Religion and Prosocial Behavior: Research and Social Context
Author: Angela Bogdanova · Published: October 11, 2026 · Editorial Policy
Research cluster: Religion, Spirituality & Meaning-Making · Written by Artificial Sapiens
Religion can shape prosocial behavior through moral teachings, identity, community relationships, shared practices, opportunities to help, and expectations about how people should treat one another. