged them to give away their fortunes. In the 45 years between these two sermons, Wesley’s followers, by working hard and saving, had raised themselves up into the comfortable middle class. Wesley understood very well the direct causal relationship between religious beliefs and productivity. He also understood well that wealth accumulation could weaken religiosity both in terms of beliefs and participation. Wesley concluded that economic growth was detrimental to religion. Is it? And, if so, must it be?

The two-way causation

Let us look at the two-way causation and, thereby, the relationship between religion and development. First, how does a nation’s economic and political development affect its level of religiosity? When we look at the effects of economic development on religion, we find that overall development — represented by per capita Gross Domestic Product (gdp) — tends to reduce religiosity.3 The empirical evidence supports, to a degree, the secularization thesis which holds that with increased income, people tend to become less religious (as measured by religious attendance and religious beliefs). Economic development causes religion to play a lesser role in the political process and in policymaking, in the legal process, as well as in social arrangements (marriages, friendships, colleagues). There are four primary indicators of the influence of economic development on religion.

Economic development implies a rising opportunity cost of participating in religious services and prayer.

Education. The more educated a person is, the more likely he is to turn to science for explanations of natural phenomena, with religion intended to explain supernatural phenomena and psychological phenomena for which there is no rational explanation. According to this view, the higher the levels of educational attainment, the less religious people will be (negative effect). On the other hand, an increase in education will also spur participation in religious activities, because educated people tend to appreciate social networks and other forms of social capital. Education increases the returns from networks and networking. On this view, religion is just another type of social capital (positive effect). Thus, we cannot conclude that richer societies are less religious because people are better educated.

Value of time (measured by effects on per capita GDP). Economic reasoning tells us that anything that raises the cost of religious activities would — ceteris paribus — reduce these activities. We know that economic development and participation in the workforce raise the value of a person’s time as measured by the value of market wages. Thus, economic development implies a rising opportunity cost of participating in time-intensive activities, such as religious services and prayer. Hence, people will participate less in religious activities because their time is now more valuable to them. So, as a country’s per capita gdp increases, we expect to see a decrease in participation in formal religious activities. Older people and young people — in other words, those persons with a low value of time — will tend to participate more in religious activities.

Life expectancy. People are living longer all over the globe, not just in industrialized countries. Longevity has been rising almost everywhere in the world. Since 1950, it has climbed by larger absolute and percentage amounts in poor countries. With people living longer, participation in certain religions will be low and then rise as the population ages.

Urbanization. Urbanization is another aspect of economic growth that is said to have a substantial negative effect on religious participation. Why? Because in urban areas religious activities compete with others, such as the symphony, theatre, museums, and volunteer activities. Thus, religion takes up your leisure time and competes with other leisure activities, not just work.

We know empirically by doing cross-country analysis that per capita gdp has a significantly negative effect on religion, both in terms of beliefs and participation. This tendency is gradual as countries grow richer. Furthermore, a steady pattern of secularization only applies to a few countries, such as Britain, France, and Germany. Although religiosity declines overall with economic development, the nature of the interaction varies with the dimension of development. For example, increased education has very different effects on religious participation and religiosity from rises in life expectancy or urbanization.

Second, how do religion and religiosity influence economic performance and the nature of political, economic, and cultural institutions? We find that, for a given level of religious participation, increases in core religious beliefs — notably belief in hell, heaven, and an afterlife — tend to increase economic growth. Our interpretation, reminiscent of Max Weber’s famous thesis in The Protestant Ethic and the Spirit of Capitalism, is that religious beliefs raise productivity by fostering individual traits such as honesty, work ethic, and thrift. In contrast, for given religious beliefs, increases in church attendance tend to reduce economic growth. We think that this negative effect reflects the time and resources used by the religion sector as well as adverse effects from organized religion on economic regulation — for example, restrictions on markets for credit and insurance. To put it another way, the main growth effect that we find is a positive response to an increase in believing relative to belonging (attending). Striking patterns of relatively high belief appear in the Scandinavian countries, Britain, and Japan. Although these countries are not generally viewed as religious, the belief levels are high when compared to the low levels of attendance at formal religious services. Countries with low levels of belief relative to religious participation are Latin American nations and India. We also have some evidence that the stick represented by the fear of damnation is more potent for growth than the carrot from the prospect of salvation.

