Interdisciplinary Journal of
Research on Religion
______________________________________________________
Volume 1 2005 Article 7
______________________________________________________
Does Religion Promote Trust?
The Role of Signaling, Reputation, and
Punishment
Richard Sosis
Senior Lecturer, Hebrew University of Jerusalem
and
Associate Professor of Anthropology, University of Connecticut*
*richard.sosis@uconn.edu
Copyright © 2005 Interdisciplinary Journal of Research on Religion. All rights reserved. No part
of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or
by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior
written permission of the publisher. The Interdisciplinary Journal of Research on Religion is
freely available on the World Wide Web at http://www.religjournal.com.
Does Religion Promote Trust?
The Role of Signaling, Reputation, and
Punishment*
Richard Sosis
Hebrew University of Jerusalem and University of Connecticut
Abstract
Scholars of religion, including such luminaries as Durkheim, Rappaport, Turner, and Weber, have
widely assumed that religion promotes intragroup trust among adherents. Recent applications of
signaling theory to religious behavior among economists, cognitive scientists, and evolutionary
anthropologists further endorse this assumption. However, trust has not been rigorously or
consistently defined across authors, making generalizations difficult to evaluate. Here I follow
Bacharach and Gambetta’s (2001) behavioral definition of trust and show that the conditions for
intragroup trust are often not met in religious communities, especially isolationist and closed
communities to which high levels of trust are typically ascribed. Rather, in such communities,
cooperation is maintained through institutional structures that effectively punish cheaters and
enhance the value of an honest reputation. These groups gainfully facilitate collective action by
offering a circumscribed social arena in which reputations can be built, evaluated, rewarded, and
efficiently punished. While face-to-face reciprocal relations obviate the need for trusting behavior
within closed religious communities, when social groups are fluid, religious practices and
symbolic markers are successful in promoting trust among in-group members and anonymous
coreligionists who reside in different communities. In addition, these religious badges of identity
may be used by non-group members as signals of trustworthiness.
*I thank Candace Alcorta, Joseph Bulbulia, Sharon Feldstein, Bradley Ruffle, and two anonymous
reviewers for very helpful comments on earlier drafts of this manuscript. I additionally thank the
Russell Sage Foundation (Initiative on Trust) and the University of Connecticut for generous
funding of this project.
Sosis: Does Religion Promote Trust? 3
Trust none; for oaths are straws, men’s faiths are wafer-cakes …
William Shakespeare
The Life of King Henry the Fifth (Pistol at II, iii)
[A] group within which there is extensive trustworthiness and extensive trust
is able to accomplish much more than a comparable group without that
trustworthiness and trust.
Coleman (1988: S101)
While it is broadly recognized that both religion and trust are directly relevant to
current world affairs, surprisingly little attention has been given to how they are
related to each other. This is particularly notable because recent theoretical
developments in the study of religion posit that trust plays an important role in
generating the cooperatively derived benefits that religious groups are presumed
to offer (see Sosis and Alcorta 2003). Economists (Berman 2000; Carr and Landa
1983; Iannaccone 1992, 1994), cognitive scientists (Atran 2002; Bulbulia 2004a,
2004b), and evolutionary anthropologists (Cronk 1994; Irons 1996a, 1996b,
1996c; 2001, 2004; Sosis 2003, 2004; Sosis and Alcorta 2003) have converged in
applying signaling theory to model and explain religious behaviors that are
puzzling from the perspectives of the egoistic-based models that serve as these
disciplines’ foundations (e.g., rational choice, evolutionary game theory). These
researchers have argued that religious behaviors signal group commitments, thus
enhancing intragroup trust and facilitating collective action.
Various authors working outside the signaling framework have also asserted
that religion enhances trust among community members (e.g., Boudon 1987;
Collins 2004; Shield 2002; Steadman and Palmer 1995; Weber 1958: 302–322).
However, inconsistency in definitions of trust have limited the ability to make
generalizations about this relationship. Even in our everyday use of the term, there
is ambiguity.1
Colloquially, we may employ the word trust to characterize the
attitude of agents toward other agents. Thus trust sometimes denotes an attitude of
confidence about another agent’s reliability. However, we also use trust to
describe behavior. Accordingly, to trust is to act on this attitude of confidence
about another agent’s reliability. The statement “Richard trusts Joseph” may refer
to a specific attitude that Richard maintains about Joseph, or it may refer to an
actual behavior that manifests itself as a consequence of the attitude—for
instance, that Richard trusts Joseph with his wallet. In the discussion that follows,
all references to trust refer to its behavioral usage, not its attitudinal one. This
focus on the behavioral aspects of trust is motivated by the fact that observers of
religion have generally commented on the remarkable levels of trusting behavior
1
In my trusty American College Dictionary, trust is defined as a noun, a verb, and an adjective,
and 23 definitions are offered. No wonder this term has caused so much confusion!
4 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
exhibited by members of certain religious communities,2
with less emphasis on
the more difficult to observe and measure attitudes and values that encourage
these behavioral patterns.
To address the question posed in the title of this article, I first offer a
definition of trust that can be easily incorporated into signaling models of
religious behavior and delineates between various types of trust. I clarify the
benefits that enhanced trust has for individuals and groups and the conditions
under which we would expect religious communities to exhibit high levels of
intragroup trust. I then examine the prevalence of trust within several well-studied
religious communities. To summarize the main argument, while many researchers
have maintained that religion increases trust, I argue that this may in fact be
illusory. Most of what observers claim is trust in religious communities,
especially those communities that demand exclusive commitment from their
members, is actually a consequence of institutional frameworks that effectively
punish cheaters and support gains for those who maintain an honest reputation.
These communities, which I refer to as “closed” (see below), can efficiently
punish untrustworthiness and offer reputational histories because social relations
are bounded. This circumscribed social arena is created by religious rituals and
symbolic markers (e.g., clothing, hairstyles) that clearly signal and demarcate inand
out-group membership. However, when group memberships are fluid and
individuals maintain multiple group identities, punishment and reputational
mechanisms are less influential. Consequently, these “open” religious
communities often rely on religiously endorsed practices and symbolic markers to
foster trust. Religious behaviors and badges also promote trust between
anonymous in-group members residing in different communities and generate
epiphenomenal benefits from increased perceived trustworthiness by nongroup
members.
WHAT IS TRUST?
As any reader of this journal will be aware, religion has as many definitions as
there are scholars who have attempted to define it (for reviews, see Klass 1995;
Spiro 1966). Trust, on the other hand, does not seem to suffer from this gratuitous
attention, as most researchers discuss trust without any clarification about what it
actually is. Indeed, while trust is the core issue of all signaling-based theories of
religion, none of this literature attempts to define or measure trust. Here, I begin
with a definition of trust offered by Bacharach and Gambetta (2001: 151): “a
2
Throughout this article, I use the term religious community to denote a self-defined group that is
united by a core set of religious beliefs and distinguishing ritual practices. Typically, but not
always, religious communities are organized around one or more churches, mosques, synagogues,
temples, or some other institutional or noninstitutional houses of worship.
Sosis: Does Religion Promote Trust? 5
person trusts someone to do X if she acts on the expectation that he will do X
when two conditions obtain: Both know that if he fails to do X, she would have
done better to act otherwise, and her acting in the way she does gives him a
selfish reason not to do X.”3
Bacharach and Gambetta define a simple two-person
trust game according to the following payoffs, where x < 3 and y > –3:
Trustee
B B'
A 3 , 1 –3 , 4
Truster
A' x , — y , —
In this trust game, A and B are trusting and trustworthy decisions by the Truster
and Trustee, respectively, whereas A' and B' are distrustful and defecting
decisions.
To illustrate, consider the following anecdote about a tourist who visited New
York City for the first time. After arriving at Port Authority by bus, he was
approached by a man who asked him whether he needed a cab. The tourist was
pleased to be offered assistance and consented. The man took both his bags and
told him to wait while he pulled his cab around. Bacharach and Gambetta’s
definition of trust accurately characterizes the payoffs in this interaction. The
Truster stood to gain an easy cab ride to his hotel but risked losing his luggage.
The Trustee stood to gain a cab fare or, more profitably, two pieces of luggage
and their contents. Unfortunately for this tourist, who happens to be a friend of
mine, the Trustee chose the latter, and the Truster never saw his belongings again.
While Bacharach and Gambetta’s definition aptly depicts my friend’s
initiation to big city life, this definition needs to be extended to capture additional
forms of trust. Here, I consider two important deviations from their payoff matrix.
First, Trusters might not always benefit from their actions, even if the Trustee is
trustworthy. The pedestrian who accedes the stranger’s request to borrow his cell
phone to make an urgent call, for example, is unlikely to receive any gain from
his trusting behavior. This type of nonprofitable trust, in which short- or longterm
material gains cannot motivate the Truster, has rarely been discussed or
3
Another influential definition of trust is offered by Coleman (1990: 99). He argues that trust
constitutes a subclass of situations involving risk, namely, those interactions in which the risk
entails the behavior of another actor. The significant difference between Coleman’s definition,
which he presents algebraically, and the definition offered by Bacharach and Gambetta is that the
preferences of the Trustee in Coleman’s definition are reduced to a probability. For Coleman, it
might or might not be in the Trustee’s self-interest to be untrustworthy; for Bacharach and
Gambetta, the risk for the Truster is that the Trustee can always gain by being untrustworthy.
