From Jesus to Christianity: The economics of sacrifice

Abstract

This article models the birth of a new religion from the ashes of apocalyptic prophecy. Christianity started around the imminent expectation of God’s Kingdom. Followers forsook worldly opportunities to prepare for the event. As the Kingdom’s arrival tarried, they found themselves “trapped” because those sacrifices—like transaction-specific investments—were wasted if they dropped out. This provided incentives to stay and transform the faith. Such effort, enhanced by reaction to the cognitive dissonance caused by prophecy failure, turned an apocalyptic movement into an established church. A survey of other apocalyptic groups confirms that dropout costs are critical to explaining outcomes.

Keywords

Apocalyptic movements birth of religion Christianity cognitive dissonance sacrifice

Introduction

One way of looking at the rise of Christianity is to describe it as a process by which a colossal failure was turned into a world-shattering success. The inspirational leader and founder of the faith, Jesus of Nazareth, was met with a mixed reception in his mission to Israel, was rejected by the Jewish authorities of Roman Palestine, and unexpectedly ended his life on a Roman cross, following which his disciples scattered in fear. Yet, today the religion that bears his name is a major world religion. While historians have devoted considerable effort to finding out who joined the Jesus movement and why, to an economist dedicated to a rational choice approach, the puzzling question is not so much why some people joined—there will always be somebody joining any movement for whatever reason—but why and how they endured failure and stayed on. In a nutshell, the answer of this article will be that the nature of the initial failure holds the key to the subsequent endurance and eventual success. Because the Jesus movement was an apocalyptic faith, centered around the imminent expectation of the end of time and the coming of the Kingdom of God on earth, the people who joined sacrificed a variety of opportunities in life to fully prepare for the impending final consummation—much like posting a bond whose value can be recouped only if and when the expectation comes true. As the Kingdom’s arrival tarried, and as dropping out and forsaking the bond were costly, the members had an incentive to stay on and devote their efforts to a gradual transformation of the faith so that it became less and less reliant on apocalyptic expectations and more and more adjusted to an extended sojourn in the world as it is. This protracted process turned an apocalyptic or millenarian movement into an established church.

Even readers who have been brought up in the Christian tradition will likely feel uncomfortable with this approach: everybody knows that there was such an inconveniently odd feature in the early faith—the Book of Revelation easily comes to mind here—but this was thankfully laid to rest on a high shelf of the religion’s library early on. And, those readers will think, this is as it should be: as historians have shown and sociologists and economists working in the rational choice tradition such as Stark (1996) and Ekelund and Tollison (2011) have elaborated, there were many good reasons why Christianity grew and finally triumphed in the Roman Empire and beyond without any recourse to the apocalyptic belief. However, those reasons for attraction and growth apply to a church that was already a going concern, while the focus of this article is on just how that church came into being in the first place. That is why we will have to retrieve that apocalyptic feature from its remote shelf and give it pride of place.

Unearthing the apocalyptic foundations of Christianity will require an extended tour of the recent historical scholarship on Jesus and Christian beginnings. This is because the church worked hard to bury those foundations and break free of apocalyptic ideas, something called “de-eschatologizing” the faith—a process that was completed only with Augustine at the beginning of the 5th century ad. Even after that time, however, the millenarian belief was not dead but time and again has kept bursting back to life from within Christianity to this day. This is a fortunate circumstance as it will allow us to test our theory against the evidence of a number of apocalyptic movements that are relatively recent and therefore better known. As we will see, some of those movements managed to survive the failure of prophecy and become established, while others fizzled out. In each case, the initial sacrifice and the way in which it was put to use were critical in explaining outcomes.

While, as mentioned, most historians have skirted the question of why the early Christians did not drop out when faced with the failure of apocalyptic prophecy, a few have addressed it. The latter group typically relies on the theory of cognitive dissonance, developed by psychologists with a view to, among other things, explaining the survival of apocalyptic groups. The idea is that the failure of prophecy produces a disconfirmation of belief, which is unsettling; in reaction to this distress, the individual re-interprets the belief so as to find that the prophecy had not really failed after all and consequently takes action to confirm the new interpretation—typically, missionary action to bring new converts to the group. Starting with Akerlof and Dickens (1982), cognitive dissonance theory has a place in economics and can be pressed into service here. However, placing the whole burden of explanation on cognitive dissonance seems to be a stretch: the sheer cost of exit from the group suggests itself as a sufficient reason to stay, following which cognitive dissonance, and the reaction thereto, can help keeping up the level of effort. In this article, then, we will model the exit cost as the key factor and the working of cognitive dissonance as conditional upon it.

Our exercise in applied economic theory requires a fresh start because the existing literature on the economics of religion has never addressed the problem of the birth of a new religion. Most of that literature, starting from Iannaccone’s (1992) seminal model of a sect, builds on the idea of sacrifice and stigma as devices that counter free riding and promote commitment and participation in a religious group. We will borrow the idea of sacrifice as a feature that marks out members from the general society, but the similarity ends there: in our framework, sacrifice works to prevent defection from the group, not free riding within the group. Free riding, or shirking, is typically not an issue in an apocalyptic group because the individual expects to be face to face with God in the imminent reckoning and one cannot cheat the monotheistic God, who is believed to be all-knowing—a perfect-information reward and punishment system. On the other hand, as mentioned above, in a different, interesting line of research not so beholden to the club model, Ekelund and Tollison (2011: ch. 4) briefly discuss the early days of Christianity and note the change in the apostles’ time discount rate, but they never ask the critical questions of why the group held together as the Lord’s return did not materialize and how the transition to the post-apocalyptic phase could be accomplished, while such questions are the novel focus of this article.¹

The remainder of the article is organized as follows. The next section takes the reader to a tour of the history of the early Jesus movement, from the time of Jesus to the end of the New Testament period. In the following two sections, we set forth the model and then go back to the beginning of Christianity with the benefit of the model’s insights. This is followed by a section that reviews a variety of apocalyptic movements, both inside and outside Christianity, looking for support for the model’s predictions. A final section concludes.

From Jesus to Paul to the Christian church: Some history

Centuries of theology-laden interpretation of the Christian message have downplayed the obvious apocalyptic content of the New Testament texts or explained it away as symbolic or allegorical discourse. Today, however, the majority—though by no means the totality—of Jesus and New Testament historians seem to agree that Jesus’ message and mission were thoroughly apocalyptic, that the earliest church documents we possess—Paul’s “authentic” letters²—are also thoroughly apocalyptic, and that the Gospels and other New Testament documents³ can be read as evidence of the protracted struggle waged by different Christian communities to extricate themselves from the grips of the apocalyptic mindset, at the same time as breaking away from the original Jewish fold.⁴

Recovering what can reasonably be taken to be authentic words and deeds of Jesus from the much later materials that comprise the New Testament is a fine act of investigative scholarship, in which all results must be qualified by probabilities and almost never by certainty. Besides relying on textual and linguistic analysis, on multiple attestations as reinforcing evidence, and on historical context known from independent sources, this detective work relies on a principle known as the criterion of dissimilarity (Fredriksen, 1999: 184–185): if a saying by, or a story about, Jesus is dissonant from the teachings of the early church, then it is likely to be authentic because later redactors would have had no interest in fabricating it or twisting the original record in that direction. This criterion may be the single best reason to accept that both Jesus and his first followers were indeed apocalyptic: since the church was soon at pains to overcome the apocalyptic tradition, if the Gospel writers included all such material it must have been because it bulked so large in the received tradition about Jesus that it could not be suppressed (Grant, 1977: 18–19)—only edited and qualified whenever possible, as we will see. Hence, it is likely to be authentic, that is, dating back to the real ministry of Jesus.