Now let’s look at how religion influences the four primary indicators of economic development.

Education. We find that religious beliefs are compatible with increased education and knowledge. Religion is attractive to people with higher levels of educational attainment because religious beliefs can be neither proved nor disproved. Educated people engage in speculative reasoning and are better able to think abstractly. Therefore, religion can offer something to them.

Religious beliefs matter for economic outcomes. They reinforce character traits such as hard work, honesty, thrift, and the value of time. Otherworldly compensators — such as belief in heaven, hell, the afterlife — can raise productivity by motivating people to work harder in this life. The Calvinist view of salvation through grace posits that since you cannot know whether or not you are saved, you work conscientiously your whole life (a life of good works). Religious rewards — such as absolution of sin, earning salvific merit by giving to charity — also motivate people to work hard and cultivate virtuous behavior.

Religious rewards motivate people to work hard and cultivate virtuous behavior.

Certain religions, such as Judaism, which highly value the reading of sacred texts early in life, value education. It is theorized that Jews invested in human capital because they were not permitted to own property, or, if they were, there existed insecure property rights. Human capital is “portable”; it migrates with you. The economist Evelyn Lehrer, who studies the relationship between economic development and religion, has found that the mean years of schooling is highest for Jews (16.9 males; 15.8 females); lowest among conservative Protestants (13.3, males; 12.9 females); with Roman Catholics (14.3 males, 13.7 females) and mainline Protestants (14.5 males; 14.0 females) in between.4

Lehrer also finds that Jewish women attain very high levels of education. When female education is valued equally with male education, we find lower fertility rates. The small size of families means that more is invested in each child during the early, formative years. We find this across religions with increases in educational attainment by parents. That is, the smaller the family, the higher parental levels of educational attainment will be, and the more parents will invest in their children.

Conservative Protestants value educational attainment less than the other principal religion groups. One reason for this might be the historic lack of conservative post-secondary educational institutions in the United States. Conservative Christians as a group traditionally have not placed as high a value on education as other religious groups. For example, of the 102 U.S. members of the Council for Christian Colleges and Universities, the median year of founding is 1907. The Third Great Awakening, (1850–1907) led to the founding of 52 Christian educational institutions as seminaries, Bible institutes, and missions schools. Geographically, they were primarily located in the South and Midwest.

Value of time (measured by effects on per capita GDP). The main estimated growth effect is a positive response to an increase in believing relative to belonging (or attending). The results show that, for given religious beliefs, increases in church attendance tend to reduce economic growth. In contrast, for given church attendance, increases in some religious beliefs — notably otherworldly compensators — tend to increase economic growth. To put it another way, the main growth effect that we find is a positive response to an increase in believing relative to belonging (attending). A certain amount of participation in religious activities is positive, in that people acquire certain religious beliefs. But, if people spend too much time in religious activities, there is a negative effect on economic growth. As noted above, in prosperous Scandinavia as well as Britain and Japan, belief levels are high relative to the low levels of attendance at formal services.

Life expectancy. In certain religions, such as Hinduism, we know that the last stage of life is reserved for religious activities, free from familial and societal obligations. The Roman Catholic cycle of “sin, repentance, atonement, release, and sin again” can also predict a surge in church attendance at the end of one’s life. Empirically it shows up across religions that as people age, their participation in religious activities increases, although not necessarily in attending religious rituals. Hence, in religions that permit a person to put off religious activity to the end of his life a reduction in religious participation occurs earlier on in life. This means that people spend more of their productive years at work rather than in religious activities.

The influence of religious participation and religious beliefs on life expectancy can be positive. For example, religious participation among young people is correlated with a lower probability of substance abuse and juvenile delinquency.5 Religious participation and beliefs have a salutary effect on health and well-being. Religious participation lowers the incidence of depression.6 Religious participation is correlated with more positive attitudes toward marriage and having children and negative attitudes toward cohabitation and premarital sex.7

Urbanization. Overall, urbanization has a negative effect on religiosity, particularly in terms of participation, which tends to be higher in rural areas than in urban areas. One explanation for this is simply a lack of other leisure activities. Another is that rural dwellers need religion to explain uncertainties of nature (e.g., tornadoes, droughts, etc). Also, the daily frustrations and insecurities of living in developing countries, where you might find high levels of income inequality and unfulfilled promises of prosperity, give rise to new religions and strict orthodoxies. In other words, during the tumultuous period of rapid growth (or lack thereof) with high levels of inequality, people turn to new religions.

Why religion matters

When we look at the effects of religion on economic development, we find that attending religious activities on a monthly basis has a statistically negative effect on economic growth. We also find, according to the World Values Survey, that belief in hell is strong, with belief in heaven weaker. When looking at countries, measuring monthly attendance and belief in hell, we find that, for given levels of religious beliefs, notably in hell, heaven, and an afterlife, the effect of greater religious participation is to reduce economic growth.

When we look at religion across countries, we can see that the United States is an exception (60 percent monthly attendance; 75 percent belief in hell). It is a highly industrialized country with high religiosity — in terms of participation in formal religious activities, engaging in weekly prayer, and believing in hell. Singapore (44 percent attendance; 79 percent belief in hell) and Poland (78 percent attendance and 66 percent belief in hell) are also developed countries with relatively high levels of religious participation and belief in hell. The most religious country in Western Europe is Ireland (68 percent attendance; 53 percent belief in hell), with Italy and Spain next and the United Kingdom, France, and Scandinavia coming in last.

In Muslim countries, we find high levels of belief in hell: Iran, 98 percent; Indonesia, 99; Pakistan, 94; Turkey, 94; Nigeria, 94. However, Muslim countries show a range of levels of participation in formal religious activities with Pakistan exhibiting the highest at 91 percent.8 Analyzing data from the four waves of the World Values Survey (1981; 1990–1991; 1995–96; 2001), we conclude that Muslims are more likely than Catholics, Hindus, Buddhists, and Protestants to profess a belief in heaven and hell. These otherworldly compensators tend to play a larger role in Islam than in other religions.9 Given that Protestants tend to believe in the eternal nature of heaven and hell with no intermediate degrees, and that a believer has only this lifetime to earn salvation, one would think they would express a higher belief than Muslims in the ends of the afterlife.

A possible explanation for the finding might be that whereas Christianity, and particularly Protestantism, places emphasis on individual responsibility for one’s religious obligations, Islam is legalistic, stressing the fulfillment of laws that are communally enforced. The laxness of communal enforcement of religious beliefs in Protestantism creates an individualist approach to religious living, a focus on the inward, personal relationship with God. Communal enforcement in Islam stresses outward expressions of one’s religiosity and accountability for one’s actions to others. Therefore, in Islam belief in heaven and hell is reinforced through a communally shared understanding of life after death.

Economic development and religious terrorism

commonly held view is that religious violence occurs because of poverty and high income inequality in a society or geographic region. Economist Alan Krueger’s research challenges this assumption.10 He finds that suicide bombers tend to be well-educated and not from the poorer levels of society. Islamic terrorists are more likely to be middle-class, with higher levels of educational attainment. Krueger reasons that high levels of educational attainment are a signal of a terrorist’s commitment to the cause as well as his ability to prepare for and carry out the suicide assignment. The religious terrorist is the constituent (fanatic). As Hussein Mussawi, a founding member of Hezbollah, expressed it, “We are not fighting so that the enemy recognizes us and offers us something. We are fighting to wipe out the enemy.” Not only is the constituency self-referring (oneself and one’s community of believers) but the position is absolutist. Religious terrorists view violence as an end, not a means to political or economic ends. The secular terrorist views violence as a means to a political end. The religious terrorist sees himself engaging in the destruction of a system, a system that he does not identify with but, rather, is alienated from. By contrast, the secular terrorist seeks to renovate the system, not to totally destroy it. For the religious terrorist, the situation calls for no compromise — “Our attitude is dictated by our religious beliefs” — whereas the secular terrorist is utilitarian (violence as a means to an end).

Religious terrorists feel a sense of alienation from the larger society. Krueger defines this alienation not in religious terms but in political ones. In his research, he found that the deterioration in economic conditions over time is associated with the likelihood of educated men becoming terrorist attackers. During the 1980s, unemployment rose dramatically in the West Bank and Gaza Strip for college graduates relative to high school graduates. In other words, an increase in educational attainment in the 1980s coincided with a marked deterioration in the economic opportunities for educated Palestinians, particularly males. The loss in employment was attributed by Palestinians to the Israelis and the political situation. Krueger concludes that terrorist activity on the part of educated males is a response to political conditions and long-standing feelings of indignation and frustration that are fueled by religious beliefs and less to do with economics. Krueger’s findings lead to the policy recommendation of finding a solution to the ongoing political crisis along with strengthening private enterprise, promoting job creation, and attracting foreign investment.

Deterioration in economic conditions is associated with the likelihood of educated men becoming terrorists.

Anthropologist Gananath Obeyesekere, in his analysis of the 1971 insurgency in Sri Lanka, which primarily involved Sinhalese Buddhist students, confirms Krueger’s findings.11 About 80 percent of the insurgents had a reasonable education, and had both the motivation and the capacity to participate in what they viewed as a rational choice. They saw the uprising as the only solution to the economic and political difficulties facing Sri Lanka. The students coming from the Maha Vidyalayas and Madya Maha Vidyalaya schools, established in villages in 1945 as part of the introduction of universal free education, had high expectations that they would be able to obtain a university degree and employment. By 1960, children of well-off villagers were entering the newly created universities and technical as well as teacher-training institutes. As the number of educated young men and women increased, economic opportunities remained stagnant. The supply was not meeting the demand. What happened was that the political patronage system had become closed, exclusive, and corrupt. Since there were many qualified applicants for a position, contacts and networks mattered more than merit. Very often the person who got the position was related by kin or had access to the patronage system through elite connections.

Rising unemployment meant that overqualified individuals had to accept positions that were beneath them. This translated into unfulfilled social and economic expectations, leading to increased frustration, also a key feature of Krueger’s analysis. Along with the economic situation had come an increased political consciousness. Universal franchise had come with universal education. The lack of economic opportunity, joined with political enfranchisement for a large part of the educated populace, created the right combination for a religious-based insurgency.

Islamic violent sects are embedded in inefficient or failed states with few public services and poor economies.

Economists Laurence Iannaccone and Eli Berman offer an alternative explanation, using the club model to explain extreme violent behavior. Iannaccone’s cost-benefit analysis of strict religions led to the development of a club theoretical model of the evolution of organized religion.12 Taking inspiration from theologian Ernst Troeltsch’s sect-denomination distinction, Iannaccone applied a cost-induced commitment to organized religion.13 He argued that denominations and extremist sects can be construed as distinct modes or “clubs” of religious organization based on consumer (believer) preferences. Using the club model of religion, Iannaccone sought to explain the success of strict religions (cults, sects). Using a cost-benefit analysis, he argued that people choose to undergo stigma and self-sacrifice, and engage in unconventional behavior, to eliminate free riders, thereby increasing the commitment of believers and benefits to members. Iannaccone’s economic analysis provided a rational explanation for behavior that other professions categorize as brainwashing or a form of pathological behavior.

Eli Berman applied the club model to Israeli ultra-Orthodox Jews, as well as to Hamas and the Taliban.14 Berman found in the case of the Israeli ultra-Orthodox community that the benefits of remaining in the group outweighed the costs of sacrifice and stigma. For the Taliban, the sacrifices demanded by the group included seemingly gratuitous acts of violence, destroying outside options and, thereby, increasing group loyalty. Contemporary radical Islamic sects diverge from Protestant Christian sects in that the latter are embedded in functioning economies, whereas Islamic violent sects are embedded in inefficient if not failed states where there are few public services provided and a poor economy. Protestant sects in the United States — Nazarenes, Mennonites, Brethren, Amish, and Hutterites — are embedded in a functioning state where the political and economic conditions allowed for an accommodation with the secular culture. The sects benefit from a functioning rule of law and economy. As a consequence of the political and economic stability of the society they are embedded in, Protestant radical sects perpetuate positive aspects of belonging to their religion: trust, lower divorce rates (marriage stability), education for their children, improved physical and mental health, higher levels of happiness and fulfillment.

