A Skeptical Review of Religious Prosociality Research

Jeremy Beahan, Luke W. Galen

It should come as no surprise to readers that the general public overwhelmingly assumes that religion is tied to morality. This can be de scribed as the “Religion Makes You Good” theory. In more scientific terms, this “religious prosociality hypothesis” predicts that religious belief is associated with a variety of positive social behaviors ranging from charitable generosity or time spent volunteering to a propensity to be helpful and honest. When discussions move beyond generalities to the realm of supporting evidence, there seems to be an impressive body of scientific work that supports this hypothesis. Indeed, studies have demonstrated that religious people do engage in more charity and volunteering1 and possess positive personal traits such as being more agreeable and conscientious.2 Beyond social behavior, religiosity has been associated with greater psychological well-being, including happiness and greater mental health.3 These conclusions are typically derived from surveys and polls (“naturalistic” methods) as well as from more controlled, laboratory-type studies of behavior such as, for example, economic games measuring cooperation and generosity. Another genre of study has used “priming” methods, that is, activating content by exposure to words or reminders on a screen, to the effect that religious “primes” make individuals more cooperative and honest. Several recent books have presented the case for religious prosociality, among them A Friendly Letter to Skeptics and Atheists: Musings on Why God is Good and Faith Isn’t Evil4 and American Grace: How Religion Divides and Unites Us.5 As the latter’s authors put it, religion makes “good neighbors.” Such seemingly overwhelming evidence for the prosocial effects of religion has been enough to convince even many nonbelievers that faith is beneficial, at least for others. In a recent Slate article titled “Don’t Trust the Godless,” Jesse Bering confessed, “Even as an atheist, I have more confidence in religious people. And now science is backing me up.”6 However, a closer examination reveals that the findings supporting the religious prosociality hypothesis are not necessarily a solid foundation. Rather, as pointed out in a recent scholarly literature review by one of us,7 they are almost always complex and tend to evaporate under scrutiny.

One factor that must be kept in mind when interpreting studies on the positive effects of religion is that they are almost always conducted in a context where a majority of individuals are religious, and a pervasive stereotype exists that religion ought to be a positive social influence. This is relevant because prosociality is often based on self-assessments that are provided by religious individuals who endorse those stereotypes. Even third-party ratings provided by family, colleagues, or peers are susceptible to the effects of a religious prosociality stereotype. For example, not only do religious individuals self-report having “nicer” personal characteristics than do nonreligious individuals, some studies have claimed to validate this greater religious prosociality by using third-party evaluators who also rated religious individuals as being nicer, more cooperative,8 altruistic, and empathetic.9 But in a predominantly religious society where 80 to 95 percent of individuals are religious and around 75 percent are at least nominally Christian, a significant proportion of those subjects’ families and peers are also religious—meaning that their judgments may be biased as well.

Ingroup favoritism is a well-studied phenomenon in social psychology. It is natural for individuals to derive self-esteem from the groups with which they are associated and to advocate for a positive public image for those who share their identity. Consistent with this theory, believers show favoritism toward other religious individuals and speak more poorly of the nonreligious or those from a different religious group.10 Often this favoritism is extended to other religious individuals regardless of whether or not they behave poorly. Believers will even rate a religious individual more highly than a nonreligious individual when the two exhibit identical behaviors. For example, in one study participants were asked to watch a video of a college student who spent his spring break volunteering for a disaster-relief organization. When the student was wearing a T-shirt with the famous “Jesus fish” symbol, he was rated as more likable, kind, and moral than the same student wearing a “Darwin fish” symbol.11 Most studies that rely on peer ratings of prosociality do not adequately control for this tendency to form ingroup biases.

Ideally, researchers should ensure that participants are unaware of the religious identity of those they are rating in order to prevent bias in the ratings. But this is rarely the case. Indeed, in the studies regarding cooperation and empathy cited above, raters were aware of the targets’ religious identities before they formed their impressions—casting doubt on their results. We know from studies in which the religious identity of the person making the evaluation and the religious identity of the person being evaluated are properly controlled that those identified as religious are consistently rated as more prosocial, while nonreligious individuals are rated as less prosocial, even when their actions are identical.12

But here’s an interesting twist: nonreligious individuals do not appear to rate their fellow nonbelievers as any more prosocial than the religious.11,13 In fact, American nonbelievers also appear to hold religious peers in high regard (though not as highly as the religious hold themselves). This indicates that nonbelievers’ judgments, too, have been swayed by a pro-religious cultural stereotype. Self-reports of happiness and personality measures such as agreeableness are more closely associated with religiosity in the United States than in the United Kingdom or Northern Europe where religion is less dominant.2 For example, in the United States, when participants have been asked to form impressions of personal characteristics based only on photographs of faces, smiling faces were judged to be more religious than nonsmiling faces,14 whereas in the United Kingdom the opposite was true.15 The relationship between religion and self-control, mental well-being, psychological adjustment, and social support is also weaker (or in some cases even reversed) in less-religious contexts. So in societies where the nonreligious are the majority, religion is not associated with prosocial outcomes.16

Naturalistic Studies, Controlled Studies, and Religious Priming

Findings about ingroup favoritism have implications for interpreting the “well-known” finding that religious people are more generous with their time and money. First, religious organizations themselves are the largest sources of charitable giving.17,18 This raises the question of whether giving by members of religious groups should be considered purely disinterested generosity or whether it might be affected by a degree of ingroup-based or nepotistic motivation. Second, many of those recipients of giving listed as “secular” often turn out to be semireligious, because many studies in the existing literature do not clearly separate religious versus secular recipients of charitable giving. For example, in Boston College’s Center on Wealth and Philanthropy study,19 the category of “religious giving” referred narrowly to houses of worship or congregations, whereas examples of what was termed secular giving also included gifts to a school, program, or hospital run by a religious organization or other recipient “that many would agree embodies spiritual values” (p. 7). One may prefer not to quibble about the recipients of assistance as long as the needy are being served, but from the standpoint of determining the generosity of the providers it is clearly necessary to separate generalized or universal prosocial giving from ingroup-specific giving.