6 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
examined, so it is not clear how frequently and under what conditions it occurs.4
In the above payoff matrix, it exists when the trusting action, A, is a dominated
strategy, namely, x > 3, y > –3. Rational choice and game theory, of course,
predict that nonprofitable trust will not occur unless short-term potential losses
can be recouped through future gains.
Second, encounters involving trust are often embedded in a series of
interactions. Accordingly, it is worth distinguishing between repeated and
nonrepeated trust. For many scholars, repeated interactions are essential for the
emergence and stability of trust. For example, Dasgupta (1988: 49) maintains that
“[f]or trust to be developed between individuals they must have repeated
encounters, and they must have some memory of previous experiences.” Cook
and Hardin (2001: 328) claim that the principal incentive to form trusting
relationships is “the desire to maintain the relationship itself for future activities.”
Intuitively, we assume that individuals will lend their cell phones more readily to
a neighbor than to a stranger, since they are more likely to interact with the
neighbor in the future.
Repeated trust interactions can take two forms. In some interactions, the roles
of Truster and Trustee are reversed over time; in other interactions, the roles are
maintained for the duration of the relationship or are rarely reversed. Coleman
(1990: 177–178) refers to these types of trust as mutual and asymmetric trust,
respectively. While the latter often characterizes the relationship between
religious leaders (Trustee) and adherents (Truster), this paper will not explore this
relationship, which deserves a separate comprehensive analysis, but rather will
focus on the symmetric and asymmetric interactions between laypeople in
religious communities.
Notice that trust has been defined without any reference to the costs and
benefits of reputations or punishments. These factors, however, can be easily
incorporated into the payoff structure and in fact may have already been assumed
in the cell phone example above. Aside from the greater knowledge we have
about the trustworthiness of our neighbor than about that of a stranger, we are also
more inclined to loan a cell phone to a neighbor because we stand to gain
reputational benefits from our trusting behavior and can impose greater
punishments on a neighbor if he calls Thailand rather than his wife who is
expecting him home for dinner. Incorporating reputations and punishments into
4
For instance, most economic experiments that are aimed at measuring trust assume, consistent
with Bacharach and Gambetta’s definition, that the Truster can potentially benefit from a trusting
relationship if the Trustee is trustworthy (e.g., Barclay 2004; Berg, Dickhaut, and McCabe 1995;
Bohnet and Croson 2004; Croson and Buchan 1999; Glaeser et al. 2000; King-Casas et al. 2005;
McCabe, Rigdon, and Smith 2003; McKelvey and Palfrey 1992). I refer to this type of trust as
profitable.
Sosis: Does Religion Promote Trust? 7
the payoffs may actually “solve” the trust dilemma by explaining it away.5
For
instance, reliable punishments can change the payoffs of an interaction so that
trust is not a behavioral option. As Coleman (1990: 115) notes, “[i]f a trustee’s
behavior is affected by the application of sanctions, then a trustor’s decision about
whether to place trust in that trustee another time should be based not simply on
his estimate of the probability of the trustee’s keeping the trust, but also in part on
the use of negative sanctions.” If a Trustee can be punished for an untrustworthy
decision (B') such that he receives 0 instead of 4, the interaction no longer offers
the possibility of trust, since it is in both the Truster’s and Trustee’s self-interest
to act cooperatively. Similarly, if Trustees can gain reputational benefits from
trustworthiness, it may be in their self-interest to choose B instead of B', thus
eliminating the trust dilemma.
Now that we have our working definitions of trust, let me clarify the primary
question of this article, which was offered in its title. To paraphrase, with
admittedly less concision: Are individuals within a religious community more
likely to make a decision that conforms to the payoffs defined by Bacharach and
Gambetta as trust than individuals who are not in a religious community? When
given the opportunity, the answer to this question is likely to be “yes.” But more
commonly, especially in closed religious groups,6
these payoffs do not depict the
range of prosocial interactions that religious adherents face within their
communities. Closed religious groups typically achieve their collective goals by
altering these payoffs rather than by relying on trust. Paradoxically, it is in
religious groups where membership is fluid, demands are fewer, and the structure
is less formal that trust as defined above is more commonly observed.
RELIGIOUS SIGNALING, COOPERATION, AND TRUSTWORTHINESS
Anthropologists have long noted that one of the primary functions of religion is
the promotion of group solidarity, and most have recognized ritual as the
5
Punishments, though, may generate second-order collective action problems that require trust to
solve (Heckathorn 1989). Johnson, Stopka, and Knights (2003) argue that belief in supernatural
punishment can solve this second-order dilemma (see below).
6
My concern here is to distinguish between communities that demand the exclusive identity of
their members, typically through isolation or enforced restrictions on the ability to interact with
non-group members, and communities that do not impose, either environmentally or socially, such
demands on their adherents. While sects and cults are often characterized by such features, they
also imply other features that are not assumed here (Stark and Bainbridge 1985). In addition, these
terms tend to be limited to the landscape of world religions and do not capture the indigenous
religions that are often of interest to anthropologists. I will therefore refer to these communities as
closed and open (or fluid), respectively, recognizing that religious groups lie along a continuum of
segregation and openness. For example, within the American Jewish community, Ultra-Orthodoxy
and Reform Judaism may lie at opposite ends of this continuum, with Modern Orthodoxy and
Conservative Judaism situated somewhere between these extremes.
8 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
mechanism through which this solidarity is achieved. Guided by Durkheim (1995
[1912]), who was among the first to appreciate the unifying nature of religious
ritual, anthropologists have argued that collective rituals enable the expression
and reaffirmation of shared beliefs, norms, and values and are thus essential for
maintaining communal stability and group harmony (e.g., Douglas 1966;
Radcliffe-Brown 1952). For Durkheim, collective rituals are the means by which
individuals bond with one another in the community. He claims that the
“effervescent” state of ritual performance minimizes individual distinctions and
emphasizes the unity of the group. Turner (1969) also views this “effervescent”
state as central to ritual’s efficacy. He observes that the temporary leveling of
initiates during rites of passage increases a sense of communitas among ritual
performers, which he characterizes as a strengthening of social bonds and
heightened communion. As Stark (2001: 183) succinctly concludes, “Participation
in religious rituals builds group solidarity.”
More recently, evolutionary scholars have argued that social bonding is not an
end in itself but a means to facilitate intragroup cooperation (Bulbulia 2004b;
Cronk 1994; Irons 2001, 2004; Nesse 1999; Rappaport 1999; Sosis 2003; Sosis
and Alcorta 2003; Steadman and Palmer 1995; Wilson 2002). Large-scale
cooperation, however, is notoriously difficult to attain without social mechanisms
that limit the potential to free ride on the efforts of others (Dawes 1980; Hardin
1982; Olson 1965). The core challenge in achieving cooperative goals is solving
problems of trust and commitment (Frank 1988, 2001; Schelling 1960, 2001).
Intragroup cooperation is most likely to emerge when individual group members
can guarantee their participation. The dilemma in human social interactions is
how to generate a credible guarantee—in other words, one that can be trusted.
Simply advertising a willingness to cooperate is fraught with potential deception.
When faced with the conditions of collective action, the incentive to display false
commitment signals is especially high, since individuals can achieve their greatest
gains by refraining from cooperation while others cooperate. Therefore, whenever
an individual can achieve net benefits from defection, credible signals of
cooperative intentions tend to be those that are too costly for defectors to imitate
(Spence 1974; Zahavi and Zahavi 1997).
Anthropologists (Cronk 1994; Irons 1996a, 2001; Sosis 2003), economists
(Berman 2000; Carr and Landa 1983; Iannaccone 1992, 1994), and cognitive
scientists (Atran 2002; Bulbulia 2004a, 2004b) have independently suggested that
religious practices and emotions are indeed such signals. Observers of religion
have repeatedly noted the costliness of religious obligations. Religious behaviors
often entail significant proximate costs, such as time, energy, and material costs,
as well as physical and psychological pain. For example, many populations
require males and females to undergo initiation rites that include beatings, genital
mutilation, exposure to extreme temperatures, tattooing, isolation, food and water
Sosis: Does Religion Promote Trust? 9
deprivation, consumption of toxic substances, and death threats (e.g., Tuzin 1982;
Whiting, Kluckholn, and Anthony 1958; Young 1965). In literate societies,
religious legal codes (e.g., Laws of Manu, Shari’a, Talmud) outlining appropriate
behavior tend to be formalized and regulate a wide range of activities, including
food consumption, work, charitable obligations, and dress. These codes also
define the frequency and structure of ritual ceremony and prayer. Although there
may be physical or mental health benefits associated with some ritual practices
(see Levin 1994; McCullough et al. 2000; Reynolds and Tanner 1995), the
significant time, energy, and material costs involved in imitating such behavior
serve as effective deterrents for anyone who does not accept the teachings of a
particular religion. Therefore religions often maintain intragroup solidarity by
requiring costly behavioral patterns of group members. The performance of these
costly behaviors signals and engenders commitment and loyalty to the group and
the beliefs of its members. Trust is enhanced among adherents, thereby
facilitating cooperative pursuits.
