From Scarcity to Spirituality: The Consequences of Religious Entrepreneurship *

Laurence R. Iannaccone Department of Economics George Mason University October 2002

* Very preliminary, based largely on material from a book manuscript (The Economics of Religion). My work on this paper was aided by grants from the Lilly Endowment, Santa Clara University, and the Harvard Project on Religion, Political Economy and Society. The work also benefited from conversations with Barry Chiswick, Carmel Chiswick, and Rodney Stark. Address comments or suggestions to larry@econzone.com.

From Scarcity to Spirituality: The Consequences of Religious Entrepreneurship *

Laurence R. Iannaccone

[H]uman society ... would crumble into nothing if mankind were not generally impressed with a reverence for those important rules of conduct. This reverence is still further enhanced by an opinion ... that those important rules of morality are the commands and laws of the Deity, who will finally reward the obedient, and punish the transgressors of their duty. Adam Smith, The Theory of Moral Sentiments, part III We owe it partly to mystical and religious beliefs, and ... particularly to the main monotheistic ones, that beneficial traditions have been preserved and transmitted. ... [T]he premature loss of what we regard as nonfactual beliefs would have deprived mankind of a powerful support in the long development of the extend order that we now enjoy, and ... even now the loss of these beliefs, whether true or false, creates great difficulties. F. A. Hayek, The Fatal Conceit.

I. Introduction
F. A. Hayek (1988) began his last book, The Fatal Conceit , with a characteristically Austrianeconomic assertion that our civilisation depends not only for its origins but also for its preservation, on ... the extended order of human cooperation ... known as capitalism. But he continued with the more provocative claim that this order arose from unintentionally conforming to certain traditional and largely moral practices many of which men tend to dislike, whose significance they usually fail to understand, whose validity they cannot prove, and which have nonetheless fairly rapidly spread by means of an evolutionary selection. For Hayek, this moral development is perhaps the least appreciated facet of human evolution. The evolutionary process has been cultural rather than biological a relatively recent development operating through group selection.1 Thus, our moral traditions and customs stand between instinct and reason logically, psychologically, [and] temporally (Hayek 1988: 23). They postdate, and in many instances contradict, our instinctive impulses (including many of the altruistic collectivist impulses that humans acquired over millions of years in small huntergatherer societies). They likewise antedate, and cannot be justified by, rational reflection or scientific analysis. Based neither on instinct nor on reason, the morality of the market (i.e., the traditions, customs, and institutions that nurture and sustain the extended order) would seem too flimsy to survive. Indeed, Hayek (1988) ends his book by asking How could traditions which people do not like or understand ... continue have been passed on from generation to generation? Even if group

selection explains their ultimate survival and spread, whence did such rules of conduct originate and how were they preserved against the strong opposition of instinct and ... reason? The answer would seem to be religion. Unnatural customs are more likely to last, and thus more likely to guide group selection, when linked to magical or religious beliefs. Thus, the professed agnostic (and nearly atheistic) Austrian concludes with the claim, quoted above, that We owe it partly to mystical and religious beliefs ... particular to the main monotheistic ones, that beneficial traditions have been preserved and transmitted at least long enough to enable those groups following them to grow. This is not to say that there is any necessary connection between religion and capitalism. Throughout history and into the present, many if not most religions have promoted values antithetical to the extended order. Yet if only because selection operates in both directions, the only religions that have survived are those which support property and the family (Hayek 1988: 137, italics original) One need not accept Hayeks argument to recognize that systems of belief in the supernatural have powerfully shaped the customs, institutions, and development of many, if not all, societies throughout human history. Nor must one be a true believer (in capitalism or Christianity) to recognize that supernaturalism comes in many forms, not all of which are equally sophisticated, sustainable, or influential. After decades of inquiry and intervention aimed at promoting progress, economists of every stripe (and sect) are talking about the norms, institutions, customs, and social capital that foster or frustrate growth. Whether such talk concerns the past, present, or (foreseeable) future religion must enter the conversation. Toward a Theory of Religion: If we economists are to extend our studies to religion (even if only so as to better understand the cultural determinants of economic development), then we must find a way to address and analyze religion that is consistent with our approach to other aspects of human behavior. Although the resulting theory will almost certainly fail to comprehend religions true scope and complexity, it nevertheless offers the prospect of methods, models, and results that economists can truly use. At the same time (and this, I confess, is what most interests me) economic theory offers a perspective on devotion, doctrine, and the institutions of faith that enhances the large, but largely un-economic, body of existing work on religion, most notably the work of sociologists, anthropologists, and political scientists. As with other incursions of economics into non-market behavior, the economics of religion aspires to a high degree of generality and a broad range of application. Economic thinking is distinctive, and even its paid practitioners must labor to apply its insights to daily life. To grasp the economic approach to religion, one must work through numerous examples and applications. I have therefore organized this essay as one might a tour, starting with broad principles, and visiting numerous examples along the way. Despite my introductory remarks about Hayek and economic development, my goal in this essay is to sketch a general economic theory of religion, rather than a theory of religions economic consequences.

I begin with a general theory of supernaturalism, then turn to magic and religion distinct submarkets that emerge to satisfy peoples demand for supernatural goods and services. The structure and products of these two sub-markets prove to be quite different, and the spread of science undermines faith in magic much more readily than faith in religion. The final sections of the paper focus on religious teachings (such as doctrines of heaven and hell) and social factors that affect their general popularity. The Economic Approach: The economics of religion is characterized by an approach rather than a domain of study. As with other applications of economics theory (Becker 1976: 5), it rests upon the combined assumptions of maximizing behavior, market equilibrium, and stable preferences used relentlessly and unflinchingly. The approach thus views people as rational religious consumers. With an eye toward costs and benefits, they choose how extensively to participate in religion and what religion (if any) they will embrace. Over time they may substantially modify their religious choices and, hence, their religious activities. They may do so even if their tastes, norms, or beliefs do not change. Indeed, the economic approach views most behavior (and most interpersonal differences) as rational responses to varying prices, incomes, skills, experiences, technologies, and endowments. Similar assumptions apply to religions supply side. Religious producers maximize members, net resources, government support, or some other basic determinant of institutional welfare. The actions of churches and clergy (or denominations and rabbis) are thus modeled as rational responses to the constraints and opportunities found in the religious marketplace. The combined actions of religious consumers and religious producers form a religious market which, like other markets, tends toward a steady-state equilibrium. As in other markets, the consumers freedom to choose constrains the producers of religion. A seller (whether of automobiles or absolution) cannot long survive without the steady support of buyers (whether money-paying customers, dues-paying members, contributors and coworkers, or governmental subsidizers). Consumer preferences thus shape the content of religious commodities and the structure of the institutions that provide them. These effects are felt more strongly where religion is less regulated and competition among religious firms is more pronounced. In a highly competitive environment, religions have little choice but to abandon inefficient modes of production and unpopular products in favor of more attractive and profitable alternatives.