One alternative to naturalistic studies with their accompanying “noise” and lack of control is to place individuals into controlled situations in the laboratory and ask them to make choices involving prosocial actions such as generous sharing or giving. These controlled studies provide more unambiguous conclusions. They also bring to light several interesting effects that compel us to qualify the religious prosociality hypothesis. One genre of such studies takes the form of economic games. In the “public goods” game, players can contribute money to a “public” fund internal to the experiment; at a trial’s end, all contributions are doubled and distributed equally among all the players regardless of their individual contribution. In the “trust game,” player A can chose whatever amount of money he or she wishes to send to player B. The amount is tripled by the experimenter, and then it is player B’s turn to choose an amount to transfer. Both games measure cooperation and trust. It is risky for any individual to give if the other players are selfish and refuse to reciprocate. But if everyone works together, each will be better off than before.

In behavioral economic studies where the religiosity of participants is known, a general trend emerges. Religious individuals cooperate more and give more money than nonreligious participants—but only in connection with those who share their religious identity.20 In one example, researchers found that clergy students extended greater monetary offers than non-clergy students, but only to those from their own group.21 These findings are most likely due to the phenomenon of ingroup favoritism. But they also indicate the pervasiveness of a pro-religious cultural stereotype, because nonreligious participants did not show the same ingroup favoritism. Rather, nonreligious participants also trusted religious participants more than their nonreligious peers and allocated more money to them22 even though, unbeknownst to them, that trust would not be reciprocated. Studies in which religious participants were paired with an outgroup member such as someone of a different religion or no religion or who was a “value violator” (for example, gay), did not identify any religious prosociality effects.

This pattern of preferential treatment is not limited to the behavioral economic studies but rather constitutes a general trend across the religious prosociality literature.23 Vassilis Saroglou has used the term minimal prosociality to refer to greater helping on the part of the religious that is extended to friends and ingroup members but not to outgroup members and those who threaten religious values.24 This could explain a body of work across different cultures and religions showing that religiosity is weakly but positively correlated with the value of benevolence (concern for close others such as friends and family) yet negatively related to the value of universalism (concern for all humans), 25,26 as well as regional discrepancies in the religious prosociality literature. For example, one study found that behavioral cooperation was greater among members of the Church of Jesus Christ of Latter-day Saints (Mormons or LDS) when the study was conducted in Utah than when the same study was conducted in Oregon.27 Mormons form the majority in Utah, but they are a minority in Oregon. Of course Mormons are not special in this regard (like other religious individuals, they are more prepared to love their neighbors when those neighbors are just like them). This highlights one potential pitfall facing researchers when interpreting data. The limited nature of religious prosociality may not be apparent in more homogeneous populations or in the absence of a comparison group. It only becomes visible in situations where religious individuals perceive a conflict of interest between their group and other religious groups. This problem plagues research on religious giving.

Religious Priming Studies

Scriptures and sermons abound with exhortations to “love thy neighbor” and to “do unto others as you would have them do onto you.” Shouldn’t frequent exposure to these messages result in some discernible increase in prosocial behaviors? The concept of “priming” involves activating concepts in study participants’ minds by exposing those participants to words, images, or contexts then measuring what effect this exposure has on behavior. In such studies, participants may be asked to read portions of religious scripture or to unscramble religious words. Priming can also be more subtle (for example, viewing religious symbols on clothing) or based on context (locating an experiment in a church as opposed to secular setting like a gymnasium). Priming can even exert influence without conscious awareness on the part of the participant. Participants’ performance on a particular task is measured relative to a control group to see if the priming had any effect.

To date, the evidence has indicated that priming with religious words results in greater honesty,28 generosity, sharing, and cooperation,29–32 as well as better self-control in distressing situations.33 For example, reading the Ten Commandments was found to reduce cheating among participants, and completing sentences with the word God had the effect of increasing participants’ resistance to temptation.34 Why are religious concepts so successful in priming such behaviors? In their review of the priming literature, Baumeister et al. (2010)35 citied “supernatural surveillance” as one possible mechanism, arguing “the belief that one’s actions are constantly and inescapably being observed by a divine being may be a strong stimulus and reminder to be aware of one’s actions.” In other words, religious primes put us on our best behaviors because they remind us that God might be watching. This interpretation of the findings does fit some of the evidence. For example, researchers found that conceptual God-related primes did increase public self-awareness among believers.36

Whatever the underlying mechanism, clearly exposure to religious concepts does increase prosocial behavior. A closer examination of the data, however, reveals a number of interesting relationships that qualify this conclusion. First, the evidence that positive religious words (like heaven, miracle, or bless) have a prosocial effect is stronger than is the case with neutral religious words like Bible, disciple, or chapel.37 Second, the positive effect of religious primes is not dependent on the participants’ personal level of religiosity. In other words, whether or not the individual participant is religious does not affect his or her ability to be primed by religious stimuli in most studies. The nonreligious also respond positively to religious primes and to the same degree. Obviously, this has implications for a pervasive general stereotype associated with religious ideas. Also, priming with secular concepts such as civil or court has the same effect in promoting honesty or lowering hypocrisy as do religious concepts.38 In fact, the mere presence of a mirror or pictures of eyespots in the laboratory has had similar effects in reducing dishonesty.39,40 So although some religious primes activate prosocial behavior, it appears that any reminder that others may be watching or any increase in self-conscious awareness is enough to promote more conscientious and socially desirable behavior.