Evolutionary models offer further insight into the efficacy and persistence of
costly religious practices. Bliege Bird and Smith (2005) outline four necessary
conditions for the evolutionary stability of a costly signal in a population: (1)
There is within-group variance in some unobservable attribute, (2) group
members can benefit from reliable information about this variance, (3) signalers
can achieve benefits at the expense of those receiving the signal, and (4) the cost
or benefit to the signaler of sending the signal is correlated with the signaler’s
quality of the attribute. Sosis (2003) has argued that religious behaviors meet
these conditions: (1) The intensity of religious beliefs varies within communities,
and this variance is unobservable; (2) individuals benefit from accurate
information about this variance because intensity of belief is related to one’s
commitment to the group and its goals; committed members are more likely to be
cooperators and thus preferred social interactants; (3) religious groups offer
various benefits for in-group members that are mutually provided and are at risk
of exploitation by those not committed to group goals; and (4) the cost or benefit
of ritual performance is weighed against ritual opportunity costs that are expected
to be higher for skeptics than for believers. Thus religious behavior can be
understood as a costly signal that reliably advertises the unobservable condition of
religious belief and group commitment. The time, energy, material, and
opportunity costs of religious activity serve to prevent those who lack sufficient
belief from displaying the signal.
While prior comparative, ethnohistorical, and experimental studies (Ruffle
and Sosis n.d.; Sosis 2000; Sosis and Bressler 2003; Sosis, Kress, and Boster
2004; Sosis and Ruffle 2003, 2004) have been largely supportive of the
application of costly signaling theory to religion, this research has focused
exclusively on demonstrating a relationship between collective ritual practice and
10 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
cooperative behavior.7
None of this work has assessed whether trust mediates the
observed relationship between costly religious practice and cooperation. Nor has
the pioneering club goods model of religious behavior developed by economists,
and its concomitant evaluation, focused on this relationship (e.g., Berman 2000;
Iannaccone 1992, 1994). To address this lacuna, below I examine the role of trust
in fostering and sustaining cooperation in several well-studied religious
populations.
As was noted above, many scholars assume that religion promotes trust,
thereby facilitating collective action, yet the lack of specificity about the meaning
of trust has resulted in numerous claims that may be unwarranted. Debates about
the existence of trust in bucolic areas (Amato 1993) are instructive. Consistent
with the idyllic stereotype, some researchers (Fukuyama 1995; Putnam 2000)
attribute high levels of trust to small towns in rural America. Cook and Hardin
(2001), however, argue that small communities achieve high levels of cooperation
and reciprocity through repeated interactions, direct information sharing about
individual behavior, monitoring, and sanctioning. In a small community, the
threat of ostracism is significant, especially if the costs of leaving the community
are high (e.g., substantial land or business investments, or extensive family or
social ties in the community) or there are barriers to entering other communities.
If migration is costly or the threat of punishment looms, prosocial reciprocal
interactions can be rational and strategic, not requiring any trust.
Rappaport (1999: 204) offers similar observations about various indigenous
populations:
In tribal societies, ethics are an immediate and perceptible aspect of relations
among people who are, for the most part, not only known to each other but stand
in well-defined relationships to each other. Reciprocal (although not necessarily
symmetrical) obligation is the cement if not, in fact, the ground of all such
relationships, and the obligations they entail are usually quite clearly specified.
Violations of obligation inevitably become evident, often quickly, and sanctions
against breach of obligation are essential elements of reciprocity’s fundamental
structure. … That prestige as much as or even more than wealth is among the
chief rewards of life properly lived in societies in which reciprocity prevails also
encourages vigorous, valorous and generous fulfillment of obligation.
7
Consistent with these results, Orbell et al. (1992) found that church attendance among Mormons
in Logan, Utah, where over 75% of the population are members of the Church of Latter-day
Saints, was positively correlated with cooperation toward anonymous strangers in prisoner’s
dilemma experiments. In a more religiously diverse area, no correlation was found, suggesting that
ritually generated cooperation is aimed at fostering intragroup prosociality rather than generalized
or universal prosociality.
Sosis: Does Religion Promote Trust? 11
Communities that are defined by their religious identity may often achieve
cooperation through similar means. Although contemporary world religions have
members dispersed across the globe, adherents generally live in small, welldefined
groups that are distinguished by their collective performance of rituals
(Luhmann 1997; Wilson 2002). Many religious communities, past and present,
have established social control through the threat of ostracism as well as formal
punishment. As in tribal and rural communities, adherents develop long-term
relationships and are able to observe the behavior of other residents, thus possibly
solving the trust dilemma through reputational rewards and punishment threats
that alter the payoffs of social interactions so that trust is unnecessary.
It is ostensibly in the economic sphere that religious groups have benefited
most from the high levels of trust attributed to them. There are countless examples
of small religious groups that have profitably dominated distinctive economic
niches that require considerable trust and cooperation (e.g., Landa 1994). I turn to
three such populations—Jewish diamond merchants, Maghribi traders, and
Muslim traders in Africa—to assess how they overcome the considerable trust
dilemmas posed by their commercial activities. Each population has been widely
studied and offers an opportunity to examine the relationship between signaling,
trust, and cooperation.
Jewish Diamond Merchants
The business dealings of Jews in diamond centers such as Antwerp, Amsterdam,
London, New York, and Tel Aviv are often cited as a classic example of trust
(e.g., Coleman 1990; Gutwirth 1968; Putnam 2000; Starr 1984). For instance,
Wechsberg (1966) describes buyers in London walking around “with a hundred
thousand pounds worth of diamonds that were handed over to them in trust” (cited
in Coleman 1990: 109). Despite the opportunity for theft and fraud between
buyers and sellers, “[a]ll deals are settled verbally” (Coleman 1990: 109). In a
rare ethnographic study of New York’s diamond district,8
Shield (2002: 103–105)
describes the extraordinary level of trust among Ultra-Orthodox Jews;
“handshakes, Yiddish, and trust still close multi-million dollar deals” (Shield
2002: 1).
As was defined above, trust therefore appears to be profitable, repeated, and
either mutual or asymmetric, depending on the specific relationship. However,
Williamson (1993: 471) argues that the “appearance of trust among diamond
dealers is deceptive.” Williamson follows Bernstein (1992), who claims that trust
is not critical or even important for business activity among New York’s Jewish
diamond dealers. Instead, reputation and social bonds are backed up by a private
8
New York’s diamond district is “more than 95% Jewish” (Shield 2002: 15).
12 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
arbitration system that uses reputation and the threat of expulsion to maintain lowcost
commercial transactions. Williamson asserts that Jews succeeded in the
diamond business not because of trust, but because they were “able to provide
cost-effective sanctions more efficiently than rivals” (1993: 472).
In a recent analysis, Richman (2004) questions why diamond merchants have
rejected public courts in favor of a private internal adjudication system. He argues
that the unique nature of the diamond industry, namely, regularly handling small,
extremely valuable goods that can be easily resold in distant markets, makes the
public legal system incapable of enforcing contracts between diamond sellers and
buyers. While Bernstein (1992) notes the importance of reputations supported by
the internal adjudication systems, Richman recognizes that the high value of
diamonds makes one-time theft more profitable than long-term benefits accrued
from an honest reputation, thus making a scrupulous business reputation alone
insufficient to maintain reliable contracts and exchanges. It is more likely that the
added social costs of defection within tight-knit Ultra-Orthodox communities
enables these communities to maintain their competitive edge in this industry by
keeping transaction costs low. In other words, because ruined reputations affect
not only business opportunities but also one’s social life, including diminishing
the marriage prospects of one’s children and siblings, these threats help to
maintain cooperative exchanges. Whether meted out socially or through the
private court system of diamond traders, these costs alter the payoffs of
transactions, so it is in everyone’s interest to pursue honest business dealings.
Therefore, the success of Jewish diamond merchants does not appear to depend
on any form of trust.
Maghribi Traders
Many scholars have claimed that medieval trade flourished as a consequence of
trust relations successfully established among “natural groups,” including
religious communities (e.g., Landa 1994; Rosenberg and Birdzell 1986; Sombart
1953; Weber 1961). Maghribi traders, a group of 11th to12th century Jewish
merchants who conducted commerce extensively throughout the Mediterranean,
would seem to be a case in point (Greif 1989). All medieval merchants faced a
significant quandary: How should they manage their goods abroad? If they
traveled with their merchandise, this would limit the amount of business they
could conduct and limit their ability to diversify their trading activity, a useful
strategy to buffer against failed ventures. However, if they hired an agent to
manage their goods abroad, they could not observe the transactions and would
have to rely on the agent to honestly report the selling prices of the goods.
Relying on the courts to punish dishonest agents was an ineffective deterrent,
since the costs to monitor such transactions were too high. To overcome these
Sosis: Does Religion Promote Trust? 13
difficulties, Maghribi traders arranged agency relations through a peer
organization, which Greif (1989, 1993) refers to as a coalition. Greif argues that
the coalition was successful at maintaining honest and trustworthy agents because
the short-term gains that could be achieved from cheating (e.g., stealing goods,
inaccurately reporting a sale price) were outweighed by the long-term gains
achieved by remaining in the coalition.