Another principle for doing good history is that for Jesus’ message to be successful—and we know it was successful, at least with some people who followed him—it must have been able to speak to the beliefs and concerns of his audience in early 1st-century Galilee and Jerusalem. Thinking otherwise—that it should make sense to us, rather than to them—is a form of anachronism. So, if an apocalyptic mission found receptive ears among Jesus’ contemporaries, it is because such a message of redemption was not foreign to late Second Temple Jews. Since the works of the biblical prophets, especially Isaiah and Ezekiel, there was an expectation that God would one day take direct control of world events, put an end to evil, pass a final judgment on both the dead and the living, and restore Israel. This expectation only grew stronger through postexilic Jewish writings, such as the book of Daniel, and the actions and writings of radical, separatist groups such as the Essenes and the sectarians at Qumran who produced the Dead Sea Scrolls. In the period leading up to the outbreak of the Jewish revolt against Rome in ad 66, a number of “prophets” gathered crowds of tens of thousands and led them out into the Judean desert or promised them the parting of the Jordan river (like that of the Red Sea of old) or the collapse of Jerusalem’s walls (like that of Jericho’s of old), each of which actions was supposed to usher in God’s deliverance. In each case, the Roman army put them down swiftly and thoroughly. Later on, as we will see in a section to follow, many Jews joined the revolt and especially the battle at Jerusalem because they believed that God was ready to deliver them, as he had at the time of the Egyptian slavery. And closest to the time of Jesus, John the Baptist preached repentance and cleansed people by immersion in the Jordan river in view of the impending arrival of God’s Kingdom. Many of these teachings and movements focused on the Messiah, supposed to be a Jewish leader of Davidic descent who would be instrumental in bringing about Israel’s redemption. Thus, in a variety of ways, the Palestinian setting of the day was ripe for Jesus’ ministry to have an impact.

Jesus picked up John’s message and carried it around in an itinerant mission through the Galilee. His very first words, as recorded in the earliest of the Gospels, say it all:⁵ “The time is fulfilled, and the kingdom of God has come near; repent, and believe in the good news” (Mark 1: 15). He taught that this Kingdom of God was to be a cataclysmic event on earth, not in heaven, that would change the world and bring history and time to an end; that there a final judgment would be meted out to all the living and the dead—the latter resurrected—with regard to both justice and mercy; that this would be the work of God alone, while all that men were to do was to repent their sins and stand ready to enter the Kingdom; and that he himself had a very special relationship with God and was to play a key role in ushering in the Kingdom—although it is not clear that he actually claimed either the title of Messiah or Son of God for himself. The timing is critical: this event was to occur soon, very soon, possibly within Jesus’ own lifetime or at the end of his career, and in any case within the lifetime of the generation of his audience.

Given this timeline and Jesus’ own role in the event, his crucifixion must have come to his followers as a sudden, catastrophic disconfirmation of the prophecy. The notion of a crucified Messiah was unheard of in the Jewish apocalyptic of the time. But then the disciples became convinced that Jesus had risen from the dead and appeared to them many times, commissioning them to spread the gospel and faithfully await his return, after which he ascended to heaven. So, the prophecy was now revived in a new form, also a novelty in the context of contemporary Jewish thinking: the Kingdom would now be established by the Second Coming (Parousia) of Jesus, whose death and resurrection signaled that the transformation and redemption of the world had indeed begun. Thus, the belief in the resurrection turned disconfirmation into re-confirmation of the prophecy, propelling the movement forward. As the book of Acts testifies, mission began in earnest.

This is where Paul comes in. He was not part of the original Jesus disciples but embraced the faith, as he himself tells us, upon his encounter with the risen Christ on the road to Damascus. To an educated, committed Pharisee like Paul, this fits perfectly in the apocalyptic framework: in contemporary Jewish thinking, the resurrection of the dead was part and parcel of the end of times, so the resurrection of Jesus was an unambiguous sign that the end had begun. His own assignment for the end times was—again in a perfectly Jewish fashion—to urge the Gentiles to believe the gospel and join the people of Israel in the Kingdom, but as Gentiles not as Jews (letter to the Galatians)—the “ingathering of the nations” foretold by Isaiah (Eisenbaum, 2009; Fredriksen, 1991b). It was an overarching task that required full commitment, and Paul devoted himself to it from the beginning to the end of his career as apostle to the Gentiles. And he was apocalyptic to the very end. In his earliest extant letter to the Thessalonians (ca. ad 50), he already struggles with doubts raised by some members of the congregation, who had seen with dismay some of their brethren die before the Lord’s return: “For since we believe that Jesus died and rose again, even so, through Jesus, God will bring with him those who have died” (1 Thessalonians 4:14). A few years later, in Paul’s most detailed discussion of life under the shadow of the apocalypse, he proclaims that “the appointed time has grown short … The present form of this world is passing away” (1 Corinthians 7: 29, 31). And in his last letter to the Romans (ca. ad 60), some 25–30 years after the crucifixion, Paul could still adamantly proclaim that “salvation is nearer to us now than when we became believers; the night is far gone, the day is near” (Romans 13: 11–12). How so? Because in his mercy God had temporarily hardened Israel’s heart to the gospel message so that more time would be available for Paul and others like him to reach out to the Gentiles and bring them to the faith—“until the full number of the Gentiles has come in” (Romans 11: 25), after which all, Jews and Gentiles, will be saved. So, in Paul’s view, Israel’s rejection of the gospel and the delay of the Kingdom are both opportunity and motive for enhanced missionary effort among the Gentiles.

About a decade after Romans, the fall of Jerusalem and the destruction of the Temple in ad 70 provided a shocking test of apocalyptic beliefs: if apocalypse meant anything, surely such an epoch-making event as the demise of the Jerusalem Temple must be a sign of the end. Mark’s Gospel, written immediately after the Destruction, is the most thoroughly apocalyptic of the canonical Gospels and includes the Destruction as a “prophecy after the fact” put in Jesus’ mouth within an extended description of the apocalypse (Mark 13). Thereafter, with the passing of time and the non-event of the Kingdom, we see the later New Testament texts gradually distancing themselves from the apocalyptic framework. For all his staunch belief in the end, Paul had already warned against tying one’s expectations to a precise timetable, thus buying time as well as giving an incentive to be alert and ready at all times: “the day of the Lord will come like a thief in the night … But you, beloved, are not in darkness, for that day to surprise you like a thief” (1 Thessalonians 5: 2, 4). (Some 20 years after this letter, Mark 13: 32–37 puts the same warning in Jesus’ mouth.) The Gospels of Matthew and Luke, composed in the 80s, still preserve the apocalyptic sayings of Jesus but surround them with amendments, qualifications, and dilutions. The book of Acts basically celebrates the church’s mission to the whole world—suggesting persistence in time, not life at the end of the world. The Gospel of John, probably written in the 90s, describes a stranger from heaven who engages in lengthy theological dissertations and barely mentions the Kingdom of God as a metaphysical reality that is already present in the believer’s heart, not an upcoming event in the material world. The latest New Testament texts are further removed from the original mindset. The letter to the Hebrews (written near the end of the 1st century) presents Jesus as the new High Priest and Christianity as the successor religion to Judaism; it speaks about holding steadfastly on the expectation of salvation, but daily life in the church is the main concern while the final judgment recedes into the indefinite future. The second letter of Peter (written probably in the early 2nd century) explains that the Lord is not slow in delivering on his promise for the last days but counts the time differently from us—one day for him is a thousand years for us. Finally, in the 2nd century, intra-Christian sectarian disputes take center stage, suggesting an organization that is groping for identity and stability.⁶ The apocalyptic belief was a diehard, recurring problem—the Book of Revelation was written in the 90s—but was gradually driven to the fringes and eventually outside of the mainline church, as exemplified by the Montanist movement in the latter half of the 2nd century.

Of particular concern to this article are the costs involved by membership in this apocalyptic movement. While centuries of Christian commentary have been busy de-eschatologizing the Christian message, they have salvaged the ethical instruction in the New Testament as a utopian blueprint for social reform and moral regeneration. At the endpoint of this process, a number of recent biographies of Jesus picture a non-apocalyptic hero who battles social inequality, hierarchy, patriarchy, nationalism, and militarism.⁷ In contrast, the majority of modern scholarship emphasizes that to make sense to their audience, those ethical prescriptions had to have a narrow target: instructions for life in the community of the last days, not for normal social life in the world as it is. Jesus’ instructions in the Sermon on the Mount (Matthew 5–7) are famous: love your enemy; turn the other cheek; the meek will inherit the earth; do not take oaths; do not worry about tomorrow but do as the birds of the air, which neither sow nor reap; do not judge, that you may not be judged; do not divorce. Elsewhere, Jesus’ followers are enjoined to give away all their possessions to the poor, for “it is easier for a camel to go through the eye of a needle than for someone who is rich to enter the kingdom of God” (Mark 10: 21, 25). And on the matter of sexual behavior, there are people “who have made themselves eunuchs for the sake of the kingdom of heaven. Let anyone accept this who can” (Matthew 19: 12).

Such “ethical perfectionism,” which involves an intensification and extension of the traditional Jewish prescriptions of the Ten Commandments (eschew not just murder but anger, not just adultery but lust in one’s heart, not just swearing falsely but swearing itself; love not only your neighbor but your enemy as well), makes sense only in the foreshortened time frame of apocalyptic expectation. For such exceptional times, traditional Jewish Law is not quite good enough: the impending redemption requires an exceptional ethics (Fredriksen, 1999: 98–110; Sanders, 1993: ch. 13). Clearly, no normal society could run for long on these principles. It is likely that these Gospel teachings go back to primitive Christian tradition, and thus to Jesus himself, because they are confirmed by the independent and earlier evidence of Paul. He too enjoins his Gentile communities to bless persecutors, never take revenge, patiently suffer injustice, love one another, not divorce, practice celibacy if one can, and put family second to preparations for the coming redemption. Interestingly, a Christian married to a pagan should not divorce so that certainly the children, and possibly the spouse, can be saved. A passage from the Corinthian correspondence is worth quoting in full:

I think that, in view of the impending crisis, it is well for you to remain as you are. Are you bound to a wife? Do not seek to be free. Are you free from a wife? Do not seek a wife. But if you marry, you do not sin, and if a virgin marries, she does not sin. Yet those who marry will experience distress in the flesh, and I would spare you that. I mean, brethren, the appointed time has grown short; from now on, let even those who have wives be as though they had none, and those who mourn as though they were not mourning, and those who rejoice as though they were not rejoicing, and those who buy as though they had no possessions, and those who deal with the world as though they had no dealings with it. For the present form of this world is passing away. (1 Corinthians 7: 26–31; author’s emphasis)

The first part of the quote basically says that anything goes “in view of the impending crisis.” Precisely in view of this, however, being unattached is best if one can manage. The second part says that one’s position in the world does not matter and one should just ignore it, “for the present form of this world is passing away.” The point to drive home here is that one should disregard and forsake everything that matters in life because the end is at hand, not for its own sake.

By contrast, later New Testament texts soften the ethics so that following in the path of Christ becomes a more manageable exercise (Rowland, 1985: 285–294). The wealthy man Zacchaeus satisfies Jesus by committing to give away half his wealth, not the whole of it, and more than compensating those whom he had harmed as a tax collector (Luke 19: 8–10). The prohibition against divorce holds “except on the ground of unchastity” (Matthew 5: 31–32). Furthermore, communities are instructed to uphold the existing social order in every way but for a touch of gentleness: traditional patterns of authority and deference should remain in place between husband and wife, parents and children, master and slave, but the stronger party should treat the weaker party nice (Ephesians 5: 21–6: 9; also Colossians 3: 18–4: 1). All worldly authorities and institutions are to be accepted and honored during this time of “exile,” as befits “a chosen race, a royal priesthood, a holy nation, God’s own people”—a re-enactment of the ancient Israelites’ plight (1 Peter 1: 17; 2: 9, 11–17). Those in the community who are sitting idle and not working—presumably in the expectation of the end—should earn their own living; if they don’t, they must be shunned, but as erring believers not enemies (2 Thessalonians 3: 6–14). The bottom line is that Christians should adapt as inconspicuously as possible to the surrounding society because the Parousia will be long in coming.

A model

Following Iannaccone (1992), we model religious activity as household production and introduce “sacrifice” as a one-time, upfront cost of joining the group (as opposed to ongoing restrictions of secular consumption, following the distinction in Berman, 2000). Unlike those articles, however, in the present model neither free riding nor screening of members’ types is an issue. Rather, here sacrifice generates an asset that enhances utility exclusively within the group and therefore plays a role similar to Montgomery’s (1996) denomination-specific, non-transferable religious capital that makes switching costly.

Consider a two-period model in which an individual may join a religious group in the first period and, if she does join, may either stay or drop out in the second period. The individual in each period is endowed with total time T to be used for work, religious activity, R, and sacrifice, K, to be discussed presently. Work at wage w allows her to buy consumption goods at price p. She derives utility from secular⁸ consumption, S, and, if a member of the religious group, from religious consumption, R, as well. For simplicity and without loss of generality,⁹ preferences are quasi-linear

U = S + α v (R)

(1)

where α is a productivity parameter and v(0) = 0, v′(R) > 0, and v″(R) < 0 in the relevant range. We further assume that there is satiety for the value of religion per unit of time or, stated differently, that the average product of time devoted to religion, v(R)/R, reaches an interior maximum.

There is no probabilistic uncertainty, and therefore no expected utility, in this model: as befits an apocalyptic expectation, believers who join in the first period blindly trust God to fulfill his promise, that is, they assign a zero probability to failure—but then in the second period they find that the promise of bliss fails so that the value of religious consumption drops. This is captured by a fall in the parameter α, with α₁ > 0 and α₂ < α₁, while α₂ is allowed to be negative (in which case R becomes a “bad”). However, dropping out in the wake of disappointment in the second period is not costless. Upon joining, the individual incurs an upfront cost in the form of a one-time sacrifice of time K, whose value, (w/p)K, is subtracted from consumption in the secular world, added on to consumption within the group with fellow believers, and left behind or lost if the person drops out.¹⁰ Importantly, this is not a sunk cost because its value is lost only if the person drops out; it is rather a transaction-specific investment of time in the group made in Period 1. It “shifts” consumption from the secular world to the group and cannot be shifted back; hence, it can be enjoyed only within the group. This sacrifice can be thought of as giving up marriage, family, employment, educational, or social networking opportunities for the sake of joining the faith, a cost that will have permanent, negative consequences upon return to the secular world. For expositional convenience, we will discuss K as a parameter set by the group’s leaders, although it can be usefully thought of as a given feature of the particular situation that predicts survival of an apocalyptic group.

With these assumptions, in Period 1 the budget constraints for a member, M, and a non-member, N, are

S_{1}^{M} = \frac{w}{p} (T - R_{1} - K)

(2a)

S^{N} = \frac{w}{p} T

(2b)

and their one-period utilities are

U_{1}^{M} = S_{1}^{M} + \frac{w}{p} K + α_{1} v (R_{1})

(3a)

U^{N} = S^{N} = \frac{w}{p} T

(3b)

Comparison of equations (2a) and (3a) shows that the value (w/p)/K is not lost but transferred to the group; hence, its enjoyment is made conditional on membership.¹¹ One should resist any temptation to interpret equation (3b) as reservation utility and conclude that a person joins the group if $U_{1}^{M} \geq U^{N}$ . For a believer in the prophecy, the imminent realization of the Kingdom of God is all that matters, so her decision to join will be based on her utility over the whole horizon, not on first-period utility alone; and utility over the whole horizon is overwhelmingly driven by utility of the second period, which is presumably infinite. In other words, a believer in the Kingdom will join whatever the sacrifice required, the utilities from different types of consumption, and the time discount rate: the joining decision is not an economic problem.

Since we are not concerned with group size or characteristics of members, the group will exist if somebody believes the prophecy. If a person joins, her level of religious effort could similarly be thought of as exogenous, determined perhaps by a leader or a group norm. But since the first period is one of awaiting the event, when humans still live in this world, it can perhaps best be thought of as optimally determined, which will be useful for later comparison. If so, maximization of equation (3a) subject to equation (2a) yields the condition for a member’s optimal level of R in the first period, ${\bar{R}}_{1}$

α_{1} v^{'} (R_{1}) = \frac{w}{p}

(4)

In Period 2, the Kingdom has not happened, and the disappointed believers are now confronted with a true economic problem: whether to stay or leave the group. The budget constraints for a continuing member, M, and a dropout, D, are

S_{2}^{M} = \frac{w}{p} (T - R_{2})

(5a)

S^{D} = \frac{w}{p} (T - K)

(5b)

and their one-period utilities are

U_{2}^{M} = S_{2}^{M} + α_{2} v (R_{2})

(6a)

U^{D} = S^{D} = \frac{w}{p} (T - K)

(6b)

while for somebody who never joined budget constraint and utility are still given by equations (2b) and (3b). Comparison of equations (2b) and (5b) provides the key to the model: the sacrifice made for the group in Period 1 cannot be made good for purposes of secular consumption in Period 2. If a member drops out, she finds that the value of K sacrificed in Period 1 is permanently subtracted from her consumption in Period 2. In other words, the initial sacrifice has long-lasting consequences, which provides a link between decisions in the two periods.¹² The member stays rather than leave the group if $U_{2}^{M} \geq U^{D}$ , which using equations (6a) and (6b) yields the staying condition

α_{2} \frac{v (R_{2})}{R_{2}} \geq \frac{w}{p} (1 - \frac{K}{R_{2}})

(7)

If she stays, maximization of equation (6a) subject to equation (5a) yields the condition for an optimal level of R in the second period, ${\bar{R}}_{2}$

α_{2} v^{'} (R_{2}) = \frac{w}{p}

(8)

Note that since K contributes linearly to utility, it does not affect the optimal R in either period, but since it cannot be taken out of the group in Period 2, K does affect the staying condition (7).

Condition (7) shows that the effect of sacrifice in this setting is to retain members,¹³ even if at a reduced level of R: it may well be that in Period 2 a member would drop out in the absence of sacrifice because condition (7) might fail with K = 0. This would be the case if $α_{2} v (R_{2}) / R_{2} < (w / p)$ . In other words, K creates a trap.

To be precise, a trap will exist if the sacrifice carried over from the first period is no lower than K*, the minimum level of K that just prevents the member from dropping out while the member is exerting her optimal second-period effort. This is found by solving condition (7) with strict equality

K^{*} = {\bar{R}}_{2} - \frac{α_{2} v ({\bar{R}}_{2})}{w / p}

(9)

where ${\bar{R}}_{2}$ is set by equation (8). Here $\partial K^{*} / \partial α_{2} < 0$ and $\partial K^{*} / \partial (w / p) > 0$ : the more serious the drop in utility from religion in Period 2 (i.e. the lower the α₂), and the higher the person’s opportunity cost of time, the higher the minimum level of sacrifice that prevents exit.

Although at K ≥ K* the member does not drop out, she will reduce her religious consumption or effort: since α₂ < α₁, clearly from equations (4) and (8), ${\bar{R}}_{2} < {\bar{R}}_{1}$ . From equations (3a) and (6a), it also follows that the achieved one-period utility level is now lower than what it was in the first period. It is here that cognitive dissonance, and the reaction thereto, comes into play. In the spirit of Akerlof and Dickens (1982), the member will feel discomfort or psychological distress at the failure of prophecy in Period 2, and because she is stuck in the group due to the sacrifice, she will try to reduce such distress by telling herself that the failure is not so bad after all—even to the point of convincing herself that no failure has in fact occurred. As a consequence, the perception of a higher-than-real marginal value of religious effort will drive the member to intensify her effort—something that could not have happened had the person dropped out. Then, since exit is not an option and hence there is no opportunity cost to choosing the “nicest” beliefs, the discomfort of cognitive dissonance will be minimized, and indeed eliminated, by going all the way to fully denying that any failure has happened and believing that the marginal productivity of religious effort is as high as it was in the first period.

Formally, the person chooses ${\hat{α}}_{2}$ , a subjective assessment of the religious productivity parameter for the second period, with $α_{2} \leq {\hat{α}}_{2} \leq α_{1}$ . With this parameter, second-period utility becomes

{\hat{U}}_{2}^{M} = S_{2}^{M} + {\hat{α}}_{2} v (R_{2})

(10)

and the level of R that maximizes this, ${\hat{R}}_{2}$ , solves

{\hat{α}}_{2} v^{'} (R_{2}) = \frac{w}{p}

(11)

Since condition (7) holds so that exit is ruled out, there is no cost to choosing ${\hat{α}}_{2} = α_{1}$ , and this will be the optimal choice as it completely resolves the cognitive dissonance and kills the associated discomfort. At this perceived productivity parameter, equation (11) becomes identical with equation (4), and the optimal level of R is back to what it was in the first period, ${\hat{R}}_{2} = {\bar{R}}_{1}$ . Consequently, equation (10) becomes identical with equation (3a), and the maximized level of one-period utility is back to whatever it was in the first period, ${\hat{U}}_{2}^{M} = U_{1}^{M}$ . Even so, of course, the failure of prophecy will always be there: whatever the one-period utility of the first period, total utility over the time horizon was driven by the expectation of the Kingdom, and this has not happened. All the revision of beliefs does is to energize the continuing member and make her happy to go on working for a re-defined purpose.

Figure 1 shows the situation with the minimum K*, full denial of the failure and ${\hat{α}}_{2} = α_{1}$ . The set of curves to the right are the marginal and average values of R in the first period, with ${\bar{R}}_{1}$ set by equation (4). The set of curves to the left are the marginal and average values of R corresponding to the “real” second-period α₂ consequent on the shock; ${\bar{R}}_{2}$ solves equation (8), and, given this, K* solves condition (7) with strict equality. Thus, this level of sacrifice is just sufficient to keep the member in. Following the full resolution of cognitive dissonance, the perceived α₂, ${\hat{α}}_{2}$ , is back to α₁, and therefore, the effective R₂, ${\hat{R}}_{2}$ , is back to ${\bar{R}}_{1}$ . Therefore, the member in denial is again enjoying the one-period utility level she was enjoying in the first period, shown by the average religious utility curve on the right;¹⁴ this is what reaction to cognitive dissonance, at its best, can accomplish.

Figure 1.

The trap and the reaction to cognitive dissonance.

Summing up, the central result of the model is that the reaction to cognitive dissonance works only within the trap: the latter explains why the member does not drop out, while the minimization of failure through psychological self-deception explains why the member, once she has chosen to stay, feels “happy” to hold out and does not reduce her effort. In other words, the decision to stay or exit the group logically precedes the onset of the reaction to cognitive dissonance: the latter involves only those who are trapped in the group, while those who leave do experience disappointment and a waste of sacrifice, but not one strong enough to prevent their exit and trigger denial of the failure; for them, mistakes are mistakes, bygones are bygones, and life goes on. The utility loss from prophecy failure is therefore real enough for everyone concerned: the Kingdom has not come true; their response to the failure, however, is contingent on how strong the initial sacrifice was.¹⁵ Thus, the burden of explaining an apocalyptic group’s survival beyond such failure does not rest on revision of perceptions alone but on the very real cost of going back to normal life. Nesting cognitive dissonance within prior sacrifice and making its resolution conditional on the latter are this article’s little new twist to this literature.

Back to Christian origins

As mentioned in section “Introduction,” the few historians who have squarely addressed the question of why the early Christians did not lapse and drop out when the anticipated end of times did not come to pass have elaborated upon the theory of cognitive dissonance, developed by psychologists Festinger et al. (1956) from the study of a modern apocalyptic group. In its original form, the theory says that a religious group faced by unequivocal disconfirmation of a key belief—such as the end of the world—will not necessarily collapse and disband. Alternatively, it may react to the distress and doubt caused by disconfirmation by taking action to reduce or eliminate the dissonance. Such action is likely to include two elements: rationalization—namely, a revision of the original belief with a view to softening the disappointment—and proselytism, the latter motivated by the assumption that “if more and more people can be persuaded that the system of belief is correct, then clearly it must, after all, be correct” (Festinger et al., 1956: 28).

In the same vein, Gager (1975: ch. 2, 1988) and Goodman (1994: ch. 8) claim that this is exactly what happened with early Christianity. In their view, the energetic missionary drive attested to by Paul’s letters and the book of Acts, and retrospectively enshrined in Jesus’ “great commission” to the apostles after the resurrection (Matthew 28: 19–20), can be interpreted as the early Christians’ reaction to the cognitive dissonance induced by the two dramatic disconfirmations of the death of Jesus and the delay of the Kingdom. Thus, “the church initially carried on its mission in an effort to maintain its eschatology” (Gager, 1975: 46; italics in the original). While this was happening, the church gradually revised its belief system so as to place it beyond disconfirmation by observable evidence, as we have seen. In support of this view of Christian evolution, we may add that, as the 1st century drew to a close and the transition to the post-apocalyptic phase got under way, open mission abated: we have no evidence of public missionary activity after the apostolic age; in the following two centuries, Christianity mainly spread through door-to-door, individual contact, keeping a low profile (Lane Fox, 1988: 314–317; MacMullen, 1984). This fits the theory in that a non-falsifiable prediction cannot trigger any cognitive dissonance and hence does not require remedial action.

However, the original theory specifies that one of the necessary conditions for disconfirmation to produce increased proselytism is that “for the sake of his belief, [the person] must have taken some important action that is difficult to undo” (Festinger et al., 1956: 4). Gager (1975: 40) emphatically concurs that this was indeed the case for the early Jesus movement. But such “action that is difficult to undo” is nothing but the extreme perfectionist ethics demanded of those who joined the apocalyptic movement, which (as detailed in the historical section above) involved a massive, non-recoupable cost. This nicely fits the twin insights derived from our model: increased effort (in the form of missionary activity) kicks in only when leaving the group upon prophecy failure turns out to be too costly.

Further support for the key role of the initial sacrifice can be found in the social composition of early Christianity. Although it is not a concern of this article to inquire about who the early Christians were or why they joined, the extant evidence on the membership is suggestive of why they did not drop out. We know (Lane Fox, 1988: 293–312) that up to about ad 250 the movement was essentially made up of the lower classes of free, urban society, with a high proportion of women of all ranks: thus, despite exceptions, not the upper classes and not the bottom rungs of the social ladder, namely, peasants and slaves. This was so presumably because the literate upper classes had not yet found a teacher they could take seriously, while nobody bothered to speak to the slaves and even less to go out to the countryside. Whatever the reason, if we believe that the social makeup of the early apocalyptic movement was similar (as a study of the Pauline communities seems to suggest: Meeks, 1983), that is exactly the type of membership that would incur the greatest sacrifice upon joining and therefore the greatest cost of exit. Slaves had nothing to lose to begin with; on the other hand, the elite could always rely on their high-placed connections even in the aftermath of a “mistake.” But what of artisans, petty traders, skilled workmen, freedmen, women—people of little education and modest means: these were the people whose prospects for upward social mobility and/or economic progress were not foreclosed and who therefore stood to lose the most from embracing the wait for the apocalypse. Once they had made the first fateful step, they were likely to hold on.

If we look ahead and ask what happened next, we find ourselves on more familiar ground. When the movement became non-apocalyptic, membership no longer required an upfront sacrifice; members’ commitment and participation could henceforth be sustained by ongoing “prohibitions” or restrictions on secular consumption and behavior, as opposed to one-time sacrifice at entry. The club model of a sect that demands of its members a behavior that is sufficiently at variance with the general society to deter free riders and screen out the low-commitment individuals, as in Iannaccone (1992) and Berman (2000), seems fully adequate as a theoretical workhorse for this situation. But what were those demands?

The one-word answer to this question is: membership itself. The sheer fact of being a Christian in the centuries before Constantine’s edict of toleration of ad 313 involved a frightening loneliness and estrangement from the surrounding society, one that was nowhere more apparent than in the Christians’ refusal to take part in the public cults that were the backbone of civic and political life in the empire. More specifically, nonparticipation in those cults meant social alienation unless one had a good excuse for it. The Jews had a good excuse: an ancestral, exclusive religion that, although odd or even despicable in the eyes of pagan writers, nevertheless deserved respect on account of its antiquity. Christianity, on the contrary, was an upstart that could claim no such ancient pedigree unless it remained within the fold of Judaism. The latter possibility very nearly happened (Brandon, 1957). As we know, Paul successfully pursued a self-appointed mission to create a network of Gentile communities on a fully apocalyptic basis.¹⁶ After his removal from the scene, however, Paul’s opponents in the mother church in Jerusalem regained ascendancy and that church could again command authority over the movement so that the future of Christianity was well set to be that of a Jewish messianic sect. This was abruptly changed by the fall of Jerusalem in ad 70: this catastrophic event completely and permanently obliterated the Jerusalem church and, in retrospect, set the communities outside of Palestine free to follow their own course. In one scholar’s view, “after the Resurrection experiences, the next most crucial event in the life of the Christian Church was the overthrow of the Jewish nation” (Brandon, 1957: 251). However, that same event also set the church free to be completely alone in the pagan world. It is a telling fact that in the following centuries, in stark contrast to the heady enthusiasm and inclusiveness of the early mission, 2–3 years of monitored apprenticeship (called catechumenate) were required before one was admitted to baptism (Lane Fox, 1988: 316–317): being a Christian was evidently understood to be so hard that defection was to be expected, hence the preliminary screening and training to reduce it.

Some comparative evidence from other apocalyptic movements

Apocalyptic movements have been a recurring feature of religious history since the time of Christian origins. While a full review of the field is beyond the scope of this article, a summary look at some important examples will provide further tests of our approach. Specifically, our model predicts that, first, the cost incurred at joining will trap believers in the group, and second, revision of beliefs will set in as a response to prophecy failure and support intensified effort.¹⁷

Broadly speaking, apocalyptic or millenarian movements fall into two classes: those who view themselves as living in the last days, in the sense that the apocalypse is an impending act of God and one’s call is restricted to adequately prepare to receive it, and those who view themselves as playing an active and crucial role in fulfilling God’s plan and bringing about the end of time. In the first case, the apocalypse is an exogenous event; in the second, it is seen as endogenous, that is, in some sense, the reward to one’s effort. The distinction is important because it implies two different responses to history’s failure to end on time.

As we have seen, infant Christianity is the paragon of the first type of movements. There, the end was thought to be very, very near, but a precise date or deadline was never set—“about that day or hour no one knows, neither the angels in heaven, nor the Son, but only the Father. Beware, keep alert; for you do not know when the time will come” (Mark 13: 32–33; repeated in Matthew 24: 36, 42). Thus, there was some flexibility in the prophecy that allowed new converts to be won to the apocalyptic movement for more than a generation following Jesus’ death (the Pauline communities and probably Mark’s community) and at the same time allowed the re-interpretation and de-eschatologization of the Christian message to be accomplished in a gradual, drawn-out fashion down into the 2nd century.

By comparison, other movements of this class set themselves an even more difficult task: expecting the apocalypse to break out on a particular date. We will look at two examples. The American preacher William Miller, after years of preaching the imminent arrival of the Kingdom of God and the need to forsake everything for it, was finally driven by the radical wing of his movement to announce that the Second Coming of the Lord would occur on the hills of Pennsylvania on 22 October 1844—a date that then went down in history as the Great Disappointment. While many of the Millerites thought the calculation was mistaken and continued to set dates for the next few years, eventually disappearing from view, a group embraced the belief that only the event, not the date, had been misinterpreted: on that fateful day,

Christ, the ‘High Priest’, had not come to earth but had moved from the holy to the most holy place in a heavenly sanctuary. The ‘cleansing of the sanctuary’ had not referred to Christ’s Second Coming but rather to the investigation of the sins of God’s people in preparation for the end of the world. (Butler, 1987: 200)

This “doctrine of the sanctuary” was buttressed by the “shut-door doctrine” which held that God’s “door of mercy” had shut on all those who had spurned the Millerite rallying cry, thus confirming the faith of those believers who had held fast in 1844. A few years later, this was replaced by renewed missionary expansion, in the belief that God in his mercy had delayed the Advent to give the first believers more time to save new souls (compare Paul’s argument in Romans, above). Soon the Sabbatarian doctrine—which by sanctifying the Saturday sharply separated the group from other Christian groups focused on the Sunday—was added to the sanctuary doctrine, by 1859 the group had passed from the original anti-clericalism to the ordination of clergy, and one of the most successful contemporary sects was born: the Seventh-day Adventists.

The Jehovah’s Witnesses followed a broadly similar pattern (Stark and Iannaccone, 1997; Zygmunt, 1970). They started out in the 1870s by issuing a date for the end of the world and then reacted to the prophecy’s failure by rescheduling the event, for a total of five dates; when the last date, 1925, again failed to deliver, date setting was suspended and the event was henceforth believed to be close enough but extending into an indefinite future. Each time, the failure was retrospectively re-interpreted as a partial fulfillment: on the date in question, some supernatural event of prophetic significance had indeed occurred, broadly in keeping with the unfolding of God’s plan, while the unfulfilled part of the prior prophecy was projected into the future through the issuance of new predictions. Being otherworldly, these “events” were not liable to disconfirmation. Concurrently, the group’s own role in the apocalypse was redefined from gathering a “little flock” of 144,000 faithful who, according to the Book of Revelation, would assist the returning Christ in his rule over the Kingdom to an expanded missionary program designed to recruit a “Great Company,” of unlimited size, who would be the blessed citizens of the Kingdom when it comes. It is well known that the Witnesses’ missionary program has been a worldwide success to this date.