Clearly, this has implications for the religious prosociality hypothesis. Religious concepts do not necessarily prime prosocial behavior because they are religious concepts. It may be that any concepts that are associated with morality in a particular culture trigger prosocial behavior or reputational concern. Norenzayan and Shariff41 conclude that any general moral stereotype (not just a religious one) will be effective in producing honesty or reducing cheating as a reminder of our reputations in the presence of third-party observers. Considering that controlled laboratory tests on honesty, cheating, and generosity have consistently failed to find any uniquely religious effect on behavior,42,43 it seems that this secular hypothesis is a better fit with the data. It may also make sense of another curious finding. Religious individuals are more likely to hold a negative stereotype of atheists as immoral or untrustworthy, but when they are primed with concepts of secular authority, their distrust of atheists is diminished.44 Believers know that atheists do not live their lives as if God is watching them. Without this supernatural monitoring, they may wonder what reason the godless have for behaving fairly. But this distrust can be ameliorated when they are reminded that morality can be monitored in other ways.

Although the positive effects of religious priming are the stuff of headlines, the dark side of this literature seldom gains the same level of attention. Numerous studies demonstrate that, in addition to positive effects, socially undesirable behaviors also manifest when subjects are exposed to religious messages. For example, in an experiment that measured aggression, participants who read passages from the Bible depicting God-sanctioned violence administered more electrical shocks to other participants than did those in the control group. In this study, although both religious and nonreligious participants became more aggressive following the prime, the effect was greater for the believers.45 In a similar study, subjects’ behavioral vengeance on other participants (by administering painful blasts of sound) was greater for those who were “extrinsically religious” (that is, religious for utilitarian reasons).46 In another example, those who had been previously screened for high levels of submissiveness became the most vengeful after being primed with religious words.47 They were also more willing to conform when pressured to be vengeful by the experimenter after religious priming.48 In another experiment, reading the biblical version of the Golden Rule had no effect in diminishing Christians’ homophobia, though strangely, reading the Buddhist version of the Golden Rule actually increased their homophobic responses,49 perhaps because the moral imperative was coming from a distrusted outgroup source. Likewise, unscrambling words associated with Christianity increased racial prejudice towards African Americans,50 while experiments conducted in a church setting (rather than a civic one) increased negative attitudes toward outgroups.51 Thus, a broad body of work indicates that religiosity is a double-edged sword, in that activating associated concepts has the effect of increasing both positive and negative behaviors.

Religiosity Takes Different Forms

Up to this point, we have been treating the concept of religiosity as if it refers to a unitary notion, such that all religious concepts are presumed to be of a piece and believers are similar to one another. However, the general concept of religiosity encompasses many components, and individuals are religious in different ways. Political and dispositional factors also complicate the picture. For example, some studies have used measures of general “intrinsic” religiosity, which refers to the degree that religion is of personal importance to an individual. However, general religious belief may contain distinct or even opposing tendencies, such as fundamentalism (exclusivist and literalist beliefs) as opposed to so-called “quest” religiosity (open-ended and spiritually seeking beliefs). Studies have also found a moderate degree of overlap between religiosity and authoritarianism (the strong valuation of obedience and submission to authorities). This is relevant to the prosociality debate because many studies have found that fundamentalist and authoritarian tendencies tend to relate to nonprosocial effects such as vengeance, prejudice, and hostility to outgroups. However, when the effects of authoritarianism are statistically removed or controlled, or when the type of religiosity in question is quest religiosity, more prosocial effects such as helpfulness, tolerance, and forgiveness are obtained. Such findings are relevant to the seemingly discrepant effects of religious priming mentioned previously, in which the activation of religious concepts has both prosocial and nonprosocial effects. More generally, the “glass half-empty/glass half-full” qualification regarding the prosocial effects of religion is due to the fact that religion can often involve contradictory and competing effects. But this also means that any reference to religion as being associated with solely positive effects (say, social justice) while ignoring its more authoritarian elements is selective at best.

Self-Report vs. Behavior

Does religious belief make one a better person? The faithful certainly think so. When asked to give an assessment of their own character, religious individuals report being more grateful, helpful, and forgiving than do nonreligious individuals. Many studies take such self-reports (employing instruments such as personality questionnaires) at face value, as constituting evidence of actual prosocial behavior. But such evaluations are in actuality affected by self-serving biases rather than by a realistic assessment of one’s own character. By now, it should be clear that self-report data tends to be unreliable. People are prone to forming positive illusions about themselves and therefore have a tendency to inflate their responses on questionnaires in a socially desirable direction. This may serve to inflate their own self-image (in a process known as “self-enhancement”), or it may seek to make a good impression on others (impression management). Although this is a pervasive tendency that is by no means confined only to the religious, it is more pronounced in those showing high intrinsic religiosity (that is, those who claim religion to be very important to them personally) than in the rest of the population. Highly religious people tend to view themselves as better than others generally,52 even other religious individuals.53 They also evaluate themselves more highly than nonreligious individuals on nonreligious attributes (for instance, intelligence). In experiments in which highly religious people are primed with God concepts, they become even more concerned with presenting positive self-assessments,36 and this priming works both ways. For example, in one study when high-self-esteem was primed in laboratory settings, Christians saw themselves as living up to Christian principles more often than their fellow believers.54 If, on the other hand, experimenters required subjects to disclose something negative about themselves or provided feedback that challenged their high self-assessments, those high in intrinsic religiosity engaged in self-deception as a means of compensating more often than those with low intrinsic religiosity.55 It seems that religiosity is connected to a presentational bias, such that highly religious people are particularly concerned with viewing themselves as moral persons and are threatened when that self-image is challenged.