Greif (1989) refers to the merchant-agent relations as one of trust,9
but it is
evident that the payoffs defined above do not characterize these relations. Agents
who cheated a merchant were punished by a permanent refusal to conduct future
business with all members of the coalition. Hence upholding an honest reputation
was crucial for remaining in the coalition and receiving its benefits. Both
merchant and agent were aware of the payoffs involved: “The agent required a
reputation of being honest, and the merchant could trust him” (Greif 1991: 460).
Yet given the institutional framework, reliance on reputations, and effective
means of punishment, Maghribi merchants did not need to trust their agents. As
Greif argues, the agents’ payoffs for honesty outweighed their payoffs for
defection. Greif (1989: 881) further contends, “trust did not reflect a social control
system or an internalization of norms of behavior.” He rejects the possibility that
trust between Maghribi traders was maintained through reputations signaling
trustworthiness “because he fears God, or has internalized an ideology of
honesty” (Greif 1989: 867). Rather, Greif claims, the successful merchant-agent
transactions are explained by the reputational mechanism and credible threats of
punishment.
Muslim Traders in Africa
Muslim merchants have been credited with the spread of Islam throughout Africa
and Southeast Asia, including Indonesia, the Philippines, and Malaysia (Stark
2001:84). Focusing on Africa, Ensminger (1997) convincingly argues that
conversions resulted from the economic advantages that Islam offered to traders
(see also Triulzi 1975). She claims that “Islam was a powerful ideology with
built-in sanctions which contributed to considerable self-enforcement of contracts.
True believers had a non-material interest in holding to the terms of contracts
even if the opportunity presented itself to shirk” (Ensminger 1997: 8). Thus Islam
provided a mechanism to overcome the collective action problems of longdistance
commerce. According to Ensminger, conversion to Islam increased trust
among traders, which reduced transaction costs, making trade more profitable. In
addition, high levels of trust among Muslim coreligionists allowed for greater
credit to be extended, thereby facilitating further trade expansion. Similar to the
9
According to the definitions offered above, this trust would be profitable, repeated, and
asymmetric.
14 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
case for Jewish diamond merchants and Maghribi traders, trust for Muslim traders
is characterized as profitable, repeated, and either mutual or asymmetric.
Ensminger contends that the steep initiation costs of entry into Islam, such as
daily prayer, abstaining from alcohol, fasting during Ramadan, and the pilgrimage
to Mecca, served as the means for establishing a reputation among merchants for
trustworthiness. In other words, these rituals and taboos are costly signals of
commitment that served to prevent free riders from achieving the benefits of more
efficient commerce.
However, Muslim traders did not rely exclusively on “non-material interests”
to secure cooperative gains. As Ensminger (1997: 9) notes, “one of the most
important institutions spread through the trading diasporas was the Muslim
commenda.” The commenda was nearly identical to the coalition used by
Maghribi traders discussed above. Perinbam (1980, cited in Ensminger 1997: 10)
describes the commenda:
It enabled an investor or group of investors to entrust capital or merchandise to
an agent-manager, who traded with it and returned the principal and previously
agreed-upon share of profits to the investor. The agent received the remaining
share of profits as a reward for his time and labor. … The commenda, which
combined the advantages of loan with those of partnership, was a contract which
regulated the use of capital, trading skills, and labor for mutual profit.
As with Maghribi traders, trust was not critical in agent-trader interactions
because agents could maximize their gains through honesty. The necessity of an
honest reputation to remain in the commenda altered the payoffs of the trust
dilemma such that it was not in the agents’ self-interest to defect on the
arrangement.
Why Trade with Coreligionists?
Many trade networks, all with their specific collective action problems to solve,
are dominated by religious communities (e.g., Iyer 1999; Udovitch 1970).
However, if Jewish diamond merchants and Maghribi and Muslim traders
exemplify conditions of other religious trading communities (i.e., they solve the
trust dilemma by modifying the payoff structure), why have these business
networks been organized around religious identities? Indeed, if the payoffs are
structured so that reputations are valued and honesty is in everyone’s interest,
why does it matter whether your trading partner shares your religious beliefs and
practices?
Three factors appear to be important. First, social relations within religious
communities are circumscribed. Ritual obligations and symbolic markers
demarcate membership within religious communities, thus clearly defining the
Sosis: Does Religion Promote Trust? 15
social arena and with whom social obligations lie. By establishing an
unambiguous social environment, religious groups are able to offer a history of
social interactions on which reputations can be built and evaluated. In other
words, the religious identities of Jewish diamond merchants, Maghribi traders,
and Muslim traders do not directly signal honesty; what keeping kashrut or hallal
(Jewish and Muslim dietary laws respectively), pilgrimages, praying three or five
times daily, wearing a yarmulke and other religious obligations signal is the
ability to obtain reliable reputational information because past social relations will
be common knowledge within these respective communities. It is easier to track
individuals and obtain information about their past interactions when community
membership is well-defined; religious behavior and symbolic markers visibly
delineate community boundaries. The second reason why successful collective
action so often involves members who share a religious identity is that
punishment can be extremely effective when it not only affects economic
associations, but affects social relations as well. Similar to the Jewish diamond
merchants who suffer social costs if they cheat on a transaction, Maghribi traders
would impose social sanctions (in addition to trade sanctions) on defecting agents
by appealing to the entire Jewish community (Greif 1989). Third, supernatural
sanctions (e.g., hell) and rewards (e.g., heaven) can alter the perceived payoffs of
an interaction so that the trust dilemma is avoided (see Bulbulia 2004b; Johnson
and Kruger 2004; Sosis 2003). Interestingly, supernatural sanctions and rewards
are not relied on exclusively to promote cooperation; they are used to support and
strengthen material punishment and reputational mechanisms. To clarify, I am not
positing that either of these systems necessarily has priority over the other. The
point is that one reason religious groups successfully overcome the challenges of
collective action (e.g., Sosis and Bressler 2003) is because their material and
supernatural punishment and reward mechanisms reinforce and complement each
other.
DOES TRUST EVER EXIST?
While the examples discussed above have been limited to economic activities,
many religious communities employ punishment and reputational mechanisms to
circumvent trust dilemmas by altering the payoffs for a wide range of withingroup
social interactions. The 19th century Perfectionists of John Humphrey
Noyes’s Oneida community, for example, formalized the reputational mechanism
through the practice of mutual criticism, in which members stood in front of the
community and endured humiliating public critiques of their character, typically
highlighting spiritual and social shortcomings (Oved 1988). Of course, many
religious groups, such as the Amish and the Church of Latter-day Saints, hold the
threat of excommunication over their constituents, a threat that they can reliably
16 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
enforce (Hostetler 1980; Shepherd and Shepherd 1998). Social sanctions,
however, need not be formalized. Although Jehovah’s Witnesses will
excommunicate those who blatantly deviate from the Society’s teachings, a
process referred to as disfellowship, informally the costs of not fully accepting the
belief system is a “loss of interaction and the cutting off of affective bonds”
(Zellner 2001:358), which effectively results in the social excommunication of the
nonbeliever.
In closed religious communities, evidently, recurrent cooperative pursuits
often do not meet the conditions for trust defined by Bacharach and Gambetta
(2001). When social relations occur within the framework of a bounded
community so that interactions are repeated and mobility between social groups is
limited, reciprocity can emerge and stabilize. Consequently, reputations and
punishment threats become relevant and influence the payoff structure of the
interactions, often making trust unnecessary. Nonetheless, trust does indeed exist
within religious populations, and I now turn to the conditions under which
religious signaling is an important factor in promoting trust.
Monitoring Costs
Although repeated interactions, reputational effects, and the threat of ostracism all
facilitate cooperative interactions, religious rituals and markers may promote trust
within religious communities by reducing monitoring costs of collective pursuits.
For example, in fieldwork that I conducted among Ultra-Orthodox Jews in
northern Israel, I found monitoring costs of cooperative activities to be especially
low. I regularly attended community religious functions at which bags and
valuables lay unguarded in a corner. Following these events, I occasionally visited
the local university library, where, in sharp contrast to the gathering I had just left,
I was required to lock my belongings in a locker, and observation cameras
monitored the book stacks and exits.
Religious groups may achieve low monitoring costs in collective pursuits, but
they pay additional costs to monitor the performance of religious obligations.
However, since many religious behaviors are formal, conspicuous, and
communally practiced—indeed, these characteristics contribute to their
effectiveness as signals (Alcorta and Sosis 2005; Sosis 2003) —they are cheaper
to monitor than other activities. Thus, religious groups lower their net monitoring
costs by transferring the costs from monitoring collective inputs to monitoring
conspicuous religious activities.
Monitoring costs are directly reduced by three additional means, all of which
increase religious group members’ opportunity costs, making defection and
untrustworthy decisions less likely. First, the high investment in adolescent
religious training rather than secular education, such as among Hutterites
Sosis: Does Religion Promote Trust? 17
(Hostetler 1997) and Ultra-Orthodox Jews (Heilman 1992), may make entry into
other communities difficult if individuals lack skills to compete economically
(Sosis 2003). Second, permanent stigmatizing markers such as tattoos, piercings,
or more severe bodily mutilations may limit one’s ability to join an outside group.