II. Origins: Supernaturalism

As the previous paragraphs suggest, standard economic insights have much to say about religion. The economic theorist can posit a demand for (suitably-defined) religious commodities and jump directly to standard insights about production, consumption, supply, demand, cost, benefit, monopoly, competition, regulation, and laissez-faire. Most contributions to the economics of religion do just this, sidestepping debates about the true nature of religion, the foundations of individual faith, and the structure of religious institutions. (See, for example, Neuman 1986.; Chiswick and Lehrer 1991; Iannaccone 1992)

A different approach is needed, however, to construct a general theory of religion one with relevance for many religions, in many times, cultures, and places. To explain the fundamental features of religious practice and religious institutions, religious commodities must be distinguished from all other commodities, and the demand for religion must be distinguished from all other demands. The following paragraphs review a theory of religion that I develop more fully in a forthcoming book. The theory is by no means complete. It does, however, get beneath the standard givens of economic analysis, working from more basic assumptions about the physical world and human nature. It defines religion, magic, and supernaturalism in terms that are broad yet amenable to economic analysis.2 It accounts for the universality of supernaturalism, the differences between religion and magic, the appeal of specific religious doctrines, and the tendency for science and technology to displace magic more readily than religion. Thanks to the meticulous work of anthropologists, sociologists, and historians, we can confirm many of the theorys predictions, drawing from a wealth of historical and ethnographic data concerning primitive, ancient, and non-Western cultures. Scarcity, Rationality, and the Demand for Supernaturalism: The theory presumes a human tendency to identify most everyday phenomena as natural or normal. Definition: The SUPERNATURAL refers to forces or beings beyond or outside this natural order, which can suspend, alter, or ignore the normal flow of events. The theory also presumes scarcity and rationality. In the beginning, in the end, and everywhere in between, there is no such thing as a free lunch, no escape from scarcity. Humans, however, are not equipped simply to accept scarcity, but are instead driven to satisfy their wants and are equipped to do so with unparalleled efficacy. The pressures of survival and the powers of their minds motivate people to defy their apparent destiny and strive for longer, more comfortable, and more meaningful lives. Most efforts fail, but some succeed spectacularly, as with the development of cooking, hunting, agriculture, writing, and the wheel, or democracy, law, literature, and art, or, in more recent times, physics, biology, chemistry, engineering, mathematics, medicine, and computing. Each of these magnificent inventions may be viewed as a technology to expand our opportunity sets and push back the limits of scarcity. Against this background, supernaturalism emerges as an alternative technology a natural, understandable, and perhaps even fully rational attempt to enhance individual and collective welfare. Supernaturalism posits the existence of mystical powers that transcend the normal physical limits of everyday life. If it is possible to influence, harness, or benefit from such powers, then the supernatural provides yet another way to combat scarcity. To be sure, its existence and efficacy eludes positive proof a fact that did not escape the notice of ancient primitives any more than it has that of modern scientists. But uncertainty cannot quench its underlying attraction, for by definition the supernatural holds forth the prospect of otherwise unattainable rewards.

Demand for the supernatural can be viewed as a reasonable response to inescapable scarcity, insatiable wants, and irrepressible hope. Supernaturalism confronts humans with the broadest form of Pascals Wager (Pascal 1910, 233). For anyone willing to grant the supernatural some probability of existence, it almost certainly makes sense to try at least one thing that one would not otherwise do offer prayers (as even most atheists admit to saying), practice rituals, embrace beliefs, wear charms, worship, meditate, study, dance, chant, conjure, or sacrifice. As with normal behavior, these acts of religion and magic increase ones expected utility wherever their costs are lower than their postulated benefits times the subjective probability that the benefits will, in fact, arise. Supernatural investments make sense, not just for the credulous, fanatic, and faithful, but also for the hopeful skeptics and cautious agnostics anyone who does not know the skies are empty. We thus arrive at a proposition that turns the standard notion of faith on its head: Conjecture: Rational individuals will seek to understand and influence the supernatural to the extent that they remain uncertain of its NON-existence. The point is not to defend supernaturalism, nor discredit it, but merely to emphasize its underlying economic rationale and universal attraction.3 Neanderthal and Cro-Magnon man left cave paintings, burial sites, and carvings that virtually all archeologists take as proof of religious beliefs. Every historic culture has left a rich record of supernatural beliefs and practices in temples and totems, traditions and texts, and art and administration. Supernaturalism is no less universal in the hundreds of preliterate societies observed and catalogued by 19th and 20th century anthropologists. Supernaturalism likewise prevails in every society of todays developed world. Although Americans seem to hold the record for religiosity, with more than 95% professing belief in God or a Universal Spirit, the supposedly secular populations of Europe maintain remarkably high levels of faith in supernatural phenomena. In every European country (except the former East Germany), atheists number fewer than fifteen percent of the population (Jagodzinski and Greeley 1997). The resilience of religiosity is nowhere more evident than in Russia, where despite sixty years of state-sponsored atheism, bloody church purges, widespread religious persecution, and pervasive secular indoctrination, only eight percent of the people remained atheistic by 1990 and religious belief is especially widespread among the young (Troyanovsky 1991; Filatov 1993: 120). The vast majority of people find atheism intellectually demanding and emotionally unsatisfying. One must push well beyond the uncertainty of the agnostic or the inactivity of the apathetic, only to be rewarded with the absence of hope. No wonder the children of atheists are more likely to abandon their parents beliefs than are the children of any (other) religion (Stark and Bainbridge 1985: 47-48). An economic perspective thus helps us to understand one of the great predictive failures in the history of the social science: the failure of secularization theory. For centuries, scholars have confidently, and incorrectly, predicted the immanent demise of religion. With powerful prose, but no real proof, Hume, Comte, Marx, Freud, the Huxleys, and other influential intellectuals, pronounced religion a dying vestige our primitive, pre-scientific past (Iannaccone and Stark 1998). Within academia, the media, and most public discourse, this received wisdom sustained a

large body of stylized facts. For example: that religion must inevitably decline as science and technology advance; that individuals become less religious and more skeptical of supernaturalism as they acquire more education, particularly more familiarity with science; and that most members of deviant cults and fundamentalistic sects are victims of indoctrination or abnormal psychology (from trauma, neurosis, or unmet needs). In contrast, decades of empirical research, using methods that range from cross-cultural comparisons, to cross-sectional surveys, leave no doubt that these generalization are simply false (Stark and Bainbridge 1985; Hadden 1987; Greeley 1989). Throughout the world in Latin America, India, Africa, Europe, the Middle East, formerly communist countries, the United States, and even Israel religion continues to flourish, especially its more fundamentalistic variants. In short, the demand for supernaturalism is as basic and irrepressible as the wants it seeks to satisfy.

III. Development: Professionalism

What many demand, others will gladly supply. The consumers hope for otherwise unobtainable payoffs provide ample opportunities for producer profit. Pie in the sky regularly sells for cash on the barrel. Markets for religion and magic thus develop as naturally as other markets, and sooner in cultural history. Self-interest and other standard economic forces promote professionalization. Supernatural specialists arise to exploit the gains from trade, specialization, and entrepreneurship. These would-be priests, prophets, and seers fan to flame the sparks of faith. Whether prompted by conviction or the lure of profit, prestige, and power, their activities promote belief and shape its form. Specialization leads, in turn, to institutionalization and product differentiation. Groups of religious professionals establish firms and franchises, offering distinctive services, terms of trade, apprenticeship programs, rites of initiation, and rules of membership. The organizational features we associate with religion in contemporary Western countries, including congregations, denominations, and clergy, are strictly analogous to those of secular industries. On the other hand, markets for the supernatural have special features, which trace back to the unique character of their underlying technology. Three deserve emphasis: First, the market is naturally competitive, with few barriers to entry or imitation, modest returns to scale, and a range of promised products for which the sky (literally) is the limit. Backed by coercive power, a single religious firm may dominate the market for legitimate religion, but diversity and competition almost always persist in an underground economy (as illustrated by Biblical Israels failure to root out pagan worship or the Medieval Catholic Churchs inability to eliminate witchcraft, heresy, and sect movements). Second, the market offers a uniquely wide range of products. In principle and practice, no desire, however large, small, abstract, or specific lies beyond the range of the supernatural. People will invoke the supernatural for everything. Individually, they will pray (and sacrifice, and chant, and meditate) for physical health, emotional support, information and advice, prosperity, good fortune, favorable weather, protection from enemies, plentiful harvests, life-

after-death, romance, revelations, and, yes, good sex. 4 Collectively, they will also petition the supernatural for peace, prosperity, domestic tranquillity, victory in battle, and a bright future. And those in positions of power will, regardless of their personal beliefs, will publicly invoke the supernatural to motivate (or terrify) their subjects, maintain loyalty, legitimate their power, and reinforce the prevailing social order. There is thus no logical limit to the size or scope of rewards that humans will seek through supernatural means.5 Jeffersonian Deists and Unitarian preachers may proclaim a distant God, who forswears miracles and magic, but rank-and-file believers the world over seem happy to seek every conceivable blessing material or spiritual, miraculous or mundane, specific or general, and here-and-now or by-and-by. Third and last, the market must address problems of information and uncertainty. The existence and efficacy of supernatural technologies remain always a matter of faith. The prospect of otherwise unobtainable rewards (including eternal life, peace on earth, and unending bliss) is tied to tremendous uncertainty. In the language of economists Darby and Karni (1973), religion offers the ultimate in credence goods, because no amount of experience (this side of death) suffices fully to evaluate its quality and efficacy. Insights from the economics of uncertainty help us understand the activities and institutions that arise in response to this dilemma. In particular, two different strategies for managing risk diversification and specialization arise in both secular and religious markets. As we shall see, the different strategies lead to different forms of religious organizations (inclusive versus exclusive) specializing in different types of products and different patterns of practice.