Like the early Christian case, both the Adventist and the Witnesses’ examples clearly show the mechanism of cognitive dissonance at work: rationalization of prophecy failure represented an effort to reduce the doubt and distress caused by disconfirmation of the original belief; this took the twin forms of re-interpreting events in a confirmatory way and engaging in missionary activity in an effort to maintain the eschatology (Gager, 1975: 46). But underlying this psychological adjustment are the very real, appalling costs incurred by those who first joined the movement. In the period leading up to, and including, the hectic summer of 1844, the Millerites (Butler, 1987: 196–198) rejected status and class distinctions and social inhibitions and turned to nudity, crawling on all fours through the streets to become like little children for the Kingdom, “holy kissing” and foot-washing within the group, sexual license, and self-starvation. Crops were left unharvested, careers were given up, all wealth was shared, and family and friends were left behind for the warmth of the new community. As a result, they were scorned, ridiculed, and ostracized, but in their eyes this only legitimated them as God’s faithful remnant and turned them ever more aggressive toward the established churches. After the Great Disappointment, then, there was not much left for them to go back to. Similarly, the Jehovah’s Witnesses engaged in adversarial behavior toward the churches and the government and literally invited persecution and, at one time, mob violence (Stark and Iannaccone, 1997: 134–137; Zygmunt, 1970: 936–939).

The second group of apocalyptic movements sees the members’ militant action as the agency that brings the millennium into being. The problem these movements have is that if the initial action fails its purpose, they may be driven to escalate their effort against all odds to such extremes as to bring about their own undoing. We will briefly review two cases of success and two cases of failure.

Islam began as an apocalyptic movement. In the 1st century of Islam, roughly spanning the beginnings under Muhammad and his first successors and the Umayyad caliphate, Jesus was thought to be the first Mahdi, the Muslim Messiah. “At this early time, Jesus is the one who is to usher in the messianic kingdom and the ultimate unity between the monotheistic faiths” (Cook, 2005: 83). There are a number of apocalyptic prophecies in the early Muslim religious literature that predict the end of the world during the year 689–690 or the year 717. Like in the New Testament case, the survival of such texts despite the obvious disconfirmation speaks to the power and authenticity of this tradition. These dates correlate with the great pushes to conquest:

Apparently the proto-Muslims believed themselves to have been allotted a bare 70 or 100 years in order to accomplish the goals of conquering the world, reforming and unifying belief in one God, and passing a messianic kingdom to Jesus who would rule it until the Last Judgment. (Cook, 2005: 85)

This belief harnessed the widespread apocalyptic expectations entertained by both Christians and Jews in the Near East around this time and included such religiously charged acts as the construction of the Dome of the Rock in Jerusalem—understood as a kind of rebuilding of the Jewish Temple—and the strenuous efforts at conquering the evil city of Constantinople in order to purify it. At the same time, jihad was not only the method of proclamation of the faith but also the means by which the fighter gained forgiveness for his sins, thus harnessing the collective goal to individual incentives. Although Constantinople did not fall, the Muslim armies did conquer “the world,” or a good part of it, and in this sense, the apocalyptic drive behind jihad was extraordinarily successful, providing a safe environment for the transition from an eschatological to an established religion.

The early Mormons of the 19th century (Arrington and Bitton, 1992) were also inspired by a vision of the Kingdom of God on earth, soon to be ushered in by the Second Coming of the Lord. But this event would not occur until and unless the necessary moral, social, and political conditions were established. This was to be accomplished through “the gathering,” the voluntary removal of the faithful to a separate place where the material foundations could be built for the working out of God’s purposes in history, in a self-conscious re-enactment of the ancient Israelites’ exodus to the Promised Land. After failed attempts in Missouri and Illinois, Utah was finally chosen as the gathering place. The upfront cost was staggering as the Saints uprooted themselves and their families from the surrounding society and their physical removal to a faraway place made any reversal extremely difficult (Arrington and Bitton, 1992: 41). But there was something for them to do, here and now, like a building enterprise whose success could be empirically gauged and which, as it occurred in the wilderness, did not require the previous overthrow of an existing power. In the event, and in no small part for these reasons, Mormon nation building was a resounding success so that, despite the failure of history to end, the transition from millenarian group to established church could smoothly be achieved.

Unlike the Muslim fighters and the Mormon trekkers, the Jewish fighters in the first war against the Romans (ad 66–70) escalated their militancy to the point of bringing on their own destruction (Lewy, 1974: ch. 4). The radical Zealot party, which took control of besieged Jerusalem in the last phase of the war, apparently believed that the God of Israel could not ignore the plight of his people and would come down on their side and ensure their triumph. As we have seen, apocalyptic currents ran deep and wide among the Jews of this period, and to many of them, the revolt seemed to herald the beginning of God’s redemption and the messianic age. The mechanism that triggered the cycle of destruction was the belief that the bleaker the situation, the surer it was that God was about to come; therefore, a purposeful worsening of conditions was believed to be the trigger of eventual salvation. Thus, in the final weeks of the siege, the Zealots burned the city’s remaining food supply, and at the very end, thousands of people crammed in the Temple precinct and awaited for the Romans to slaughter them while joyfully trusting the Lord to appear at any moment. Mindful of the tragedy, the rabbis later condemned the very idea of “forcing God’s hand”—but for now the apocalyptic mission ended in utter disaster.¹⁸

The “kingdom” established by the radical wing of the Anabaptists in the German city of Münster in 1534–1535 provides a striking replica of the Jerusalem revolt (Cohn, 1961: ch. 12; Lewy, 1974: ch. 5). The similarities strongly suggest a common underlying pattern because the Anabaptists could not possibly have read the chief surviving source on the earlier event—the Jewish historian Josephus—and so they could not be consciously re-enacting the fall of Jerusalem. Here too, the believers took over the city, forced the Lutheran population to re-baptize or flee, and established the New Jerusalem on the model of the Book of Revelation. A prophet set himself up as king of the “reign of the saints,” all property as well as women were held in common use, and a militia sought out and executed the lukewarm and the weak, in a spiral of progressive radicalization. Since the apocalypse would be worldwide, although initiated in Münster, the fighters sent out appeals to insurgency to nearby cities and took on an overwhelmingly superior force mustered by the local bishop, which laid siege to the city. The more dire the situation, with famine and disease weakening the defenders, the more enthusiastic and defiant became their attitude, in the certainty that God would not fail to come to their help. In the end, those who did not starve to death were nearly all slaughtered.¹⁹

In these militant apocalyptic groups, then, the reaction to cognitive dissonance is obvious—the bleaker the predicament, the greater the certainty of the final redemption—and the upfront costs of joining the movement are truly enormous—high risk of death, not just social ostracism. But unless they are lucky enough for unrelated reasons, as the early Muslims were, or wise enough to resettle in a desert, as the Mormons were, the use to which they put their initial investment is such as to preclude post-eschatological transition and invite self-destruction.

Conclusion

This article has suggested an economic explanation for the birth of Christianity: the high cost of dropping out of a failed apocalyptic movement. Unlike most of existing historical research, the emphasis here is not on why some people joined but why they stayed. The only rational explanation that had previously been suggested—reaction to cognitive dissonance—is modeled here as an important factor but subordinate to exit costs. A selective review of other apocalyptic movements in religious history has been shown to provide additional support for the central argument of this article.