Third, Sosis (2003) has argued that through processes of cognitive dissonance
(Festinger 1957) or social perception (Bem 1965, 1972), repeated ritual
performance fosters belief in the doctrines of the group; it facilitates internalizing
group norms. These beliefs typically include a nonmaterial system of reward and
punishment that consequently reduces the costs of monitoring members’ behavior
(Bulbulia 2004b; Johnson 2005; Johnson and Kruger 2004). Alcorta and Sosis
(2005) further note that humans use ritual to conditionally associate emotions and
culturally defined sacred symbols; they propose that such symbols represent
powerful tools for promoting trust and cooperation. Since emotions are generated
from limbic structures that are out of conscious control, they are difficult to fake
(Ekman, Levenson, and Friesen 1983, Levenson 2003) and can consequently serve
as reliable signals of trustworthiness and commitment, thereby reducing
monitoring costs (Alcorta and Sosis 2005; Bulbulia 2004b).
While not directly concerned with supernatural reward and punishment,
Coleman (1990) makes two germane points for understanding the conditions
under which a society will aim to internalize its norms. First, Coleman (1990:
296) notes that norm internalization is efficient when there is a range of actions
that the society seeks to control. This aptly characterizes religious communities,
which generally seek members who behave prosocially toward coreligionists
under diverse conditions; in other words, religious communities wish to
encourage cooperation and trust between members regardless of the situation that
arises. Second, Coleman argues that external policing becomes less efficient when
observations are more costly to observe, and therefore under these conditions,
societies are more likely to rely on internalization strategies. Since the intragroup
trust and cooperation that are encouraged within religious communities are not
limited in time (e.g., just during work hours) and place (e.g., just in a house of
worship) but are a continuous obligation, it is impossible to monitor members’
commitment to this ethic all of the time. Johnson and Bering (n.d.) additionally
note that negative sanctions are more effective than positive reinforcement in
eliciting a desired behavioral response. Indeed, reviewing the results of various
economic experiments that employed reward and punishment incentives,
Andreoni, Harbaugh, and Vesterlund (2003: 894) conclude that “one might expect
less cooperation in societies where good behavior is rewarded than in those where
poor behavior is punished.” Hence, supernatural punishment is an efficient
complement to the material punishment system that deters defection from
cooperative obligations among community members, resulting in lower
monitoring costs.
18 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
Returning to Bacharach and Gambetta’s (2001) definition, low investment in
secular training, stigmatization, and norm internalization make it more likely that
a Truster will choose A, the trusting decision, since she recognizes that Trustees
are more likely to choose B because the potential gains for defection (B') are
reduced, or at least perceived to be reduced, in the case of supernatural sanctions.
Intergroup Mobility
Exponential population growth, greater demographic mobility, and the trend
toward increased urbanization in the modern era have influenced the scope of
trust in contemporary societies. The effects of these factors on trust might seem
counterintuitive; where social groups are open and fluid, there will be greater
dependence on trust in collective pursuits than there is in segregated or isolated
closed communities. In environments in which there is high mobility between
groups, reputations and punishments are less effective as means of securing
cooperation (Hardin 2000). Since individuals can relocate and align themselves
with a different community when faced with the prospect of social sanctions,
punishments are less effective in generating cooperation. Likewise, mobility
enables individuals to escape and leave behind damaged reputations, making
reputations in open groups a less reliable source of information than in closed
communities. Ironically, trusting behavior may be most vital in environments in
which social groups are fluid and individuals maintain multiple group identities.
Under these conditions, cooperation is less frequent and more challenging to
sustain than in closed groups (e.g., Sosis and Bressler 2003); however, when
collective goals are pursued, trusting behavior is essential. This is not to claim
that in open communities with high intergroup mobility, reputations are not
valued and punishments are not employed with some efficacy. They are. The
point is that these mechanisms for motivating collective action will be less
effective in open communities than in closed ones. Concomitantly, in open
communities, the payoffs for cooperative interactions are more likely to mirror the
conditions for trust established above; hence cooperation is more difficult to
achieve. In other words, open communities rely on trust to achieve their collective
goals to a greater degree than do closed communities, which in contrast are more
dependent on institutional structures that enable punishment and reputational
mechanisms to foster cooperation.10
Religion plays another significant role in environments with high intergroup
mobility. Individuals who maintain multiple group identities can face conflicts of
interest when diverse group ideals do not coincide. In these circumstances,
10
To clarify, I am not arguing that individuals’ general perceptions or attitudes of trust will be
higher in open communities than in closed communities. Indeed, Alesina and La Ferrara (2002)
find that interpersonal trust is somewhat lower in communities where members are more transient.
Sosis: Does Religion Promote Trust? 19
religion often serves to prioritize values. Theologians (Hefner 1993; Tillich 1951,
1952) as well as social scientists (Klass 1995; Rappaport 1999) have noted that
one of religion’s defining characteristics is that it is the ultimate concern of its
followers. As such, religions not only offer moral guidelines to settle individual
conflicts of interest, but also establish a hierarchy of values that can be
incorporated into other identities. This integration will be facilitated by the fact
that others equally recognize the priority of religious values and convictions.
Nonrepeated Trust
Costly religious rituals and markers not only lower monitoring costs of
community members and encourage repeated trust in environments of high
mobility; they also serve to enhance trust among fellow adherents in the absence
of long-term reciprocal relations. These interactions were characterized above as
nonrepeated trust and include the payoffs that I defined as nonprofitable, as well
as profitable trust. In large-scale societies, generalized norms of cooperation are
susceptible to free riders because the ease of relocating to another community
makes punishment and reputational benefits ineffective means of social control
(Hardin 2000). In addition, higher population densities result in more frequent
interactions between individuals who do not have long-term reciprocal relations.
Costly rituals and stigmatizing markers enable mutually beneficial relationships
between strangers or new acquaintances without the costly investment of longterm
relationships. These religious obligations are likely to be most valuable in
communities with adherents in separate locations (whether geographically distant
or near) who interact occasionally but not frequently enough to establish
cooperative networks without the support of additional commitment signals.
Trade, marital exchanges, and financial aid to support struggling and nascent
communities are all realized benefits of these intercommunity relations between
dispersed communities that share a common faith (e.g., Ensminger 1997;
Hostetler 1997; Udovitch and Valensi 1984).
While religion undeniably facilitates collective action among community
members, its greater influence may be the proliferation of nonrepeated trust
between members of the same religion who are strangers (see Macy and Skvoretz
1998), and it likely offers the uncommon conditions under which nonprofitable
trust can occur. In contrast to intragroup or intergroup trust, we can refer to this
type of trust as extended-group trust: trust that occurs among anonymous (or
nearly anonymous) members of the same religion who reside in different
communities. Computer simulations conducted by Riolo, Cohen, and Axelrod
(2001) show that cooperation can emerge and stabilize between agents that share
identifying markers and have no prior history of interaction and no expectations
of future encounters. Costly religious markers and rituals enable adherents who
20 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
share a common faith to assess mutual trustworthiness without investing in longterm
reciprocal relations. For example, Triulzi (1975: 59) describes how 19th
century migrant Muslim traders from the Sudan traveled to their coreligionists in
Ethiopia, where “[t]he stranger was received as a brother.” Similarly, during
fieldwork among Ultra-Orthodox Jewish communities in Israel, I repeatedly
observed invitations for meals, lodging, and rides by residents to unknown UltraOrthodox
travelers. On several occasions, I witnessed cars being loaned to
complete strangers, and interviews revealed a surprising number of interest-free
loans offered and accepted between people who had previously not known each
other.
It is puzzling why such markers are perceived as honest signals of
trustworthiness. To imitate nonpermanent markers with intentions of deception is
generally not too difficult. Palestinian suicide bombers, for instance, have often
dressed up as Ultra-Orthodox Jews, escaping detection and successfully fulfilling
their missions. With less devastating effects, reporters, anthropologists (including
myself), and Peace Corps workers, among others, have regularly donned the garb
of those they are working with to gain their trust. Symbolic markers have long
been recognized as critical for establishing ethnic and cultural boundaries,
allowing individuals to selectively direct altruistic benefits toward in-group
members (Nettle and Dunbar 1997). Barth (1969) accurately notes, “The
identification of another person as a fellow member of an ethnic group implies a
sharing of criteria for evaluation and judgment” (cited in Carr and Landa
1983:150). We appear to have psychological biases that favorably predispose us
toward others with whom we share symbolic identity markers (e.g., Fershtman
and Gneezy 2001). In our evolutionary history, such markers were likely to be
reliable indicators of in-group identity, as they generally, but not always, are
today. Despite these possible biases, when reputational information is available, it
is rarely ignored; it is evaluated along with symbolic messages, often taking
priority. For instance, in her study of Ultra-Orthodox diamond merchants, Shield
(2002: 110) discerns that “religious piety does not substitute for ethical business
practice; in fact, expressions of piety without ethical behavior garner considerable
contempt. Traders are too savvy to let superficial emblems of religiosity blind
them to dubious marketplace behaviors.” Jains from India offer another
interesting example. Iyer (1999: 107) observes that among Jains, “business
reputation is derived not only from economic activity but also from religious
activities of family members.” However, religious practices appear to play a
signaling role that is secondary to economic status. In Jainism, wealth is
considered a reward for one’s piety, whereas involuntary poverty is a result of
moral failings. Consequently, “wealth emerges as a signifier not only of moral
piety but also as the signal for economic reputation” (Iyer 1999: 105). Wealth
begets the advantages of extended credit, which of course begets more wealth.