IV. Differentiation: Religion versus Magic

Of all the factors that promote product differentiation and sub-market formation in the market for the supernatural, none are more striking than those that separate religion from magic. These two conceptions of the supernatural imply two different modes of interaction, leading to differences in form, content, outcomes, and organizational structure. Definition: MAGIC refers to beliefs and practices concerning impersonal supernatural forces or the impersonal manipulation of supernatural beings. RELIGION refers to beliefs and practices concerning one or more supernatural beings. My use of the words magic and religion parallels that of 19th century anthropologists. Religion consists of beliefs, practices, and institutions that relate to one or more supernatural beings. The one great God of Israel, the many small gods ancient Greece, the ancestral spirits of Chinese Shinto worshippers, and, I suppose, the ghost of Christmas past all qualify as supernatural beings. At a minimum, every such being possesses consciousness and desire in addition to superhuman powers (Stark and Bainbridge 1987: 82). Magic, by contrast, consists of beliefs, practices, and institutions that concern impersonal supernatural forces or the impersonal manipulation of supernatural beings. When astrologers cast horoscopes, or athletes carry good luck charms, or fortune-tellers read the future from a deck of cards, their activities qualify as magic. Most New Age religion likewise qualifies as magic, be it crystals, pyramid power, palmistry, or mystical auras.6 On the other hand, Las Vegas-style illusions do not qualify as magic because they are not viewed as genuine manifestations of supernatural power.

The distinction between religion and magic might seem arbitrary or unimportant just one of many ways to divide the supernatural domain. It is not. Humans have no choice but to describe the supernatural world in terms derived from the natural world. No other vocabulary is available to them. Language, thought, and experience thus lead people to characterize supernatural phenomena as matters of personal versus interpersonal agency. We can envision other-worldly forces or other-wordly beings, but thats about it. From an economic perspective, magic and religion yield benefits in fundamentally different ways. Magic involves supernatural production; religion involves supernatural exchange. Production is impersonal; specific inputs, manipulated in specified ways, yield specific outputs. Exchange is, by contrast, interpersonal, a process of communication and bargaining that involves character no less than competence. Here again, magic and religion recapitulate a natural distinction: the two ways humans overcome scarcity in everyday life. The magical specialist is, by definition, a kind of supernaturally-oriented technician, but a technician with a problem for by all accounts (scientific, ethnographic, and often even autobiographic) he wields no truly special powers. Rain dancers cannot influence the weather, Tarot readers cannot see the future, and witches cannot fly with or without their brooms. Although good magicians will hide this fact, perhaps even from themselves, they cannot avoid its consequences. And the most important consequence is that when put to the test (a proper test that is), magic will fail. Magicians must therefore seek to avoid tests.7 The logic of production traps magic. On the one hand, magicians must promote themselves as masters of the supernatural. On the other hand, they cannot properly account for their failures, except to emphasize their personal limitations or the limitations of the forces they control, neither of which plays well in the market. No magician can fully escape the dilemma: impersonal forces do admit testing, and effective methods for their control should work on demand. True magicians should succeed here and not just there, now and not just then. Yet by all accounts, scientific, ethnographic, and even autobiographical, they do not. Religion avoids the trap of magic. In a religious system, maximizing humans seek to interact with maximizing gods supernatural beings with their own preferences. This makes religion less susceptible to refutation, because failure can be explained in terms of the gods autonomy and personality. The implications, as summarized below and in table 1, are immense: Religious practitioners can ask whereas magical practitioners demand: Humans approach the gods as supplicants listening, petitioning, hoping, and waiting, rather than summoning, compelling, projecting, and manipulating. Religion can excuse failure while taking credit for success: A priest, pastor, or petitioner can always counter frustration or disappointment with the thought that God did hear and did answer, but the answer was No. Religion can focus on the future: The gods are not obligated to deliver on demand or even in this life. They can take their time, promising only that it shall come to pass (Isaiah 2:2).8 Religion can remain general: The gods can avoid specifics, even as they promise to sustain, nurture, and save. 8

TABLE 1: MAGIC VERSUS RELIGION

Magic
supernatural forces demanding/compelling utilitarian/matter-of-fact a-moral individualistic client-oriented episodic/crisis-oriented secretive fee-for-service here-and-now small/specific rewards

Religion
supernatural beings asking/pleading idealistic/systematic normative/moralistic collective congregational cyclical/routine open/public contributions and other forms of funding future orientation large/general rewards (and small/specific)

Religion can refuse tests: Gods can play by their own rules. Thus, the same biblical tradition that scoffs at magicians, applauds the king who tests them, and proclaims its superior efficacy, can nevertheless denounce those who would test the Lord your God (Deuteronomy 6:17, see also Daniel 2-3). Religion can demand sacrifice, whereas magic can only demand technique: The gods can ask things of you as a matter of need, preference, or concern for your own good. Religion can ask why; magic remains matter-of-fact: Religion can produce systematic theology (though this is certainly more characteristic of Western monotheism than other religious traditions). If only to protect itself against refutation, magic tends to be highly unsystematic. Religious specialists can work clearly, openly, publicly, and regularly: With all the protections described above, religion has relatively little to hide. It can offer instruction instead of mysterious mumbo-jumbo, and it can regularly meet in public rather than work in secret. Magic centers on one-on-one client-practitioner relationships governed by the customers sporadic needs and satisfaction with the specific services rendered. The technology of religion thus serves to promote group activities, intense commitment, longterm relationships, loyalty, exclusivity, contribution-based financing, and collective goods. These, in turn, sustain religions most important contributions to culture: collective identity, shared values, and individual moral constraint.9 Some scholars have, in fact, gone so far as to define religion in terms of its collective dimension. Arguing that [i]n all history we do not find a single religion without a Church, mile Durkheim (1965: 59, 62) proceeded to describe religion as a unified system of beliefs and practices which unite into a single moral community called a Church, all those who adhere to

them. Though other scholars would dispute whether all religions revolve around communities, few can reject his complementary observation that There is no Church of magic (italics original). Comparing religion and magic, Durkheim (1965: 44-45) correctly noted that:
whenever we observe the religious life, we find it has a definite group as its foundation It is quite another matter with magic. Between the magician and the individuals who consult him, as between these individuals themselves, there are no lasting bonds which make them members of the same moral community, comparable to that formed by the believers in the same god The magician has a clientele and not a Church, and it is very possible that his clients have no other relations between each other, or even do not know each other A Church is a moral community . But Magic lacks any such community.

In short, despite the fact that many religions that fail (by any measure of success), there can be no doubt that religion sustains commitments, cultures, and institutions far richer than those of magic. Historical and cross-cultural records highlight the universality of this proposition. The growth of science and technology help explain why it is especially true today.
SCIENCE VERSUS THE SUPERNATURAL:

Secularization: As I have shown elsewhere (Iannaccone and Stark 1998), generations of Western scholars, from the 18th century through the present, have proclaimed the death of the gods in the face of science.10 Yet notwithstanding the many eloquent, emphatic, and apparently wellreasoned claims of these distinguished scholars, and notwithstanding the widespread public impression that religion has declined over time, religion has in fact shown very little sign of serious, long-term retreat, much less immanent or eventual death. The continuing vitality of religious faith and institutions underscore the inadequacy of traditional approaches to the socialscientific study of religion The general theory of religion and magic outlined above offers an alternative to secularization theory, and indeed a new approach to the social-scientific study of religion. It explains, among other things, the persistence of demand for the supernatural even after of centuries of scientific discovery yielding a vast increase in human knowledge and incredible technological progress revolutionizing virtually every aspect of our lives. It also explains why, despite continuing demand for both religion and magic, the long-run prospects of religion are much more favorable than those of magic. Scientific progress does alter the market for supernatural goods and services, but its effects are most strongly felt by magic (and the magical elements of religion). Magic versus Science (again): As we have seen, magic cannot overcome its fundamental weakness weakness grounded in the logic of impersonal supernatural forces. Despite its superficial similarities to science (including a matter-of-fact approach, a utilitarian orientation, and a penchant for elaborate procedures, as opposed to poetry, myth, and narrative), magic cannot withstand scientific scrutiny. Its methods consistently violate the canons of science they are intuitive rather than empirical, analogical rather than analytical, and secret rather than open to peer review. Its claims fail in the face of observation and attempted replication. The spread of science and technology thus tend to undermine the demand for magic.

10

Religion, by contrast, has much less trouble coexisting with science. It persists, not because it can be proved to work, but because it cannot be proved to not work. It remains attractive, not because it looks like good science, but because its approach and claims are so obviously distinct from those of science. Having already analyzed the relative strengths of religion over magic, particularly its resistance to refutation, it suffices draw the obvious conclusions: Conjecture: As people become better acquainted with the character and content of science, the methods of magic come to look more and more anti-scientific. The spread of science thus tends to undermine faith in magic but does little to reduce faith in (nonmagical) religion. Together, the two observations above suggest a new and modified form of the secularization hypothesis: Conjecture: As science and technology grow, the relative status (and, probably, absolute prevalence) of magic will tend to decline. The status and prevalence of religion need not change, however, although its overtly magical aspects will tend to diminish. Historical accounts appear to support this generalization. See, for example, Keith Thomas (1973) study tracing the decline of magic in England. Changes in Catholic doctrine and practice, from medieval times through the present, follow a similar path. D. Michael Quinn (1987)(zz) documents the prevalence of magic in 18th and 19th century America (as opposed to modern times). Change even appears within American Protestant denominations, whose Colonial and frontier churches promoted many magical teachings and practices that their successors now utterly reject. Quantitative data seem also support the argument. As I have noted elsewhere (Iannaccone and Stark 1998), virtually all time series data refute the notion that religious commitment and religious involvement have declined over time (as education, income, or familiarity with science has increased). There thus there appears to been no decline in religion that parallels the decline in magic. Moreover, contemporary cross-sectional surveys, such as Gallup polls and NORCs General Social Surveys, almost never find a significant negative relationship between measures of religiosity (such as church attendance and belief in God) and measures of income, education, or familiarity with science and technology. In contrast, however, people with higher levels of education do display less belief astrology and other popular forms of magic.

Recap: Sections I - IV
The general theory of supernaturalism refutes once and for all the common complaint that rational choice models have little relevance beyond the consumer cultures of the developed West. Indeed, our brief forays into the anthropology of religion and magic in non-Western and primitive societies suggests that the economic perspective may be even more relevant in these societies than it is to Christianity, Islam, and Judaism the great monotheistic religions of the contemporary West. Ironically, it is Western monotheism that is, in principle if not in practice, most hostile to the simple economic notion that through religion, humans may trade with god(s) in order to accumulate current and future material benefits. Although thinly veiled trade still

11

exists within Christianity, Islam, and Judaism, materialism is routinely denounced (again, in principle, if not in practice) and Gods terms of trade are carefully nuanced. Economic considerations help us understand why supernaturalism persists in every known time and place. They likewise help us understand the emergence of religion and magic as contrasting orientations toward the supernatural. The different orientations imply that, despite their apparent similarity, magic tends to be more refutable than religion and its institutions more fragile. In practice, religion and magic will tend to concern themselves with different problems, and the markets for religion and magic will be structured very differently. Most importantly, magicians will not establish churches or congregations, merely collections of clients, whereas religion will abound in churches, drawing their strength and numbers from the moral communities that religion can sustain. The theory also addresses the relationship between science and supernatural, and leads to an important prediction: scientific advance will tend to undermine magic far more readily than religion. Although this scaled-down variant of traditional secularization theory lacks the boldness and scope of its predecessor, it does at least enjoy the advantage of empirical support. In sections to follow, I will extend the theory of religion and show certain doctrines tend to predominate in religious markets. There are, for example, very good reasons why most religions offer their best clients large afterlife rewards. Conversely, afterlife penalties, while not the least bit attractive to those on their way to Hell, can be of use in constraining peoples behavior and will, for this reason, be promoted by various groups and elites. Having traced the appeal of certain doctrines to the characteristics of individuals and groups, we can also explain the conditions under which one type of doctrine gives way to another, including the conditions that have led contemporary Christians to largely reject the traditional view of Hell.

V. The Market for Doctrine: Teachings That Sell, Pay, and Evolve
Within all religious traditions, but especially those known for their tendency to create moral communities, doctrines concerning Gods will provide the foundation for numerous demands and activities. Such teachings are usually sold in terms of their benefits to the religious consumer (though the terms of exchange are rarely stated as such). We turn now to the market for such teachings, focusing as always on perceived costs and benefits, and starting with a definition of doctrine precise enough for economic analysis, yet broad enough to cover most actual religious teachings. Analyzing Doctrines The typical religious teaching is an assertion or assurance of the form A leads to R, where A specifies a set of actions (including, perhaps, subjective attitudes) to be undertaken by the supplier and/or demander, and where R is the promised set of results. The action A might thus be lay a sacrifice at the foot of the god Zogar, and the result R might be an increased probability of a good harvest. Formally, we may denote any such assertion with the symbol, [A,R]. Although this semantic structure characterizes virtually all cost-benefit assertions,

12

whether secular or supernatural, our interest centers on religious assertions, which by definition concern results achieved through divine agency.11 The validity of most religious assertions is, as we have seen, difficult or impossible to determine since the promised results will usually depend upon the consent of god(s), involve very general benefits, and/or occur in the distant future.12 To reduce all doctrine to assertions of the form [A,R] is, of course, a simplification, but not as great a simplification as might at first appear. Religious statements come in many forms, as the following examples illustrate, but most are at least implicitly equivalent to [A,R] assertions. Consider, for example, the following broad classes of statements: 1) Cause and Effect: These statements fall directly within the [A,R], action/result framework. They arise whenever a religious authority demands or advises an act or attitude so as to produce a valued outcome (or discourages an act on the basis of its harmful consequences).13 2) Predictions and Promises: Though superficially unconditional, and thus of the form R without A, most religious predictions and promises involve implicit demands for future action, warnings against inaction, or assurances that R is the reward for some previous action.14 3) Simple commands and exhortations: Here the statements form seems to be A without R. Closer reading, however, of the text or subsequent interpretation almost always reveals a promised result that justifies, motivates, or accompanies the command.15 4) Simple declarations: Even most (semantically) simple declarations fit the [A,R] model insofar as they exist to build faith. Whether by original intent or subsequent interpretation, the matter-of-fact, historical statements within most religious texts and traditions fulfill this function.16 In short, regardless of their explicit form and perhaps even regardless of their original intent, the standard interpretation of a religious statement can almost always be expressed in [A,R] form. Thus, [A,R] assertions are the bread-and-butter of religious texts and teachings. This is hardly surprising, for as we shall see below, religious consumers are not likely to buy a doctrine that offers no results or unsatisfactory results, nor are religious producers likely to sell a doctrine that stipulates no actions. Evaluating doctrines: In the market for religion, success inevitably depends on developing an appropriate set of teachings, backed up by appropriate rituals, requirements, organizational structures, all of which vary depending upon the religions history and customer base. It is clear, however, that some religious assertions are inherently better suited to survive the market test a test which requires, at minimum, both willing buyers and willing sellers. The preacher-sellers: In any society there will exist people who have the opportunity to promote religious assurances of the form [A,R]. In competitive religious markets, such as that of the United States, all citizens have the right to promote their religious beliefs. In more regulated environments, such as medieval Europe, the right is far more restricted. In any setting, however, the promoter makes a sale whenever some other person agrees to undertake the action, A, in the hope of receiving the promised result, R. The required action will usually involve activities 13