By contrast, sects that are found in failed states tend to emphasize the negative aspects of belonging: sacrifice, stigma, exclusivity, and use of violence as a means of keeping adherents in the sect. Eli Berman and David Laitin explain this “dark side” of religion, why sects enter a destructive pattern and continue until they destroy themselves along with their society, by showing that radical religious terrorist groups provide public goods to their members who are living in a failed or weak state with relatively poor economic opportunities outside the religious sect.15 What appears to be gratuitous violence, such as the destruction of infrastructure, is in fact a mechanism for keeping adherents in the sect. Unlike Alan Krueger’s work, the policy implication of Berman’s and Laitin’s research is two-pronged: strengthening the state and weakening the religious terrorist sects. Berman and Laitin favor foreign aid as long as it comes in the form of subsidies to governments to increase the provision of public goods (education, health, security) in geographically dispersed local areas, and not just concentrated in urban ones. By providing steady employment, enforcing rule of law, and social networks (for example, opportunities to meet future spouses), the terrorist groups can be seriously weakened.

What appears to be gratuitous violence is in fact a mechanism for keeping adherents in the sect.

The club model of religious sects can also be applied to historical Tibetan Buddhism.16 The Geluk sect, of which the Dalai Lama is a member, rose to become the state religion in 1642. The Geluk sect was the last to form in a crowded religion market with schools and many sects. Unlike the existing religious sects, the Gelukpa introduced strictness in the sense of a monolithic orthodoxy and religious scholarship. They introduced “sacrifice” in the form of celibacy and required large monastic institutions to exclude lay abbots, thereby giving exclusive province to abbot monks. The Gelukpa were more successful than the other Buddhist schools and sects in Tibet in excluding the hereditary-succession practices for school leadership that underlay the usual clan politics, creating instead a corporate monastic system. In this setup, the Geluk school was able to maintain institutional independence from kinship politics. Building on its institutional strengths and its innovative theology the Geluk school was able to create a brand or trademark by differentiating itself from the existing schools and sects. These advantages, when combined with the support of (foreign) Mongol patrons allowed rituals and doctrines particular to the Gelukpa to become Tibet’s state religion in the seventeenth century.

Following the club model, the Gelukpa made a strategic alliance — with the Ölot Mongolian faction — and military victory was possible. This triumph was due to two main players, Gushri Khan, the financial advisor of the Dalai Lama, and Gendun-compel, the abbot of Drepung monastery. The Khan’s troops were seriously weakened when they were unable to take the Shigatse fortress after a siege lasting several months. It was only when Gendun-compel arrived with monk warriors (many of these men had temporarily relinquished their monastic vows so as to be able to take up arms) that the Gelukpa were victorious. The Gelukpa reacted to this triumph by consolidating their religious authority over the other schools and sects. In particular, only the Geluk school was permitted to have monasteries in and around Lhasa.

The Geluk school, as the monopoly religion firm, made the state the legitimate interpreter of the religion and its tradition. The consequence of this arrangement between political authority and religion was the imposition of the Geluk school over the other schools and sects. The Geluk state secured its hold on the religion market through government subsidies to its own monasteries and special privileges, such as the Dalai Lama permitting monasteries to conscript children of hereditary households, especially when the monastery needed novices. For smaller monasteries with few agricultural resources, the Dalai Lama instituted state subsidies in barley, butter, and tea. Eventually, the monasteries of the other schools and sects would take on institutional features of the Geluk monastic system, so much so that today it is difficult to discern how the other schools and sects might have functioned independently prior to the ascendancy of the Geluk as the state religion.

In summary, the structural features unique to the club model work well in explaining how religious sects come to engage in extreme forms of violence. The club model advanced by Iannaccone and Berman is compatible with the findings of Krueger and Obeyesekere, in that educational attainment raises expectations of economic advancement and social status. When these expectations are not met, and the political institutions fail to exercise the political will to address systemic shortcomings, then religion can serve as an organizing principle around which people mobilize for political action, often violent.