Sosis: Does Religion Promote Trust? 21
Signaling Trustworthiness to Non-Group Members
Few studies have examined how signals that may have emerged for one purpose
can be coopted for other purposes and still maintain stability (e.g., Johnstone and
Norris 1993). With regard to religious symbolic markers, it appears that they not
only signal in-group commitment and trustworthiness; outsiders also use religious
behaviors and markers to evaluate trustworthiness. For example, Frank (1988)
observes that affluent New York City families place advertisements in the
newspapers of Salt Lake City for Mormon governesses for their children.
Apparently, “persons raised in the Mormon tradition are trustworthy to a degree
that the average New Yorker is not” (Frank 1988: 111).
Paxson (2004) has similarly shown that Sikhs achieve economic benefits from
perceived trustworthiness. Adult male Sikhs can be easily distinguished by
various external markers, often referred to as the five K’s (Kes, Kangha, Kara,
Kirpan, and Kachera): unshorn hair and beard and wearing a comb, steel bracelet,
saber, and breeches. Additional constraints on their behavior, such as refraining
from alcohol and tobacco and the requirement to pray five times daily, further
mark their distinctiveness. Paxson (2004) argues that Sikhs are recognized by
non-Sikhs as trustworthy trading partners, even without a history of prior
exchanges. Non-Sikhs can utilize Sikh symbolic markers as a “seal of approval”
signaling trustworthiness. The external displays indicate that the individual has
already endured the monitoring systems within Sikh communities that allow him
to maintain his membership.
CONCLUSIONS
Religious communities clearly achieve high levels of cooperation and collective
action. Trust, however, is often unnecessary to mediate these cooperative
activities. Closed religious communities tend to support effective punishment
systems, be well defined, and have members who repeatedly interact with each
other, enabling the establishment of reliable reputations. These religious
communities additionally facilitate collective action by offering a credible history
of social relations and extending their punishment measures outside of the
economic sphere and into the social arena. In religious communities where
identities are shared between multiple groups and individuals can transfer groups,
reputations and punishments will be less successful at marshaling cooperation,
and trust will be indispensable for realizing cooperative goals. Religious markers
and rituals also promote successful trust between anonymous interactants. Not
only do religions achieve this among extended-group members, that is, those who
22 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
share a faith but live in disparate communities, but religious rituals and markers
can also signal trustworthiness to non-group members.
Nonetheless, religious adherents are unlikely to extend trust indiscriminantly.
Indeed, various researchers argue that in social systems with close-knit families or
small groups, the strong trusting relations that occur within the family or group
are at the expense of trust with individuals outside of these relations (Fukuyama
1995; Gellner 1988; Ruffle and Sosis 2005; Yamagishi 1996). Experimental
evidence appears to support this assertion (Yamagishi and Yamagishi 1994;
Yamagishi, Cook, and Watabe 1998). Among religious communities that engage in
high levels of costly signaling and encourage strong group ties (even in highpopulation-density
environments), we would similarly expect members to be less
trusting of outsiders than are individuals living in large urban settings
characterized by networks of reciprocation (see Cook and Hardin 2001).
There may be conditions, however, under which indiscriminant, or less
discriminating, trust is a successful strategy. For example, following Weber,
Runciman (2005) argues that the success of the Puritan ethic was that
indiscriminant fairness and cooperation were profitable because in small
communities, there were significant public pressures and enforceable punishments
that maintained this norm of behavior. Simulations conducted by Macy and
Skvoretz (1988) indicate that trust can spread across a large population if trust is
first stabilized in embedded social interactions.
The discussion presented here raises a number of questions and suggests
several avenues for further inquiry. Within the anthropological signaling literature
(e.g., Irons 2001; Sosis 2003), it has been assumed that religious behaviors
directly signal group commitments. However, the case studies examined here
suggest that religious signals have multiple effects and that their messages and
significata vary environmentally in fundamental ways. For example, in addition to
signaling group commitment, within closed groups, religious signals can also
indicate membership in a community where reputational information is valued
and available. The prevalence of punishments and reputations in many of these
communities to which high levels of trust have often been ascribed attests to the
importance of these mechanisms in facilitating cooperative goals. When given the
opportunity to behave in a trusting manner according to the payoffs defined
above, individuals within these communities will undoubtedly often rise to the
challenge. What is significant is that many religious groups do not rely on the
goodwill that their communities cultivate but rather establish institutions that
strengthen the efficacy of social penalties and reputational status. As Sigmund,
Hauert, and Nowak (2001: 10757) conclude from their analysis of minigame
simulations of public goods experiments, “reputation is essential for fostering
social behavior among selfish agents, and … it is considerably more effective
with punishment than with reward.” Indeed, recent experimental work has found
Sosis: Does Religion Promote Trust? 23
reputations (King-Casas et al. 2005) and punishment (Fehr and Gachter 2002) to
be powerful predictors of trust and cooperation. Future work should further
explore the efficacy of these mechanisms within religious communities and the
relative costs of their maintenance.
Future research also needs to examine trust relations in open religious
communities. I have asserted here that collective action in these communities will
require trust owing to weak punishment and reputational mechanisms, yet
empirical data are lacking. Observational as well as experimental data are
imperative for evaluating this hypothesis. Moreover, future work must examine
trust within religious populations that constitute the majority in their respective
environments. When minority coreligionists from disparate communities interact,
we expect markers and ritual signals to facilitate social interactions. Are such
signals as effective when coreligionists constitute the majority of the religious
landscape? In the United States, for example, is wearing a Star of David more
effective at fostering trust among Jews than wearing a cross is at fostering trust
among Christians? While experimental data could evaluate this, it seems likely
that when coreligionists constitute a majority, they will require additional identity
information to narrow the individual’s affiliation to a specific subpopulation
within the majority culture. In contrast, minority religious populations are
probably less concerned with denominational affiliations when evaluating trust
relationships with unfamiliar extended-group members.
I opened this article by noting that current world affairs highlight the
importance of understanding the relationship between religion and trust. I close
with the same message. As more destructive means of warfare continue to
develop and the dangers posed by international conflicts increase, understanding
the factors that promote and sustain trust may be more relevant now than at any
time in history. Examining religion should be central to this endeavor; even when
conflicts are not directly motivated by religious concerns, religion is often used as
a catalyst to unify groups for military purposes (MacNeill 2004; Sosis 2004;
Sosis, Kress, and Boster 2004). In addition to its relevance in developing conflict
resolution strategies, religion’s ability to enhance trust may also be invaluable for
promoting economic growth (Barro and McCleary 2003) and economic exchange
in underdeveloped economies in which property rights and other law
enforcement, judicial, and economic institutions are weak (see Ruffle and Sosis
n.d.). Over a decade ago, Coleman (1990: 180) suggested “that combined
empirical and theoretical work on the dynamics of trust in social relationships
should be productive.” I of course concur and hope that as this work continues to
develop, future research will also explore the dynamics of trust within religious
communities.
24 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
REFERENCES
Alcorta, C., and R. Sosis 2005 “Ritual, Emotion, and Sacred Symbols: The Evolution of
Religion as an Adaptive Complex.” Human Nature 16 (in press).
Alesina, A., and E. La Ferrara. 2002. “Who Trusts Others?” Journal of Public Economics
85: 207–234.
Amato, P. 1993. “Urban-Rural Differences in Helping Friends and Family Members.”
Social Psychology Quarterly 56: 249–262.
Andreoni, J., W. Harbaugh, and L. Vesterlund. 2003. “The Carrot or the Stick: Rewards,
Punishments, and Cooperation.” American Economic Review 93: 893–902.
Atran, S. 2002. In Gods We Trust: The Evolutionary Landscape of Religion. Oxford, UK:
Oxford University Press.
Bacharach, M., and D. Gambetta. 2001. “Trust in Signs.” In Trust and Society, edited by
K. Cook, 149–184. New York: Russell Sage Foundation.
Barclay, P. 2004. “Trustworthiness and Competitive Altruism Can Also Solve the
‘Tragedy of the Commons.’” Evolution and Human Behavior 25: 209–220.
Barro, R. J., and R. M. McCleary 2003. “Religion and Economic Growth Across
Countries.” American Sociological Review 68: 760–781.
Barth, F. 1969. Ethnic Groups and Boundaries. Boston: Little, Brown.
Bem, D. 1965. “An Experimental Analysis of Self-Persuasion.” Journal of Experimental
Social Psychology 1: 199–218.
Bem, D. 1972. “Self-Perception Theory.” In Advances in Experimental Social
Psychology, Volume 6, edited by L. Berkowitz, 1–62. New York: Academic Press.
Berg, J., J. Dickhaut, and K. McCabe. 1995. “Trust, Reciprocity and Social History.”
Games and Economic Behavior 10: 122–142.
Berman, E. 2000. “Sect, Subsidy and Sacrifice: An Economist’s View of Ultra-Orthodox
Jews.” Quarterly Journal of Economics 115: 905–953.
Bernstein, L. 1992. “Opting Out of the Legal System: Extralegal Contractual Relations in
the Diamond Industry.” Journal of Legal Studies 21: 115–157.
Bliege Bird, R., and E. Smith. 2005. “Signaling Theory, Strategic Interaction, and
Symbolic Capital.” Current Anthropology 46: 221–248.
Bohnet, I., and R. Croson, eds. 2004. “Trust and Trustworthiness.” Journal of Economic
Behavior and Organization 55: 443–597.
Boudon, R. 1987. “The Individualistic Tradition in Sociology.” In The Micro-Macro
Link, edited by J. Alexander, B. Geisen, R. Munch, and N. Smelser, 45–70. Berkeley:
University of California Press.
Bulbulia, J. 2004a. “Area Review: The Cognitive and Evolutionary Psychology of
Religion.” Biology and Philosophy. 18: 655–686.
Bulbulia, J. 2004b. “Religious Costs as Adaptations That Signal Altruistic Intention.”
Evolution and Cognition 10: 19–38.
Carr, J., and J. Landa. 1983. “The Economics of Symbols, Clan Names and Religion.”
Journal of Legal Studies. 13: 135–156.
Coleman, J. 1988. “Social Capital in Creation of Human Capital.” The American Journal
of Sociology 94: S95–S120.
Sosis: Does Religion Promote Trust? 25
Coleman, J. 1990. Foundations of Social Theory. Cambridge, MA: Harvard University
Press.
Collins, R. 2004 Interaction Ritual Chains. Princeton, NJ: Princeton University Press.
Cook, K., and R. Hardin. 2001. “Norms of Cooperativeness and Networks of Trust.” In
Social Norms, edited by M. Hechter and K. D. Opp, 327–347. New York: Russell
Sage Foundation.
Cronk, L. 1994. “Evolutionary Theories of Morality and the Manipulative Use of
Signals.” Zygon 29: 81–101.
Croson, R., and N. Buchan. 1999 “Gender and Culture: International Experimental
Evidence from Trust Games.” American Economic Review Papers and Proceedings.
89: 386–391.
Dasgupta, P. 1988. “Trust as a Commodity.” In Trust: Making and Breaking Cooperative
Relations, edited by D. Gambetta, 49–72. Oxford, UK: Basil Blackwell.
Dawes, R. 1980. “Social Dilemmas.” Annual Review of Psychology 31: 169–193.
Douglas M. 1966. Purity and Danger. New York: Frederick A. Praeger.
Durkheim, E. 1995 [1912]. The Elementary Forms of Religious Life. New York: Free
Press.
Ekman, P., R. Levenson, and W. Friesen. 1983. “Autonomic Nervous System Activity
Distinguishes Among Emotions.” Science 22: 1208–1210.
Ensminger, J. 1997. “Transaction Costs and Islam: Explaining Conversion in Africa.”
Journalof Institutional and Theoretical Economics 153: 4–29.
Fehr, E., and S. Gachter. 2002. “Altruistic Punishment in Humans.” Nature 415: 137–
140.
Fershtman, C., and U. Gneezy. 2001. “Discrimination in a Segmented Society: An
Experimental Approach.” Quarterly Journal of Economics 116: 351–377.
Festinger, L. 1957. A Theory of Cognitive Dissonance. Stanford, CA: Stanford University
Press.
Frank, R. H. 1988. Passions Within Reason: The Strategic Role of the Emotions. New
York: W.W. Norton.
Frank, R. 2001. “Cooperation Through Emotional Commitment.” In Evolution and the
Capacity for Commitment, edited by R. Nesse, 292–309. New York: Russell Sage
Foundation.
Fukuyama, F. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York:
Free Press.
Gellner, E. 1988. “Trust, Cohesion, and the Social Order.” In Trust: Making and
Breaking Cooperative Relations, edited by D. Gambetta, 142–157. Oxford, UK:
Blackwell.
Glaeser E., D. Laibson, J. Scheinkman, and C. Soutter. 2000. “Measuring Trust.”
Quarterly Journal of Economics 115: 811–846.
Greif, A. 1989. “Reputation and Coalitions in Medieval Trade: Evidence on the Maghribi
Traders.” Journal of Economic History 49: 857–882.
Greif, A. 1991. “The Organization of Long-Distance Trade: Reputation and Coalitions in
the Geniza Documents and Genoa During the Eleventh and Twelfth Centuries.”
Journal of Economic History 51: 459–462.
26 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
Greif, A. 1993. “Contract Enforceability and Economic Institutions in Early Trade: The
Maghribi Traders.” American Economic Review 83:525–548.
Gutwirth, J. 1968. “Antwerp Jewry Today.” Jewish Journal of Sociology 10: 121–137.
Hardin, R. 1982. Collective Action. Baltimore, MD: Johns Hopkins University Press.
Hardin, R. 2000. “Trust and Society.” In Competition and Structure, edited by G.
Galeotti, P. Salmon, and R. Wintrobe, 17–46. Cambridge, UK: Cambridge University
Press.
Heckathorn, D. 1989. “Collective Action and the Second-Order Free-Rider Problem.”
Rationality and Society 1: 78–100.
Hefner, P. 1993. The Human Factor: Evolution, Culture, and Religion. Minneapolis:
Fortress Press.
Heilman, S. 1992. Defenders of the Faith. Los Angeles: University of California Press.
Hostetler, J. 1980. Amish Society. Baltimore, MD: Johns Hopkins University Press.
Hostetler, J. 1997. Hutterite Society. Baltimore, MD: Johns Hopkins University Press.
Iannaccone, L. 1992. “Sacrifice and Stigma: Reducing Free-Riding in Cults, Communes,
and Other Collectives.” Journal of Political Economy 100: 271–291.
Iannaccone, L. 1994. “Why Strict Churches Are Strong.” American Journal of Sociology
99:1180–1211.
Irons, W. 1996a. “In Our Own Self-Image: The Evolution of Morality, Deception, and
Religion.” Skeptic 4: 50–61.
Irons, W. 1996b. “Morality as an Evolved Adaptation.” In Investigating the Biological
Foundations of Morality, edited by J. P. Hurd, 1–34. Lewiston, NY: Edwin Mellen
Press.
Irons, W. 1996c. “Morality, Religion, and Human Nature.” In Religion and Science:
History, Method, and Dialogue, edited by W. Richardson and W. Wildman, 375–399.
New York: Routledge.
Irons, W. 2001. “Religion as a Hard-to-Fake Sign of Commitment.” In Evolution and the
Capacity for Commitment, edited by R. Nesse, 292–309. New York: Russell Sage
Foundation.
Irons, W. 2004. “An Evolutionary Critique of the Created Co-Creator Concept.” Zygon:
Journal of Religion and Science 39: 773–790.
Iyer, G. 1999. “The Impact of Religion and Reputation on the Organization of Indian
Merchant Communities.” The Journal of Business and Industrial Marketing 14: 102–
121.
Johnson, D. 2005. “God’s Punishment and Public Goods: A Test of the Supernatural
Punishment Hypothesis in 186 World Cultures.” Human Nature 16 (in press).
Johnson, D., and J. Bering. n.d. “Supernatural Agency and the Evolution of
Cooperation.” www.princeton.edu
Johnson, D., and O. Kruger. 2004. “The Good of Wrath: Supernatural Punishment and
the Evolution of Cooperation.” Political Theology 5: 159–176.
Johnson, D., P. Stopka, and S. Knights 2003. “The Puzzle of Human Cooperation.”
Nature 421: 911–912.
Johnstone, R., and K. Norris 1993. “Badges of Status and the Cost of Aggression.”
Behavioral Ecology and Sociobiology 32: 127–134.
Sosis: Does Religion Promote Trust? 27
King-Casas, B., D. Tomlin, C. Anen, C. Camerer, S. Quartz, and P. R. Montague 2005.
“Getting to Know You: Reputation and Trust in a Two-Person Economic Exchange.”
Science 308: 78–83.
Klass, M. 1995. Ordered Universes: Approaches to the Anthropology of Religion.
Boulder, CO: Westview Press.
Landa, J. 1994. Trust, Ethnicity, and Identity: The New Institutional Economics of Ethnic
Trading Networks, Contract Law, and Gift-Exchange. Ann Arbor: University of
Michigan Press.
Levin, J. 1994. “Religion and Health: Is There an Association, Is It Valid, Is It Causal?”
Social Science Medicine 38: 1475–1482.
Luhmann, N. 1997. “Globalization or World Society: How to Conceive of Modern
Society?” International Review of Sociology 7: 67–79.
MacNeill, A. 2004. “The Capacity for Religious Experience Is an Evolutionary
Adaptation to Warfare.” Evolution and Cognition 10: 43–60.
Macy, M., and J. Skvoretz. 1998. “The Evolution of Trust and Cooperation Between
Strangers: A Computational Model.” American Sociological Review 63: 638–660.