A comparison of self-reports with controlled experiments using actual behavioral measures reveals the degree of this discrepancy. For example, religious individuals claim to value forgiveness more than others,56 but any effect of their religiosity on actual forgiveness in a given situation is negligible.57 Similarly, those high in intrinsic religiosity report a more grateful disposition but do not demonstrate greater behavioral gratitude.58 High intrinsic religiosity doesn’t reduce aggression (and can even increase aggression under certain circumstances), though it is likely to make one think that he or she is less aggressive.59,60 Likewise, religious people are no more honest than the nonreligious, though they are more likely to report that they are.61 In regards to helping others, fundamentalists perceive themselves as behaving altruistically toward everyone, but in reality they are more willing to help friends or like-minded individuals—not strangers or those with different values.62 We must remember that most people tend to overestimate how moral they actually are, so it is common to find a gap between how individuals predict they will behave and how they actually do behave.63,64 However, since the highly religious show a greater tendency towards self-enhancement on questionnaires, this disjunction between self-report and measured behavior is even wider among the religious than in the rest of the population. Indeed, the greatest gap between altruistic beliefs and altruistic behaviors is found in those who rate religion as more important to them personally.65

Although the experimental evidence fails to support the general stereotypes of religious prosociality, some researchers have described the results in such a way as to give a more positive impression. For example, in their review of the religiosity and forgiveness literature, McCullough and Worthington56 suggested that religious people are conscious that they should value forgiveness highly in order to be consistent with religious teachings and that “even if religious people are no more facile at forgiving in real-life situations than are less religious people, they do desire to be forgiving” (p. 1152). But praising believers for their moral intentions misses the point. They do not just desire to be better—in some cases they think they actually are superior. There are dangers inherent in having an unrealistic assessment of one’s own character and limits. The social fallout of abstinence-only sex education in America is just one of many possible examples. Researchers have found that greater religious commitment was associated with a “temptation bias”—a belief that one can resist temptation better than others, especially in sexual situations.66 Yet the vast majority of Christian youth who take abstinence pledges still engage in premarital intercourse. This would not be such a problem if they didn’t also show greater risk for pregnancy and STDs due to having unprotected sex! If promoters of abstinence-only sexual education were more realistic about the likelihood of failure, perhaps they would see the value in teaching young adults to take precautions.

Compared to Whom?

Often the wording of study findings (for example, “religious people are more prosocial”) contains an implied comparison (“than nonreligious individuals”). Typically, the methodology used in studies is to compare in a general population the highly religious to the less religious, such as the unaffiliated or infrequent church attenders. However, if one is attempting to identify effects due specifically to belief, this method combines belief with other factors. Chief among these is that more religious individuals also attend church and are therefore strongly socially embedded within a group. Unaffiliated individuals (often called the “Nones”) by definition are not members of such a group, so that those at the “low end” of religiosity scales are therefore not “joiners.” As a result, most studies represent a comparison between religious group members and weakly religious nongroup members, along with a minority of atheists and agnostics.

The nature of relative comparison is important to the religious prosociality debate because frequently the strongest prosocial effects associated with religion tend to be connected with church attendance and the social context of religious groups, rather than with metaphysical beliefs. That is, frequent church attenders do indeed show a pattern of more prosocial behaviors than those who don’t frequently attend church. However, when measured as a function of metaphysical beliefs (say, how certain one is of God’s existence), the findings are less clear. For example, although more frequent church attendance has been linked to modestly lower rates of mental illness such as depression, the effect is negligible when religiosity is measured as strength of belief.67 In another example using Gallup survey data, differences in generosity when measured as a function of religious importance were smaller than those measured as variation in religious attendance.68 In fact, many studies that statistically control or separate purely social factors such as group attendance and the strength of social relationships have found greatly diminished or nonexistent effects of religious belief on prosocial measures.69

This clearly has implications for the role of purely secular factors, such as group involvement and social embeddedness in the community, as opposed to any unique effect of religious engagement. For example, could a committed secular group have effects on its membership similar to that of a church? In the above-mentioned book American Grace, Putnam and Campbell conclude that “religious beliefs . . . turn out to be utterly irrelevant to explaining the religious edge in good neighborliness” (p. 465), and “even an atheist who happened to become involved in the social life of a congregation . . . is much more likely to volunteer in a soup kitchen than the most fervent believer who prays alone” (pp. 472–473).5 Indeed, this “social group-based prosociality” is a pervasive effect.

Moreover, when we rank individuals by their level of prosocial behavior, it is unaffiliated or nominally religious individuals, not committed secular individuals, who are at the opposite end of the prosocial continuum from highly engaged religious believers. The prosocial continuum separates the religiously committed from the unaffiliated and uncommitted; it does not separate committed believers from committed unbelievers. As we will see below, measurements that distinguish only between religious and nonreligious respondents have the effect of lumping principled atheists and socially engaged nonbelievers together with those merely claiming no religious affiliation. (Though it is worth remembering that most of those who claim no religious affiliation are still religious.)

Curvilinear Effects

The effect of lumping committed atheists and nonbelievers together with the merely unaffiliated is an imprecise “watering down” of any effect of religious belief or the lack thereof. Not surprisingly, studies that lump in this way appear to show a linear relationship between religiosity and prosociality—the more religious one is, the more helpful, happier, and healthy one will be. However, there is a smaller number of studies comparing the highly religious with the confidently nonreligious, and this group of studies shows interesting effects relevant to the religious prosociality hypothesis. Some studies compare church attenders with committed atheists and/or humanists. Instead of a linear relationship, these studies often reveal a curvilinear or U-shaped relationship between belief and prosociality. That is, confident religious believers and confident nonbelievers show nearly identical outcomes, with the nominally religious and unsure nonreligious in between. For example, nominal believers (not atheists) show the highest levels of depression, the poorest mental health, and generally report less satisfaction with life.70 In fact, data gathered from over fifty nations corroborates these findings—the World Values Survey found that both those who claimed religion is “very important” and those who claimed it was “not important at all” were the happiest.71