on the parts of both seller and the buyer. For example, a disciple-buyer may pay the priest-seller to petition the gods for a good harvest. The buyer-adherents: Rational consumers will evaluate religious assertions in terms of their expected costs and benefits. The relevant costs are, of course, those that the buyer incurs by undertaking his or her part of the required action, A. Let the symbol C(A) designate the perceived magnitude of these costs. The magnitude of benefits will likewise depend on the value of the promised reward, R, but also on the probability, , that this reward actually materializes as a result of the action, A. The expected value associated with the assertion [A,R] is thus V(R) C(A). In most settings, a person also ascribes less utility to a more risky prospect, even when the magnitude of the payoff and probability of payoff are offsetting. To incorporate risk aversion in a formal model of doctrine, it suffices to assume that buyers evaluate doctrines not merely in terms of their expected payoff, but rather in terms of their expected utility.17 Doctrines that Sell Although the complexity of most real-world doctrines makes it difficult to assess their costs and benefits, we can identify major classes of doctrines that will not survive market pressures. For example, there will be no market for assurances, [A,R], whose required actions are more costly than the promised results. To be specific: a priest may be correct that a whole-burnt offering of two cows will cause the gods to bless your one remaining cow with more milk, but right or wrong this is not the kind of teaching that will attract a following. Similarly, there will be no market for assurances whose costly actions are promised to yield no results at all or only undesirable results. (Doctrines of eternal punishment, an apparent contradiction to this rule, are discussed at length below.) In short: Conjecture: Religious teachings will emphasize promises of beneficial results. Moreover, these results will principally benefit the buyer-adherents and those closest to them. The implications of a cost-benefit approach to doctrine are not always obvious. The following extended example demonstrates an important principle: Conjecture: It is not enough for a religion to promise large rewards, even if its potential customers believe the promise. A successful religion must offer an appropriate reward gradient, so as to guarantee greater payoffs to its more dedicated adherents. To understand the basic argument, imagine a religious market consisting of three competing churches. The first, The Church of Conditional Salvation, demands a specific lifestyle A1 that has observable, here-and-now opportunity costs, C1 . It also teaches that those who join the church and abide by its teachings will be rewarded with a blissful life after death worth R1 , whereas those who reject church and its teachings will suffer afterlife penalties, P1 (where P1 < 0 < R1 ).18 The second church, The Church of Universal Salvation, advocates a lifestyle costing C2 and promises a blissful afterlife, R2 . Unlike the first church, however, it believes in an infinitely loving God who will somehow save all (or nearly all) humans regardless of their behavior in this

14

life. Those who reject this church will thus receive rewards P2 which (nearly) equal the members rewards, R2 . Finally, there is The Church of No Tomorrow, which advocates a lifestyle costing C3 , but which denies the existence of an afterlife and so offers the same reward, R3 = P3 = 0, to those who accept or reject its teachings. This last churchs afterlife prospects are, of course, equal to those anticipated by an atheist (or, for that matter, Auguste Comte and his Church of Humanity). Only the first church can flourish in a competitive environment, because only it offers net benefits to those who accept its teachings. In either of the other churches, peoples afterlife prospects are equally good, or equally bad, whether or not they join the church. Indeed, they would do better to join no church and thereby avoid all lifestyle costs associated with church teachings. [Formal analysis: Appendix 1 and formalizes this result, and introduces a symbolic framework suitable for more complex cases. Potential church members have four options join the first church (of conditional salvation), the second (of universal salvation), the third (of no salvation), or join no church at all. Their afterlife rewards will depend on which church, if any, they join and which, if any, is true. They thus face 16 possible outcomes as depicted in the payoff matrix of figure A.1.19 Without yet knowing which church is true, each potential member must choose the membership/lifestyle option that maximizes his or her expected utility. In other words, they must each choose a column of the payoff matrix, then wait until the true row (corresponding to the true church) is finally revealed and the actual afterlife payoff (at the intersection of column and row) is finally received. But given the scenario described above, people will always maximize their payoffs by choosing the first church, regardless which church is true, For if the first church is true, they will be rewarded for choosing it and punished for choosing another, but if the second or third church turns out to be true their payoffs will be the same regardless of which church they attends. So, as in the initial example, rational people will always choose the first church (or no church at all).]20 Applications: Despite its limitations, the church choice framework highlights a serious problem facing the liberal churches of mainline Protestantism. In the ongoing competition for members, they repeatedly loose to their more conservative counterparts, not because they promise too much or too little, but because they cannot bring themselves to promise their members substantially more than they offer nonmembers. Indeed, as culturally-sensitive liberals, they hesitate to criticize any belief system be it Christian, Jewish, Moslem, Hindu, Gnostic, agnostic, or Wiccan. Such openness, however admirable, does little to motivate conversion. Atheists face the same problem, or worse, selling their beliefs. With little to offer and less to withhold, true atheists remain scarce (about 4% of the American population and fewer than 10% in most other developed countries), and rarely do their children follow in their disbelieving footsteps.21 Indeed, viewed as a type of religious identity (analogous to Protestant, Catholic, Jewish, Moslem or Hindu), atheism manifests far less intergenerational stability than any other affiliation. Let us push the analysis a bit further, confining our attention to conservative, conditional salvation churches like the first, since we have already demonstrated their tendency to dominate. Consider two such churches, each of which promises its members afterlife rewards (R1 and R2 ,

15

respectively) contingent upon membership and behaviors (costing C1 and C2 , respectively). Each also threatens its nonmembers with punishments or lesser rewards (P1 and P2 , respectively). It follows from the analysis of figure 4.1, that that potential members will tend to choose one church over the other: the higher its net afterlife benefits (R-P), the lower its lifestyle costs (C), and the more plausible (i.e., the more subjectively probable) its claims to truth. Leaving issues of cost and plausibility, we obtain the following important conclusion: Conjecture: Religions face a strong inflationary incentive to promise huge afterlife payoffs net of afterlife punishments. Truly, the sky is the limit, since no one can prove them wrong. Traditional Christian theology provides a prime example, offering faithful members eternal bliss the highest conceivable reward while threatening heretics and heathen with the greatest possible punishment eternal torment. Entrepreneurial suppliers of religion are thus led, step by step, to Heaven and Hell or their doctrinal equivalents pie-in-the-sky and fire below. My analysis of salable doctrines is, of course, oversimplified and incomplete. Yet the results are suggestive, and surprisingly predictive, just the same. Let us turn therefore to the other side of the market, and consider which doctrines are most attractive to the churches that produce them. Doctrines that Pay Costs and benefits shape the supply side of a religious market just as they do the demand side. In particular, they constrain the supplied content of religious teachings, since no doctrine will attract many preacher-sellers if it does not benefit them as well. Otherwise, why should the sellers bother to expend resources trying to convince potential buyer-adherents that the doctrine is true? Even if most preachers are so saintly as to rise above self-interest (a rare and difficult feat, even among Gods servants), and even if they are willing to preach a particular doctrine gratis, how can market their message how can they disseminate it effectively unless they obtain resources from those they have convinced? Even it is true that God blesses believers who mail all their contributions to competing denominations (and I for one have always believed it is), a church that preaches such doctrines will rarely survive, much less prosper. Here again we observe a kind of invisible hand at work, guiding the market toward self-serving doctrines: Conjecture: No matter how much a doctrine appeals to demanders, it must also attract potential suppliers. Successful doctrines must therefore generate beneficiaries who will either promote the doctrines themselves or else finance or force others to do so for them. Implications and examples: [Notes: In religion, as in other enterprises, inefficiency and pure altruism are the exceptions, not the rule. And even the gods (as distinct from their promoters) do not receive much in the way of sacrificed resources. Most sacrificial offerings are, in fact, divided between the priest and the petitioner, with little more than the smoke, intestines, and burned fat reserved for the gods. Examples of cost-conscious sacrifice abound, not only in the Hebrew Torah, but also in many other religious traditions. In many respects, Christianity lowered the price of admission further still, requiring no physical sacrifices at all, not even the traditional Jewish sacrifice of circumcision. This is, however, an overstatement, since early Christianity 16

and many subsequent Christian orders and sect movements instituted their own, high-cost lifestyle demands (such as poverty, celibacy, and absolute obedience). How they managed to succeed, and continue to manage, is the primary subject of Iannaccone (1992; 1994). Potlatch and apparently gratuitous sectarian demands thus deserve attention, not because they typical of religion, but precisely because they appear to violate the general pattern, which might be called dollars from donors, but pennies for heaven.] The supply side perspective: The economic approach continually reminds us that religious outcomes, like those of all other markets, reflect the activities and interests of both demanders and suppliers. This contrasts with traditional religious scholarship, which has all but ignored the supply-side of the religious markets, thus leading researchers and the media to erroneously interpret all manner of religious trends and innovations as the consequence of psychic shifts or socio-economic upheavals (including the so-called Great Awakenings of the 18th and 19th century America, the rise of New Religious Movements in the 1960s and 70s, and the explosive growth of Protestantism in Latin America). For more on this, see (Iannaccone, Finke et al. 1997). Here, it suffices to emphasize the critical role played by religious suppliers of all kinds the mainstream suppliers/clergy who staff conventional churches and satisfy most current religious demand, the advertisers/evangelists who market religion to lapsed or potential consumers, and the innovators/cult leaders who build demand for entirely new religious products. [An aside on the problem of plausibility (developed more fully in my book): The plausibility of a teaching (as indexed by the perceived probability of reward, , is central to our theory since it determines the expected value associated with promised results, R. It comes, therefore, as no surprise that all religious traditions emphasize the importance to faith. More importantly, all religious suppliers work to enhance the plausibility of their products. Standard ways to enhance plausibility include: testimonials, collective activity, evidence of commitment on the part of the seller, emotionally-charged meetings, childhood indoctrination, social ties to those who already believe, and the use of powerful images (derived from narrative, myth, ritual, drama, or poetry).] Application: The decline of Hell The Christian doctrine of eternal torment would seem to contradict the preceding predictions. It fact, however, both the doctrine and the social conditions that sustain it are well within the theorys scope. The puzzle, of course, is why anyone would buy into a church or church doctrine that threatens members with the prospect of, to put it mildly, very large negative rewards. (Note that the issue here is not the presence of what I have termed a reward gradient, which offers more to devoted members than casual members or nonmembers. The issue concerns the horrifyingly large, absolute losses associated with Hell.) The solution probably lies within the realm of group dynamics, wherein people choose not only for themselves but for others as well. Although no religious adherent wishes to buy Hell for himself, he may well appreciate its use as a threat against others. Above and beyond Hells

17

value in establishing a reward gradient as described above, the threat of afterlife punishment may benefit:
n

Parents trying to raise loyal children, Governing elites trying to maintain a loyal citizenry, Economic elites trying to reduce the threat of theft and shirking among their worker and tenants, and Religious elites trying to reduce defection from their religion.

It is not hard to see why any or all of the former groups might be willing to pay to put the fear of Hell into members of the latter groups (see Becker 1996:227 for a similar argument). They might even be willing to live with increased fear among themselves as long as they thereby raise the level belief and fear in others. This last point is important, for it helps us understand how self-interested people might genuinely accept a doctrine that, viewed merely at the individual level, reduces their anticipated utility. Where Hell no longer serves these functions or is no longer needed to serve them, then its popularity should decline (since it continues to threaten the buyers with the disturbing prospect of personal damnation). For example, as greater personal wealth and the welfare state leave parents less desperate to constrain their childrens behavior and ensure their long-term loyalty, parents will become less anxious to expend resources teaching their children about Hell.22 Likewise, certain political changes such as increased democracy, religious freedom, and the separation of church and state tend undermine the governing elites capacity and desire to indoctrinate, regulate, and legitimate through this or any doctrine. These arguments about Hell embrace a bit of the old Marxist religion-as-opiate thesis, but also much more. New conditions may also alter Hells intrinsic plausibility. For example, people may be less inclined to believe in damnation (or more precisely, in a God who would punish people with such torments) as conditions in this life become less hard and, especially, as parents, employers, and rulers treat their subordinates less harshly. Finally, Hell becomes less attractive as societies become more religiously diverse. The notion that torment awaits all non-believers yields less utility when such people include friends, coworkers, and family members (and it takes a truly terribly marriage to make a husband relish the prospect of his mother-in-law burning forever in Hell). We might even say that the road from Hell is paved with good intentions good intentions, that is, toward those who might otherwise suffer its pains.23 24 Evidence: Historical data tend to support the predictions above. For example, content analyses of American sermons track a dramatic decrease in reference to Hell and damnation over the past two centuries [cite]. (Ekelund, Hbert et al. 1996) argue that the medieval Catholic church manipulated the doctrines of hell and purgatory to enhance their power and wealth. [Conjecture: The Christian doctrines of eternal torment became more elaborate and more widely accepted as the mobility and pluralism of the Greco-Roman world gave way to the medieval European world of isolated small, static, and religiously-homogenous towns.] Finally, there exist relevant survey 18

data, including GSS or Gallup, that indicate how attitudes toward Hell vary with the respondents socio-economic status. At the level of organizations rather than individuals, it appears that the afterlife teachings and doctrinal emphases of different denominations do correlate, in the anticipated directions, with the average socio-economic status of their members (after controlling for purely historical doctrinal effects). Thus, denominations with richer and more educated members abandoned the emphasis on hell sooner and more thoroughly than those whose members tend to poorer and less educated.

Recap: Section V
Although the theory of religious teachings sketched out above is far from complete, it does provide a new and useful way to define, analyze, and compare different doctrines. Indeed, the framework of this paper constitutes the first serious attempt to develop social-scientific theory of religion and religious doctrine amenable to formal modeling and mathematical analysis. As a practical matter, the theory helps us recognize, several things. First, most religious teachings have a common (action/result) underlying structure. Second, peoples religious behavior seems broadly consistent with the notion that they evaluate religious teachings in terms of their anticipated costs and benefits and, where possible, choose to follow those teachings that offer higher net returns. Third, religious doctrines will not flourish unless they some manage to generate net benefits for both sides of the market the buyer-adherents and seller-clergy. Fourth, competition and the untestable character of afterlife claims leads religious suppliers to offer their members huge afterlife payoffs. Fifth, in the market for members, high these high payoffs to members must be coupled with much lower payoffs to non-members, or else potential members will have little or no strategic incentive to join the religion. Sixth, and last, members will tend to embrace the threat of religious penalties, including large afterlife penalties, if the threat functions to constrain the behavior of people with whom they interact. To put it more bluntly, people may buy into the doctrine of Hell if they can thereby scare the Hell out of others. Hells decline in the face of increased material comfort and social diversity is a striking implication of the theory, and an interesting case of pluralism undermining one traditional aspect of (Christian) religion without necessarily reducing religiosity per se.

Conclusions
For more than two centuries, academics confidently predicted (and often enthusiastically proclaimed) the immanent demise of religion the victim progressive and irreversible secularization whereby all forms of supernatualism withered before the advance of science and rationality. With the benefit of hindsight, not to mention mountains of data, it now appears that secularization theory never had much evidence, let alone any real theory, to justify its broad claims. Religious beliefs and activities remains popular with large portions of the population in most parts of the world, and religious institutions remain influential for good or ill nearly everywhere. And if Hayek is right about the historic functions of religion, this outcome may not be such a bad thing. Humans may still require religions to sustain their collective faith in the moral customs and traditions that provide the foundation for a market-based civilization.

19

In any case, an economic perspective sheds new light on the underlying demand for religion and the diverse forms of supply that arise to satisfy it. Economic concepts and categories extend quite naturally to the realm of religion in part because economic concepts and categories (such as cost and benefit or production and trade) are fundamental to all human activities. But the extension is also facilitated by the fact that market institutions and activities are not as radically unlike religious institutions and activities as people normally assume. (For example, monetary transactions are more common in the realm of religion and magic than most contemporary Americans realize. Exchange is fundamental to the functioning of both realms, as is trust in ones exchange partners and the mediums of exchange [forgive the pun]. In both realms, competition, imitation, innovation, and entrepreneurship are pervasive unless forcefully suppressed. And both religions and markets are emergent institutions, developing spontaneously and unpredictably over long periods of time, and arising out of innumerable exchanges and activities calculated to promote the interest of individuals, but ultimately affecting much larger groups.) Ultimately, of course, the economic approach to religion must be judged by capacity to advance our understanding of beliefs, behavior, and institutions. This paper is designed to illustrate that it does just that.

20

Endnotes:
1

For a discussion of group selection, see Field (Field 2002) who notes that only three other economists Hirschleifer, Samuelson, and Becker acknowledged the importance of group selection prior to the 1990s.
2

Many elements of the theory are present or presaged in two seminal works by Rodney Stark and William Sims Bainbridge (1980; 1987). See also One True God, Starks (2001) recently published study of the historical consequences of monotheism.
3

My argument for the universality of supernaturalism is fundamentally different from, though not entirely inconsistent with, the evolutionary account put forth by Pascal Boyer (2001).
4

The day after writing this sentence, I happened to visit a Christian bookstore where I encountered shelf upon shelf of Christian self-help literature, including hundreds of books on getting out of debt, disciplining children, coping with death, healing, mother-daughter relationships, addiction, retirement, cooking, exercise, home-schooling, adoption, dating, divorce, marriage, and, yes, sex. One particularly striking volume, A Celebration of Sex (xxx 199x), carried the subtitle A guide to enjoying Gods gift of married sexual pleasure.
5

This conclusion contrasts with that of most early-to-mid twentieth-century sociologists and anthropologists, who sought to define, explain, and analyze religion in terms of its supposed functions, most notably social integration (e.g., Durkheim, Radcliffe-Brown, Malinowski, Parsons, and Yinger).
6

Magical elements are, to varying degrees, present in all religious traditions. Orthodox Catholicism, with its emphasis on miracles, saints, and relics, is considerably more magical than traditional Protestantism, which emphasized faith and Biblical teaching above all else. Reform Judaism is considerably less magical than Hassidic Judaism.
7

My analysis of magic has benefited from numerous studies of magic by cultural anthropologists (such as ). It has also benefited from McCloskeys (1990) perceptive and entertaining discussion of the poetics and economics of magic. McCloskey does not, however, develop what one might call a theory of magic, much less a formal model, since he is writing to demystif[y] economic theory and rescue its wisdom from the pretensions of experts by stressing that models are metaphors, theories are stories, and [s]cience is rhetoric, all the way down [McCloskey 1990: dust jacket].
8

Except where otherwise noted, all biblical quotations come from the New International Version (International Bible Society 1988).
9

This conclusion recapitulates Durkheims (Durkheim 1965: 59, 44-45) famous observations contrasting religion and magic, which went so far as to define religion in terms of its collective dimension. According to Durkheim, In all history we do not find a single religion without a Church It is quite another matter with magic. The magician has a clientele and not a Church, and it is very possible that his clients have no other relations between each other, or even do not know each other A Church is a moral community . But Magic lacks any such community.
10

[Review and cite Voltaire, Spencer, Marx, Freud, Huxley, Wallace, Dobbelaere, etc.]

21

11

My model of assertions includes Stark and Bainbridges (1980; 1985: 6; 1987: 36) compensators as a special class of [A,R] assertions in which the results are highly general and/or future-oriented.
12

An aside on evaluation costs: To formalize this notion, we imagine a cost function, C(.), defined over the set of all possible assertions {[A,R]}. By definition, C([A,R]) denotes the cost of evaluating the truth or falsity of assertion [A,R]. Religious assertions tend to have very high evaluation costs. The same, however, is true of many secular goods, such as the forms of medical services, psychological counseling, and investment advice that economists call credence goods (because their value can scarcely be determined even after their consumption). For those concerned that no assertion can ever really be proved, we might take a more Bayesian approach and let C([A,R],p) denote the cost of arriving at an evaluation that one views as correct with probability p. Thus, C([A,R],.95) denotes the cost of becoming 95% sure that [A,R] is or is not true. [Include more here on the difference between search, experience, and credence goods?] 13 Biblical examples include: Honor your father and your mother, so that you may live long in the land the Lord your God is giving you (Exodus 20:12), Ask and it will be given to you; seek and you will find; knock and the door will be opened to you (Matthew 7:7), and Repent and be baptized and you will receive the gift of the holy spirit (Acts 2:38). 14 Consider, for example, Gods famous promise to Abraham, reiterated in numerous texts: I will make you into a great nation and I will bless you; I will make your name great (Genesis 12: 2). This apparently unconditional blessing is, in fact, linked to Abrahams faith in God, as is each subsequent repetition or elaboration of the promise (as in Genesis 15:1-6 and 22:15-18). More importantly, Jewish and Christian interpreters of these texts have for thousands of years viewed them as proof that God rewards faith and keeps His promises (e.g., Romans 4:16-25, Hebrews 11:8-12, 17-19).
15

Most of the Bibles ten commandments illustrate this form of statement: You shall have no other gods before me. Remember the Sabbath day by keeping it holy. You shall not murder. You shall not commit adultery. You shall not steal. You shall not give false testimony You shall no covet (Exodus 20:3, 8, 13-17). The relevant promise is nearby, in Exodus 19:5: Now if you obey me fully and keep my covenant, then out of all nations you will be my treasured possession. Note also the no-so-implicit threat in Exodus 15:26: If you listen carefully to the voice of the Lord and keep all his decrees, I will not bring on you any of the diseases I brought on the Egyptians, for I am the Lord, who heals you.
16

For example: In the beginning, God created the heavens and the earth. (Genesis 1:1).

In the beginning was the Word. Through him all things were made; (John 1:1-3) The message of these texts is no less clear, though certainly less crude, than that of the evangelical bumper sticker they inspired: God and Son, Inc. doing business with folks like you for thousands of years. Indeed, religious narrative almost always exists to promote faith in the religion and its assertions. The previously-cited book of John is especially up-front about this fact: Jesus did many other miraculous signs in the presence of his disciples, which are not recorded in this book. But these are written that you may believe that Jesus is the Christ, the Son of God, and that by believing you may have life in his name. (John 20:30-31).

22

17

The consumers goal then becomes the maximization of a risk-averse expected utility function EU, evaluated over the costs and benefits associated with all his activities. Letting I denoted the magnitude of value associated with all activities except [A,R], his objective function may be written: EU[ V(R)-C(A)+I].
18

In terms of the framework described above, the church teaches two complementary assertions: [A1 ,R1 ] and [not-A1 ,P1 ], with C1 = C(A1 ) and P1 < R1 .
19

For the sake of generality, I have designed the printed matrix to accommodate any set of churches, numbered 1 through n, and to track two other factors the observable, here-and-now benefits, B, associated with membership in each church and the subjective probability, , that the person assigns to the truth of each church.
20

It is not difficult to extend the formal model of religious choice so as to also accommodate choices concerning how much time and money to devote to religion, and which takes account of the interplay between an individuals assessment of costs, benefits, and probability and the activities of his or her fellow members. This general model plays a major role in my book, The Economics of Religion.
21

The son of the famous atheist-activist, Madeline Murray OHare, is merely the best-known recent example. This one-time child plaintiff in the court case leading to abolition of school prayer in 196x now writes books and routinely appears on TV, radio, and other public forums to renounce his atheistic upbringing and promote Christianity.
22

One must consider both the income and substitution effects at work here, but the basic intuition seems sound.
23

This argument also applies to 19th-century disputes over the doctrine of Predestination, which seemed much more fair in tight, religiously homogeneous societies where most everyone was supposedly among the elect (Finke 19xx). Likewise, as knowledge of and sympathy for people outside the West has grew, Christians became increasingly concerned about the damnation of unbaptized children and ignorant heathens.
24

Should we view this as an example of secularization la Peter Berger (1967)? Only if we can show that this doctrinal change amounts to reduced religiosity.

23

APPENDIX Prices, pleasures, promises, and plausibility a formal model of church choice and religious involvement
Let i = actors choice(s) of church j = the true church/true state of the world and let X = the actors church-related activities S = the actors stock of religious preferences/experiences Schu = the churchs (median or mean) stock of religious preferences/experience Sref = the religious preferences/experiences of the actors reference group C = opportunity cost of the actors church-related activities R = the expected mystical (future + current but unknowable) rewards/punishments B = here-and-now, observable benefits from church participation = subjective probability that a given churchs R-claims are true. then EU = utility from nonreligious activities + j{ j(Si,Sref,Schu) Rj(Xi,Si) } + Bi(Xi,Si,Sref,Schu) (credence benefits) (observable benefits) C(Xi) (opportunity costs)

By assumption, Si is a function of the Xi and Si from previous periods and perhaps also the Sref and Schu from previous periods. In this model, subjective probabilities, expected rewards, observable benefits, and opportunity costs all vary continuously as a function of the actors religious activities and preferences and/or experiences and the preferences and/or experiences of his reference group(s) and his church. The game-theoretic payoff matrix in figure 4.1 is a simplified variant of this model in which all the payoffs and probabilities are fixed.

24

Figure A.1 Choosing your Church: Game-Theoretic Matrix of Expected Payoffs

State of the world 1 is true k is true j is true n is true none is true n 1i Pn+B1-C1 B1-C1 Pn+Bk-Ck Bk-Ck Pn+Bj -Cj Bj -Cj Rn+Bn-Cn Bn-Cn Pn 0 j Pj +B1-C1 Pj +Bk-Ck Rj +Bj -Cj Pj +Bn-Cn Pj k Pi+B1-C1 Rk+Bk-Ck Pi+Bj -Cj Pi+Bn-Cn Pi Subjective probability 1 Join church 1 R1+B1-C1 Join church k P1+Bk-Ck Join church j P1+Bj -Cj Join church n P1+Bn-Cn Join none P1

Deriving the payoff formulae: Let k denote any church and let j denote the particular church to which a person adheres (j,k=1,2, n or none). Let Rk denote the unverifiable (e.g., afterlife) reward associated with adhering to church k if it is true, and let Pk denote the unverifiable penalty associated with not adhering to church k if it is true. (Pk need not be negative, and is more properly termed an alternative payoff.) Let Bk and Ck denote the observable, here-and-now benefits and costs associated with adhering to church k. Finally, for each church k, let k denote the subjective probability that the person assigns to church k being true. It follows that the persons total payoff will be Rk +Bk -Ck if j=k, Pk +Bj-Cj if jk, Pk if he adheres to no church (j=none), and Bj-Cj if no church is true (k=none). The matrix above indicates the full set of such payoffs, with rows corresponding to possible true churches and columns corresponding to the persons possible choice of churches. To obtain the expected payoff associated with any particular church choice, j, the individual works down the corresponding join church j column, multiplying each cells payoff by its associated probability and then summing over all possiblities (k = 1,2, ,n, none). For jnone, the resulting expected payoff is Bj-Cj + k j( k Pk ) + jRj. For j=none, it becomes k j( k Pk ). And for k=none, but jnone, it becomes Bj-Cj.

25

Bibliography: Becker, G. S. (1976). The Economic Approach to Human Behavior. Chicago, University of Chicago Press. Becker, G. S. (1996). Accounting for Tastes. Cambridge, Harvard Unversity Press. Berger, P. L. (1967). The Sacred Canopy. Garden City, NY, Doubleday. Boyer, P. (2001). Religion Explained: The Evolutionary Origins of Religious Thought. New York, Basic Books. Chiswick, C. U. and E. L. Lehrer (1991). "Religious Intermarriage: an Economic Perspective." Contemporary Jewry 12: 21-34. Darby, M. R. and E. Karni (1973). "Free Competition and the Optimal Amount of Fraud." Journal of Law and Economics 16(1): 67-88. Durkheim, E. (1965). The Elementary Forms of the Religious Life. New York, NY, Free Press. Ekelund, R. B., R. F. Hbert, et al. (1996). Sacred Trust: The Medieval Church as an Economic Firm. New York, NY, Oxford University Press. Field, A. J. (2002). Altruistically Inclined?: The Behavioral Sciences, Evolutionary Theory, and the Origins of Reciprocity, University of Michigan Press. Filatov, S. (1993). On Paradoxes of the Post-Communist Russian Orthodox Church. Religions San Frontiers? Present and Future Trends of Migration, Culture, and Communication. R. Cipriani. Rome, Rome University: 117-125. Greeley, A. M. (1989). Religious Change in America. Cambridge, MA, Harvard University Press. Hadden, J. K. (1987). "Toward Desacralizing Secularization Theory." Social Forces 65(3): 587-611. Hayek, F. A. (1988). The Fatal Conceit: The Errors of Socialism. Chicago, University of Chicago Press. Iannaccone, L. R. (1992). "Sacrifice and Stigma: Reducing Free-Riding in Cults, Communes, and Other Collectives." Journal of Political Economy 100(2): 271292. Iannaccone, L. R. (1994). "Why Strict Churches are Strong." American Journal of Sociology 99(5): 1180-1211. Iannaccone, L. R., R. Finke, et al. (1997). "Deregulating Religion: The Economics of Church and State." Economic Inquiry 35(2): 350-364. Iannaccone, L. R. and R. Stark (1998). "Rationality and the 'Religious Mind'." Economic Inquiry 36(3): 373-389. Jagodzinski, W. and A. M. Greeley (1997). The Demand for Religion: Hard Core Atheism and 'Supply Side' Theory. McCloskey, D. N. (1990). If You're So Smart: The Narrative of Economic Expertise. Chicago, University of Chicago Press. Neuman, S. (1986). "Religious Observance within a Human Capital Framework: Theory and Application." Applied Economics 18(11): 1193-1202. Pascal, B. (1910). Pascal's Pensees. Stark, R. (2001). One True God: Historical Consequences of Monotheism. Princeton, Princeton University Press.

26

Stark, R. and W. S. Bainbridge (1980). "Towards a Theory of Religion: Religious Commitment." Journal for the Scientific Study of Religion 19(2 June): 114-128. Stark, R. and W. S. Bainbridge (1985). The Future of Religion. Berkeley, University of California Press. Stark, R. and W. S. Bainbridge (1987). A Theory of Religion. Bern, Peter Lang Publishing. Thomas, K. (1973). Religion and the Decline of Magic. Harmondsworth, Middlesex, Penguin. Troyanovsky, I., Ed. (1991). Religion in the Soviet Republics. San Francisco, HarperCollins.

27