McCabe, K., M. Rigdon, and V. Smith. 2003. “Positive Reciprocity and Intentions in
Trust Games.” Journal of Economic Behavior and Organization 52: 267–275.
McCullough, M., W. Hoyt, D. Larson, H. Koenig, and C. Thoresen. 2000. “Religious
Involvement and Mortality: A Meta-Analytic Review.” Health Psychology 19: 211–
222.
McKelvey, R., and T. Palfrey. 1992. “An Experimental Study of the Centipede Game.”
Econometrica 4: 803–836.
Nesse, R. 1999 “The Evolution of Commitment and the Origins of Religion.” Science
and Spirit 10: 32–33.
Nettle, D., and R. Dunbar. 1997. “Social Markers and the Evolution of Reciprocal
Exchange.” Current Anthropology 38: 93–99.
Olson, M. 1965. The Logic of Collective Action: Public Goods and the Theory of Groups.
Cambridge, MA: Harvard University Press.
Orbell, J., M. Goldman, M. Mulford, and R. Dawes 1992. “Religion, Context, and
Constraint Toward Strangers.” Rationality and Society 4: 291–307.
Oved, Y. 1988. Two Hundred Years of American Communes. New Brunswick, NJ:
Transaction Books.
Paxson, N. 2004. “The Entrepreneurial Ethic of Sikhs.” Paper presented at meetings of
the Association for the Study of Religion, Economics, and Culture, October 22–24,
2004, Kansas City, Missouri.
Perinbam, B. 1980. “The Julas in Western Sudanese History: Long-Distance Traders and
Developers of Resources.” In West African Cultural Dynamics: Archaeological and
Historical Perspectives, edited by B. K. Swartz, Jr., and R. E. Dumett, 455–476. The
Hague: Mouton.
Putnam, R. 2000. Bowling Alone: The Collapse and Revival of American Community.
New York: Simon and Schuster.
Radcliffe-Brown, A. 1952. Structure and Function in Primitive Society. London: Cohen
and West.
28 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
Rappaport, R. 1999. Ritual and Religion in the Making of Humanity. Cambridge, UK:
Cambridge University Press.
Reynolds V., and Tanner R. 1995. The Social Ecology of Religion. New York: Oxford
University Press.
Richman, B. 2004. “Community Enforcement of Informal Contracts: Jewish Diamond
Merchants in New York.” Paper presented at meetings of the Association for the
Study of Religion, Economics, and Culture, October 22–24, 2004, Kansas City,
Missouri.
Riolo, R., M. Cohen, and R. Axelrod. 2001. “Evolution of Cooperation Without
Reciprocity.” Nature 414: 441–443.
Rosenberg, N., and L. Birdzell. 1986. How the West Grew Rich. New York: Basic Books.
Ruffle, B., and R. Sosis. 2005. “Cooperation and the In-Group-Out-Group Bias: A Field
Test on Israeli Kibbutzim and City Residents.” Journal of Economic Behavior and
Organization 58 (in press).
Ruffle, B., and R. Sosis. n.d. “Does It Pay to Pray? Evaluating the Economic Return to
Religious Ritual.”
Runciman, W. G. 2005. “Puritan American Capitalism and Evolutionary Game Theory.”
Max Weber Studies 5 (in press).
Schelling, T. 1960. The Strategy of Conflict. Cambridge, MA: Harvard University Press.
Schelling, T. 2001. “Commitment: Deliberate Versus Involuntary.” In Evolution and the
Capacity for Commitment, edited by R. Nesse, 292–309. New York: Russell Sage
Foundation.
Shepherd, G., and G. Shepherd. 1998. Mormon Passage: A Missionary Chronicle.
Chicago: University of Illinois Press.
Shield, R. 2002. Diamond Stories. Ithaca, NY: Cornell University Press.
Sigmund, K., C. Hauert, and M. Nowak. 2001. “Reward and Punishment.” Proceedings
of the National Academy of Sciences 98: 10757–10762.
Sombart, W. 1953. “Medieval and Modern Commercial Enterprise.” In Enterprise and
Secular Change, edited by F. Lane and J. Riemersma, 25–40. Homewood, IL: Irwin.
Sosis, R. 2000. “Religion and Intra-Group Cooperation: Preliminary Results of a
Comparative Analysis of Utopian Communities.” Cross-Cultural Research 34: 70–
87.
Sosis, R. 2003. “Why Aren’t We All Hutterites?:Costly Signaling Theory and Religion.”
Human Nature 14: 91–127.
Sosis, R. 2004. “The Adaptive Value of Religious Ritual.” American Scientist 92: 166–
172.
Sosis, R., and C. Alcorta. 2003. “Signaling, Solidarity and the Sacred: The Evolution of
Religious Behavior.” Evolutionary Anthropology 12: 264–274.
Sosis, R., and E. Bressler. 2003. “Cooperation and Commune Longevity: A Test of the
Costly Signaling Theory of Religion.” Cross-Cultural Research 37: 211–239.
Sosis, R. H. Kress, and J. Boster. 2004. “Scars for War: A Cross-Cultural Study of Male
Rites as Costly Signals of Commitment for Warfare.” Presentation for the British
Academy Centenary Project, The Social Brain, Chester, England, June 9–11.
Sosis: Does Religion Promote Trust? 29
Sosis, R, and B. Ruffle. 2003. “Religious Ritual and Cooperation: Testing for a
Relationship on Israeli Religious and Secular Kibbutzim.” Current Anthropology 44:
713–722.
Sosis, R., and B. Ruffle. 2004. “Ideology, Religion, and the Evolution of Cooperation:
Field Tests on Israeli Kibbutzim.” Research in Economic Anthropology 23: 89–117.
Spence, M. 1974. Market Signaling, Cambridge: Harvard University Press.
Spiro, M. 1966. “Religion: Problems of Definition and Explanation.” In Anthropological
Approaches to the Study of Religion, edited by M. Banton, 85–126. London:
Tavistock Publications.
Stark, R. 2001. One True God: Historical Consequences of Monotheism. Princeton, NJ:
Princeton University Press.
Stark, R., and W. Bainbridge. 1985. The Future of Religion: Secularization, Revival, and
Cult Formation. Berkeley: University of California Press.
Starr, R. 1984. “The Real Treasure on 47th Street.” Editorial in New York Times, March
26.
Steadman, L. B., and C. T. Palmer. 1995. “Religion as an Identifiable Traditional
Behavior Subject to Natural Selection.” Journal of Social and Evolutionary Systems
18: 149–164.
Tillich, P. 1951. Systematic Theology, Volume 1. Chicago: University of Chicago Press.
Tillich, P. 1952. The Courage to Be. New Haven, CT: Yale University.
Triulzi, A. 1975. “Trade, Islam, and the Mahdi in Northwestern Wallagga, Ethiopia.”
Journal of African History 16: 55–71.
Turner V. 1995 [1969]. The Ritual Process. Chicago, Aldine.
Tuzin, D. 1982. “Ritual Violence Among the Ilahita Arapesh.” In Rituals of Manhood:
Male Initiation in Papua New Guinea, edited by G. H. Herdt, 321–356. Berkeley:
University of California Press.
Udovitch, A. 1970. Partnership and Profit in Medieval Islam. Princeton, NJ: Princeton
University Press.
Udovitch, A., and L. Valensi. 1984. The Last Arab Jews: The Communities of Jerba
Tunisia. New York: Hardwood Academic Publishers.
Weber, M. 1961. General Economic History. New York: First Collier Books.
Weber, M. 1958. From Max Weber: Essays in Sociology, edited by H. H. Gerth and C.
W. Mills. New York: Oxford University Press
Wechsberg, J. 1966. The Merchant Bankers. Boston: Little, Brown.
Williamson, O. 1993. “Calculativeness, Trust, and Economic Organization.” Journal of
Law and Economics 36: 453–486.
Wilson, D. S. 2002 Darwin’s Cathedral. Chicago: University of Chicago Press.
Whiting, J., R. Kluckholn, and A. Anthony. 1958. “The Function of Male Initiation
Ceremonies at Puberty.” In Readings in Social Psychology, edited by E. Maccoby, T.
Newcomb, and E. Hartley, 359–370. New York: Holt, Rinehart, and Winston.
Yamagishi, T. 1996. “In-Group Bias and Culture of Collectivism.” In Proceedings of the
Korean Psychological Association Fiftieth Anniversary Conference, 119–133.
Yamagishi, T., K. Cook, and M. Watabe 1998. “Uncertainty, Trust and Commitment
Formation in the United States and Japan.” American Journal of Sociology 104: 165–
194.
30 Interdisciplinary Journal of Research on Religion Vol. 1 (2005), Article 7
Yamagishi, T., and M. Yamagishi. 1994. “Trust and Commitment in the United States
and Japan.” Motivation and Emotion. 18: 129–166.
Young, F. 1965. Initiation Ceremonies. New York, Bobbs-Merrill.
Zahavi, A., and A. Zahavi. 1997. The Handicap Principle: A Missing Piece of Darwin’s
Puzzle. New York: Oxford University Press.
Zellner, W. 2001. Extraordinary Groups: An Examination of Unconventional Lifestyles,
7th Edition. New York: Worth Publishers.