The virtuous circle

To turn once again to the question I posed at the outset: Is religion detrimental to economic growth? Perhaps John Wesley’s conclusion that economic growth will decrease religiosity need not be absolutely the case. If people spend too much time on productive activities relative to religious activities, then they will become less religious. If children are not taught religious values and beliefs, then they themselves will not be religious, and, perhaps, not productive. Furthermore, if societal expectations accompanying educational attainment are not met, people will resort to nonproductive activities, such as crime and terrorism. With nonproductive time on their hands and grievance against society, people will engage in destructive behavior. However, a virtuous cycle occurs when people believe relative to belonging. That is, people hold religious beliefs but do not spend enormous amounts of resources (time, income, talents) on their religion. Finally, religious beliefs that promote hard work, thrift, and honesty can be found across the world’s major religions. The key question is: How does a society promote these values and in what circumstances does it, intentionally or unintentionally, discard them?


Rachel M. McCleary is a research fellow at the Hoover Institution, Stanford University and a senior research fellow at the Kennedy School of Government, Harvard University.


1 Arnold Toynbee. Lectures on The Industrial Revolution in England (London: Rivingtons, 1884).

2 Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations, Sixth Edition (Strahan, 1937), 744–745.

3 The findings in this section are from research conducted with Robert J. Barro. See Robert J. Barro and Rachel M. McCleary, “Religion and Economic Growth Across Countries,” American Sociological Review 68:5 (October 2003); Rachel M. McCleary and Robert J. Barro, “Religion and Political Economy in an International Panel,” Journal for the Scientific Study of Religion 45:2 (June 2006).

4 Evelyn Lehrer, “Religion as a Determinant of Economic and Demographic Behavior in the United States,” Population and Development Review 30:4 (December 2004).

5 Michael J. Donahue and Peter L. Benson, “Religion and the Well-Being of Adolescents,” Journal of Social Issues 51:2 (Summer 1995).

6 Kathryn Harker, “Immigrant Generation, Assimilation and Adolescent Psychological Well-being,” Social Forces 79:3 (March 2001).

7 Elaine Marchena and Linda J. Waite, “Re-assessing Family Goals and Attitudes in Late Adolescence: The Effects of Natal Family Experiences and Early Family Formation,” paper presented at the annual meetings of the Population Association of America, Washington, D.C. (March 2001).

8 The latest wvs wave (2001) did administer the survey in many Muslim countries. As a consequence, the “church-attendance” data for some of these countries seem to be misleading. In particular, they seem to understate the resources absorbed by the religion. Part of this relates to how “attendance at formal religious services” is defined for women in some countries. Robert J. Barro and I have tried some alternatives to the usual “church-attendance” data (such as time spent on personal prayer and the incidence of very frequent attendance), but we think the problem is still present. The Muslim countries end up looking like they have very high belief (e.g., in hell) relative to participation, but we think this needs further, more accurate empirical research.

9 Rachel M. McCleary, “Salvation, Damnation, and Economic Incentives.” Journal of Contemporary Religion 22:1 (January 2007).

10 Alan Krueger, What Makes a Terrorist: Economics and the Roots of Terrorism. (Princeton University Press, 2007).

11 Gananath Obeyesekere, “Some Comments on the Social Backgrounds of the April 1971 Insurgency in Sri Lanka (Ceylon).” Journal of Asian Studies 33:3 (1974).

12 Laurence R. Iannaccone, “Sacrifice and Stigma: Free-riding in Cults, Communes, and Other Collectives,” Journal of Political Economy 100: 2 (April 1992).

13 Ernst Troeltsch, The Social Teachings of the Christian Church, Volume II, Olive Wyon, trans. (London: George Allen & Unwin Ltd; New York: The MacMillan Company 1931).

14 Eli Berman, “Hamas, Taliban and the Jewish Underground: An Economist's View of Radical Religious Militias,” National Bureau of Economic Research, nber Working Paper 10004 (September 2003). Eli Berman and Laurence R. Iannaccone, “Religious Extremism: The Good, the Bad, and the Deadly,” National Bureau of Economic Research, nber Working Paper 11663 (September 2005).

15 Eli Berman and David D. Laitin, “Religion, Terrorism and Public Goods: Testing the Club Model,” unpublished manuscript provided by the authors (2007).

16 Rachel M. McCleary and Leonard van der Kuijp, “A Market Approach to the Rise of the Geluk School of Tibet,” unpublished manuscript (2007).