In general, the message of this article to economists is that the birth of a new religion—a problem never addressed before in economics—can be modeled while fully accounting for participants’ rationality; in particular, the importance of the central case study here under examination—the birth of Christianity—does not need to be belabored. Our central finding is that one important way to start a new religion from scratch—an exceedingly difficult proposition if people are rational—may be to start from a mistake, provided it is the “right” kind of mistake. This is because mistakes are costly and that cost can be turned into a constraint that prevents defection. On the other hand, the message of this article to historians of Jesus and Christian origins is that an economic approach can provide additional support to the apocalyptic view of both Jesus and early Christianity, whereas no such support can be extended to any non-apocalyptic view in which, logically, the upfront cost of joining must be secondary or nonexistent.

Footnotes

Acknowledgements

An earlier draft of this article was presented at the annual meetings of the Canadian Economic Association in Ottawa, 3–5 June 2011, where participants provided interesting discussion. The author is indebted to Maryam Dilmaghani, Hugh Neary, and two referees of this journal for useful comments and especially to Franco Cugno for extensive discussion and advice on the model.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Akerlof

Dickens

(1982) The economic consequences of cognitive dissonance. American Economic Review 72(3): 307–319.

Arrington

Bitton

(1992) The Mormon Experience: A History of the Latter-day Saints. 2nd ed. Urbana, IL and Chicago, IL: University of Illinois Press.

Bader

(1999) When prophecy passes unnoticed: new perspectives on failed prophecy. Journal for the Scientific Study of Religion 38(1): 119–131.

Berman

(2000) Sect, subsidy and sacrifice: an economist’s view of Ultra-Orthodox Jews. Quarterly Journal of Economics 115(3): 905–953.

Brams

(2011) Game Theory and the Humanities: Bridging Two Worlds. Cambridge, MA: MIT Press.

Brandon

SGF

(1957) The Fall of Jerusalem and the Christian Church. 2nd ed. London: SPCK Publishing.

Butler

(1987) The making of a new order: Millerism and the origins of Seventh-day Adventism. In: Numbers

Butler

(eds) The Disappointed: Millerism and Millenarianism in the Nineteenth Century. Bloomington, IN and Indianapolis, IN: Indiana University Press, pp. 189–208.

Cohn

(1961) The Pursuit of the Millennium. 2nd ed. New York: Harper & Row.

Cohn

(2001) Cosmos, Chaos, and the World to Come: The Ancient Roots of Apocalyptic Faiths. 2nd ed. New Haven, CT and London: Yale University Press.

10.

Cook

(2005) The beginnings of Islam as an apocalyptic movement. In: O’Leary

McGhee

(eds) War in Heaven, Heaven on Earth. London and Oakville, CT: Equinox Publishing, pp. 79–93.

11.

Crossan

(1994) Jesus: A Revolutionary Biography. San Francisco, CA: HarperSanFrancisco.

12.

Eisenbaum

(2009) Paul Was Not a Christian. New York: HarperCollins.

13.

Ekelund

Tollison

(2011) Economic Origins of Roman Christianity. Chicago, IL and London: University of Chicago Press.

14.

Ferrero

(2008) The triumph of Christianity in the Roman Empire: an economic interpretation. European Journal of Political Economy 24(1): 73–87.

15.

Ferrero

(2013) The cult of martyrs. Journal of Conflict Resolution 57(5): 877–900.

16.

Ferrero

(2014a) Competition between Judaism and Christianity: Paul’s Galatians as entry deterrence. Kyklos 67(2): 204–226.

17.

Ferrero

(2014b) Jesus and the Ratchet (Unpublished manuscript). Alessandria: University of Eastern Piedmont.

18.

Festinger

Riecken

Schachter

(1956) When Prophecy Fails: A Social and Psychological Study of a Modern Group that Predicted the Destruction of the World. New York: Harper & Row.

19.

Fredriksen

(1991a) Apocalypse and redemption in early Christianity: from John of Patmos to Augustine of Hippo. Vigiliae Christianae 45(2): 151–183.

20.

Fredriksen

(1991b) Judaism, the circumcision of Gentiles, and apocalyptic hope: another look at Galatians 1 and 2. Journal of Theological Studies 42(2): 532–564.

21.

Fredriksen

(1995) What you see is what you get: context and content in current research on the historical Jesus. Theology Today 52: 75–97.

22.

Fredriksen

(1999) Jesus of Nazareth, King of the Jews. New York: Vintage Books.

23.

Fredriksen

(2000) From Jesus to Christ: The Origins of the New Testament Images of Christ. 2nd ed. New Haven, CT and London: Yale University Press.

24.

Frend

WHC

(1984) The Rise of Christianity. London: Darton, Longman & Todd.

25.

Gager

(1975) Kingdom and Community: The Social World of Early Christianity. Englewood Cliffs, NJ: Prentice-Hall.

26.

Gager

(1988) Proselytism and exclusivity in early Christianity. In: Marty

Greenspahn

(eds) Pushing the Faith: Proselytism and Civility in a Pluralistic World. New York: Crossroad, pp. 67–77.

27.

Goodman

(1994) Mission and Conversion: Proselytizing in the Religious History of the Roman Empire. Oxford: Clarendon Press.

28.

Grant

(1977) Jesus: An Historian’s Review of the Gospels. New York: Charles Scribner’s Sons.

29.

Iannaccone

(1992) Sacrifice and stigma: reducing free riding in cults, communes, and other collectives. Journal of Political Economy 100(2): 271–291.

30.

Kümmel

(1957) Promise and Fulfilment: The Eschatological Message of Jesus. Naperville, IL: Alec R. Allenson.

31.

Lane Fox

(1988) Pagans and Christians. San Francisco, CA: Harper & Row.

32.

Latourette

(1954) A History of Christianity. London: Eyre & Spottiswoode.

33.

Lewy

(1974) Religion and Revolution. New York: Oxford University Press.

34.

MacMullen

(1984) Christianizing the Roman Empire, AD 100-400. New Haven, CT: Yale University Press.

35.

Meeks

(1983) The First Urban Christians: The Social World of the Apostle Paul. New Haven, CT and London: Yale University Press.

36.

Montgomery

(1996) Dynamics of the religious economy: exit, voice and denominational secularization. Rationality and Society 8(1): 81–110.

37.

Rowland

(1985) Christian Origins: From Messianic Movement to Christian Religion. Minneapolis, MN: Augsburg Fortress Publishers.

38.

Sanders

(1993) The Historical Figure of Jesus. London: Penguin Books.

39.

Stark

(1996) The Rise of Christianity: A Sociologist Reconsiders History. Princeton, NJ: Princeton University Press.

40.

Stark

Iannaccone

(1997) Why the Jehovah’s Witnesses grow so rapidly: a theoretical application. Journal of Contemporary Religion 12(2): 133–157.

41.

Zygmunt

(1970) Prophetic failure and chiliastic identity: the case of Jehovah’s Witnesses. American Journal of Sociology 75(6): 926–948.