Curvilinear effects are also found in the moral realm as well. For example, physicians’ likelihood of practicing among the underserved exhibits a curvilinear relationship with intrinsic religiosity.72 In the personality trait of conscientiousness, strongly secular individuals are equivalent to strongly religious individuals.73 One reason for this curvilinear effect may have to do with the tendency to behave according to the dictates of one’s conscience when faced with pressure to conform. For example, in Milgram’s infamous obedience experiments, participants were instructed to give “electric shocks” (later revealed to be fake) to a confederate who pretended to be distressed and in pain. When Bock and Warren replicated the study, they found both “extreme believers” and “extreme non-believers” were the most likely to disobey the researcher’s unethical orders.74 Moderate religious believers were the most compliant. These lab results are consistent with one of history’s darkest real-world tests of obedience to authority. Studies of rescuers of Jews during World War II found that the highest proportion of rescuers were either highly religious or completely nonreligious.75 The researchers suggested that moderates were less likely to help because those who have strength in their convictions (regardless of whether those convictions are religious in nature) are more likely to act according to personal conscience than simply to obey prevailing social norms. It appears that confidence in one’s worldview and regular affiliation with like-minded people are more important to well-being and moral integrity than the particular content of an individual’s metaphysical beliefs. But most studies miss this deeper relationship. When nonbelievers are “unevenly yoked” with the nominally religious, proper comparisons cannot be made, and the curvilinear effects described above remain invisible to researchers.

Criterion Contamination

Such unequal comparisons in the religious-prosociality literature are sometimes due to the question-begging manner in which religion or spirituality is defined. Studies will often describe how the “religiously-engaged” or those with “spirituality” have advantages in their mental well-being or social relationships. For example, the U.S. military includes a “Spiritual Fitness” dimension in its instrument assessing soldiers’ wellness and mental health, referring to greater resiliency among those who are spiritually inclined.76 However, an examination of the actual question items on such measures indicates a flaw in the way concepts such as spirituality are measured—a problem known to researchers as “criterion contamination.” When making a prediction of some criterion, it is generally more meaningful (and difficult) when the items used in prediction do not themselves contain elements of the items being predicted. To do otherwise virtually guarantees a positive correlation. For instance, it is entirely circular to say that “religiously-engaged individuals have greater social networks” when “religious engagement” is itself defined by having church social contacts. This is equivalent to suggesting that “socially engaged religious believers are socially engaged.” Similarly, many “spirituality” scales measure concepts that do not necessarily refer to supernatural beliefs such as a sense of universalism, personal meaning, or social relationships, with items such as “I believe there is a larger meaning to life”; “It is important for me to give something back to my community”; “I believe that humanity as a whole is basically good”; “I am concerned about those who will come after me in life”; and so on. Clearly, agreeing with those statements is not tantamount to endorsing a belief in some sort of spiritual power or God. But the findings are often interpreted as representing such an endorsement—although any atheist with a broader sense of meaning or purpose will score as being “spiritual” on these scales. This has the effect of artificially inflating the apparent relationship between religiosity or spirituality and positive prosocial outcomes because those with a sense of meaning and purpose will also be socially engaged and have greater well-being. Numerous studies ostensibly demonstrating that religiosity is related to prosocial outcomes are in actuality examples of criterion contamination effects.

How to Critically Examine Studies

The question “Does religion make us prosocial?” does not admit a simple answer. The conclusion one reaches depends on the measure of religiosity used and the way prosociality is defined. To achieve a clear understanding, one must take into account a host of complicating factors. We would like to suggest ten questions to aid the critical consumer in assessing the merits of research on religious prosociality:

  1. Has the research controlled for the possibility that stereotypes, such as the expectation that religious individuals will be more prosocial, have affected self-reports and ratings?
  2. Are the results based on evidence that has been compromised by ingroup favoritism bias?
  3. When prosocial effects follow the priming of religious concepts, would those same effects follow secular primes?
  4. Is the study also able to detect potential negative as well as positive effects of religious priming?
  5. Is the research based on self-reports (such as hypothetical behaviors or attitudes) or does it measure actual behaviors?
  6. Could the context of the study have an impact on the results? For example, a study conducted in highly religious contexts like the United States may not obtain like results in predominantly nonreligious contexts like northern Europe.
  7. Are the results solely attributable to religious belief itself, or could they be due to group affiliation? If church-attending believers are compared to nonchurch attenders, the source of any differences may not be clearly attributable.
  8. Related to number 7, does the study conflate nonbelief with low religiosity?
  9. Do the groups under comparison allow for an examination for curvilinear effects?
  10. Has “religion” or “spirituality” been defined in a way that would also include prosocial behavior?


Recommended Reading

  • Galen, L. W. 2012. “Does Religious Belief Promote Prosociality?: A Critical Examination.” Psychological Bulletin 138 (2012): 876–906.
  • Galen, L. W. “The Complex and Elusive Nature of Religious Prosociality: Reply to Myers and Saroglou.” Psychological Bulletin 138 (2010): 918–923.


  1. A. C. Brooks, Who Really Cares: The Surprising Truth about Compassionate Conservatism (New York, NY: Basic Books, 2006).
  2. V. Saroglou, “Religiousness as a Cultural Adaptation of Basic Traits: A Five-Factor Model Perspective.” Personality and Social Psychology Review 14, No.1 (2012): 108–25.
  3. D. G. Myers, “The Funds, Friends, and Faith of Happy People.” American Psychologist 55, No. 1 (2000): 56–67.
  4. D. G. Myers, A Friendly Letter to Skeptics and Atheists: Musings on Why God Is Good and Faith Isn’t Evil (San Francisco, CA: Jossey-Bass, 2008).
  5. R. D. Putnam and D. E. Cam
    pbell, American Grace: How Religion Divides and Unites Us ( New York, NY: Simon & Schuster, 2010).
  6. J. Bering, “Don’t Trust the Godless.” Slate, July 1, 2012. Accessed at http://www.salon.com/2012/07/01/dont_trust_the_godless/.
  7. L. W. Galen, “Does Religious Belief Promote Prosociality?: A Critical Examination.” Psychological Bulletin 138, No. 5 (2012): 876-906.
  8. C. G. Ellison, “Are Religious People Nice People? Evidence from the National Survey of Black Americans.” Social Forces 71, No. 2 (1992): 411–30.
  9. V. Saroglou, I. Pichon, L. Trompette, M. Verschueren, and R. Dernelle, “Prosocial Behavior and Religion: New Evidence Based on Projective Measures and Peer Ratings. Journal for the Scientific Study of Religion 44, No. 3 (2005): 323–48.
  10. W. C. Rowatt, L. M. Franklin, and M. Cotton, “Patterns and Per­sonality Correlates of Implicit and Explicit Attitudes toward Christians and Muslims.” Journal for the Scientific Study of Religion 44, No. 1 (2005): 29–43.
  11. L. W. Galen, C. M. Smith, N. Knapp, and N. Wyngarden, “Per­cep­tions of Religious and Non-religious Targets: Exploring the Effects of Perceivers’ Religious Fundamentalism.” Journal of Applied Social Psychology 41, No. 9 (2011): 2123–2143.
  12. D. R. Widman, K. E. Corcoran, and R. E. Nagy, “Belonging to the Same Religion Enhances the Opinion of Others’ Kindness and Morality.” Journal of Social, Evolutionary, and Cultural Psychology 3, No. 4 (2009): 281–89.
  13. L. W. Galen, and A. Ver Wey, “Unpacking Religious Prosociality: Personality Ratings Are Contaminated by Religious Stereotype and Ingroup Bias.” Symposium presented at the 16th meeting of the European Conference on Personality, July 2012, Trieste, Italy.
  14. L. P. Naumann, S. Vazire, P. J. Rentfrow, and S. D. Gosling, “Per­sonal­ity Judgments Based on Physical Appearance.” Personality and Social Psychology Bulletin 35, No. 12 (2009): 1661–1671.
  15. R. Highfield, R. Wiseman, and R. Jenkins, “In Your Face.” New Scientist 201, No. 2695 (2009): 28–32.
  16. E. Diener, L. Tay, and D. G. Myers, “The Religion Paradox: If Religion Makes People Happy, Why Are So Many Dropping Out?” Journal of Personality and Social Psychology 101, No. 6 (2011): 1278–1290.
  17. American Association of Fundraising Counsel Trust for Philan­thropy. Giving USA: The Annual Report on Philanthropy For The Year 2002 (New York, NY: American Association of Fundraising Counsel, 2002).
  18. V. A. Hodgkinson and M. S. Weitzman, Giving and Volunteering in the United States: Findings from a National Survey (Washington, DC: Independent Sector, 1996).
  19. Center on Wealth and Philanthropy, “Geography and Giving: The Culture of Philanthropy in New England and the Nation (Boston, MA: Boston Foundation, 2007).
  20. A. Ben-Ner, B. P. McCall, M. Stephane, and H. Wang, “Identity and Ingroup/Out-Group Differentiation in Work and Giving Behaviors: Experimental Evidence.” Journal of Economic Behavior & Organization 72, No. 1 (2009): 153–70.
  21. C. Fershtman, U. Gneezy, and F. Verboven, “Discrimination and Nepotism: The Efficiency of the Anonymity Rule.” Journal of Legal Studies 34, No. 2 (2005): 371–96.
  22. J. H. W. Tan and C. Vogel, “Religion and Trust: An Experimental Study.” Journal of Economic Psychology 29, No. 6 (2008): 832–48.
  23. A. Norenzayan and A. F. Shariff, “The Origin and Evolution of Religious Prosociality.” Science 322, No. 5898 (2008): 58–62.
  24. V. Saroglou, “Religion’s Role in Prosocial Behavior: Myth or Reality?” Psychology of Religion Newsletter 31, 1–8 (Spring 2006).
  25. M. Pepper, T. Jackson, and D. Uzzell, “A Study of Multidimensional Religion Constructs and Values in the United Kingdom.” Journal for the Scientific Study of Religion 49, No. 1 (2010): 127–46.
  26. V. Saroglou, V. Delpierre, and R. Dernelle, “Values and Religiosity: A Meta-Analysis of Studies Using Schwartz’s Model.” Personality and Individual Differences 37, No. 4 (2004): 721–34.
  27. J. Orbell, M. Goldman, M. Mulford, and R. Dawes, “Religion, Con­text, and Constraint Toward Strangers.” Rationality and Society 4, No. 3 (1992): 291–307.
  28. B. Randolph-Seng and M. E. Nielsen, “Honesty: One Effect of Primed Religious Representations.” International Journal for the Psychology of Religion 17, No. 4 (2007): 303–15.
  29. A. M. Ahmed and O. Salas, “Is the Hand of God Involved in Human Cooperation?” International Journal of Social Economics 36, No. 1/2 (2009): 70–80.
  30. B. Paciotti, P. Richerson, B. Baum, M. Lubell, T. Waring, R. McElreath, . . . E. Edsten, “Are Religious Individuals More Generous, Trust­ing, and Cooperative? An Experimental Test of the Effect of Religion on Prosociality.” In Research in Economic Anthropology: Vol. 31. The Economics of Religion: Anthropological Approaches, series edited by D. C. Wood and volume edited by L. Obadia and D. C. Wood (Bingley, England: Emerald, 2011), pp. 267–305.
  31. A. F. Shariff and A. Norenzayan, “God Is Watching You: Priming God Concepts Increases Prosocial Behavior in an Anonymous Economic Game.” Psychological Science 18, No. 9 (2007): 803–09.
  32. A. M. Ahmed and O. Salas, “In the Back of Your Mind: Subliminal Influences of Religious Concepts on Prosocial Behavior,” Working Papers in Economics No. 331. Gothenburg, Sweden: Gothenburg School of Business, Economics and Law, University of Gothenburg, 2008).
  33. J. Y. Sasaki and H. S. Kim, “At the Intersection of Culture and Religion: A Cultural Analysis of Religion’s Implications for Secondary Con­trol and Social Affiliation.” Journal of Personality and Social Psychology 101, No. 2 (2011): 401–14.
  34. K. Laurin, A. C. Kay, and G. M. Fitzsimons, “Divergent Effects of Activating Thoughts of God on Self-Regulation.” Journal of Personality and Social Psychology 102, No. 1 (2011): 4–21.
  35. R. F. Baumeister, I. M. Bauer, and S. A. Lloyd, “Choice, Free Will, and Religion.” Psychology of Religion and Spirituality 2, No. 2 (2010): 67–82.
  36. W. M. Gervais and A. Norenzayan, “Like a Camera in the Sky? Thinking About God Increases Public Self-Awareness and Socially Desir­able Responding.” Journal of Experimental Social Psychology 48, No. 1 (2012): 298–302.
  37. I. Pichon, G. Boccato, and V. Saroglou, “Nonconscious Influences of Religion on Prosociality: A Priming Study. European Journal of Social Psychology 37, No. 5 (2007): 1032–1045.
  38. A. F. Shariff and A. Norenzayan, “God Is Watching You: Priming God Concepts Increases Prosocial Behavior in an Anonymous Economic Game.” Psychological Science 18, No. 9 (2007): 803–09.
  39. M. Bateson, D. Nettle, and G. Roberts, “Cues of Being Watched Enhance Cooperation in a Real-World Setting.” Biology Letters 2, No. 3 (2006): 412–14.
  40. C. D. Batson, E. R. Thompson, G. Seuferling, H. Whitney, and J. A. Strongman, “Moral Hypocrisy: Appearing Moral to Oneself Without Being So.” Journal of Personality and Social Psychology 77, No. 3 (1999): 525–37.
  41. A. Norenzayan and A. F. Shariff, “The Origin and Evolution of Religious Prosociality.” Science 322, No. 5898 (2008): 58–62.
  42. R. E. Smith, G. Wheeler, and E. Diener, “Faith Without Works: Jesus People, Resistance to Temptation, and Altruism.” Journal of Applied Social Psychology 5, No. 4 (1975): 320–30.
  43. W. P. Williamson and A. Assadi, “Religious Orientation, Incentive, Self-Esteem, and Gender as Predictors of Academic Dishonesty: An Experimental Approach.” Archive for the Psychology of Religion 27, No. 1 (2005): 137–58.
  44. W. M. Gervais, “Finding the Faithless: Perceiv
    ed Atheist Prevalence Reduces Anti-Atheist Prejudice.” Personality and Social Psychology Bulletin 37, No. 4 (2011): 543–56.
  45. B. J. Bushman, R. D. Ridge, E. Das, C. W. Key, and G. L. Busath, “When God Sanctions Killing: Effect of Scriptural Violence on Aggression.” Psychological Science 18, No. 3 (2007): 204–07.
  46. M. M. Leach, M. E. Berman, L. Eubanks, “Religious Activities, Religious Orientation, and Aggressive Behavior.” Journal for the Scien­tific Study of Religion 47, No. 2 (2008): 311–19.
  47. V. Saroglou, O. Corneille, and P. Van Cappellen, “Speak, Lord, Your Servant Is Listening”: Religious Priming Activates Submissive Thoughts and Behaviors.” International Journal for the Psychology of Religion 19, No. 3 (2009):143–54.
  48. P. Van Cappellen, O. Corneille, S. Cols, and V. Saroglou, “Beyond Mere Compliance to Authority Figures: Religious Priming Increases Conformity to Informational Influence Among Submissive People.” International Journal for the Psychology of Religion 21, No. 2 (2011): 97–105.
  49. O. Vilaythong Tran, N. M. Lindner, and B. A. Nosek, “Do Unto Others”: Effects of Priming the Golden Rule on Buddhists’ and Christians’ Attitudes Toward Gay People.” Journal for the Scientific Study of Religion 49, No.3 (2010): 494–506.
  50. M. K. Johnson, W. C. Rowatt, and J. LaBouff, “Priming Chris­­­tian Religious Concepts Increases Racial Prejudice.” Social Psycho­­logical & Personality Science 1, No. 2 (2010): 119–26.
  51. J. LaBouff, W. C. Rowatt, M. K. Johnson, and C. Finkle, “Differences in Attitudes Towards Outgroups in a Religious or Non-Religious Context in a Multi-National Sample: A Situational Context Priming Study.” International Journal for the Psychology of Religion 22, No. 1 (2012): 1–9.
  52. W. C. Rowatt, A. Ottenbreit, K. P. Nesselroade Jr., and P. A. Cunningham, “On Being Holier-Than-Thou or Humbler-Than-Thee: A Social-Psychological Perspective on Religiousness and Humility.” Journal for the Scientific Study of Religion 41, No. 2 (2002): 227–37.
  53. C. T. Burris and L. M. Jackson, “Social Identity and the True Believer: Responses to Threatened Self-Stereotypes Among the Intrinsically Religious.” British Journal of Social Psychology 39, No. 2 (2000): 257–78.
  54. M. Alicke and C. Sedikides, “Self-Enhancement and Self-Protection: What They Are and What They Do.” European Review of Social Psychology 20, No. 1 (2009): 1–48.
  55. C. T. Burris and G. S. Navara, “Morality Play or Playing Morality? Intrinsic Religious Orientation and Socially Desirable Responding.” Self and Identity 1, No. 1 (2002): 67–76.
  56. M. R. McCullough and E. L. Worthington Jr., “Religion and the Forgiving Personality.” Journal of Personality 67, No. 6 (1999): 1141–1164.
  57. R. P. Brown, C. D. Barnes, and N. J. Campbell, “Fundamentalism and Forgiveness.” Personality and Individual Differences 43, No.6 (2007): 1437–1447.
  58. J.-A. Tsang, A. Schulwitz, and R. D. Carlisle, “An Experimental Test of the Relationship Between Religion and Gratitude.” Psychology of Religion and Spirituality 4, No. 1 (2011): 40–55.
  59. M. M. Leach, M. E. Berman, and L. Eubanks, “Religious Activities, Religious Orientation, and Aggressive Behavior.” Journal for the Scientific Study of Religion 47, No. 2 (2008): 311–19.
  60. T. Greer, M. Berman, V. Varan, L. Bobrycki, and S. Watson, “We Are a Religious People; We Are a Vengeful People.” Journal for the Scientific Study of Religion 44, No.1 (2005): 45–57.
  61. R. W. Hood Jr., P. C. Hill, and B. Spilka, The Psychology of Religion: An Empirical Approach, 4th ed. (New York, NY: Guilford Press, 2009).
  62. J. Blogowska and V. Saroglou, “Religious Fundamentalism and Limited Prosociality as a Function of the Target.” Journal for the Scientific Study of Religion 50, No. 1 (2011): 44–60.
  63. C. D. Batson, The Altruism Question: Toward a Social Psychological Answer (Hillsdale, NJ: Erlbaum, 1991).
  64. T. D. Wilson, Strangers to Ourselves: Discovering the Adaptive Unconscious (Cambridge, MA: Bel­knap Press/Harvard University Press, 2002).
  65. C. C. Ji, L. Pendergraft, and M. Perry, “Religiosity, Altruism, and Altruistic Hypocrisy: Evidence from Protestant Adolescents.” Review of Religious Research 48, No. 2 (2006): 156–78.
  66. S. Garos, J. K. Beggan, and A. Kluck, “Temptation Bias: Seeing Oneself as Better Able Than Others to Resist Temptation.” In Research in the Social Scientific Study of Religion, Volume 15, edited by R. L. Peidmont and D. O. Moberg (Leiden, the Netherlands: Brill, 2004), pp. 235–60.
  67. T. B. Smith, M. E. McCullough, and J. Poll, “Religiousness and Depression: Evidence for a Main Effect and the Moderating Influence of Stressful Life Events.” Psychological Bulletin 129, No. 4 (2003): 614–36.
  68. B. G. Smith and R. Stark. “Religious Attendance Relates to Generosity Worldwide.” September 4, 2009. Accessed at http://www.gallup.com/poll/122807/Religious-Attendance-Relates-Generosity-Worldwide.aspx.
  69. C. Lim and R. D. Putnam, “Religion, Social Networks, and Life Satisfaction.” American Sociological Review 75, No. 6 (2010): 914–33.
  70. L. W. Galen and J. Kloet, “Mental Well-Being in the Religious and the Non-Religious: Evidence for a Curvilinear Relationship.” Mental Health, Religion & Culture 14, No. 7 (2011): 673–89.
  71. T. Rees, “The Happiness Smile.” Web log message, August 6, 2009. Retrieved from http://epiphenom.fieldofscience.com/2009/08/happiness
  72. F. A. Curlin, L. S. Dugdale, J. D. Lantos, and M. H. Chin, “Do Religious Physicians Dis­proportionately Care for the Underserved?” Annals of Family Medicine 5, No. 4 (2007): 353–60.
  73. L. W. Galen and J. Kloet, “Personality and Social Integration Factors Distinguishing Nonreligious From Religious Groups: The Importance of Controlling for Attendance and Demographics.” Archive for the Psychology of Religion 33, No. 2 (2011): 205–28.
  74. D. C. Bock, and N. C. Warren, “Religious Belief as a Factor in Obedience to Destructive Demands.” Review of Religious Research 13, No. 3 (1972): 185–91.
  75. S. P. Oliner and P. M. Oliner, The Altruistic Personality: Rescuers of Jews in Nazi Europe (New York, NY: Free Press, 1988).
  76. B. B. Hagerty, “Army’s ‘Spiritual Fitness’ Test Angers Some Soldiers.” Accessed January 13, 2011, at http://www.npr.org/2011/



Jeremy Beahan

Luke W. Galen is associate professor of psychology at Grand Valley State University in Michigan. Jeremy Beahan teaches world religions, aesthetics, and introduction to philosophy at Kendall College of Art and Design in Grand Rapids. He is the producer and cohost of the popular skeptical podcast Reasonable Doubts.

Luke W. Galen

Luke W. Galen is associate professor of psychology at Grand Valley State University in Michigan. Jeremy Beahan teaches world religions, aesthetics, and introduction to philosophy at Kendall College of Art and Design in Grand Rapids. He is the producer and cohost of the popular skeptical podcast Reasonable Doubts.

It should come as no surprise to readers that the general public overwhelmingly assumes that religion is tied to morality. This can be de scribed as the “Religion Makes You Good” theory. In more scientific terms, this “religious prosociality hypothesis” predicts that religious belief is associated with a variety of positive social behaviors ranging from …

This article is available to subscribers only.
Subscribe now or log in to read this article.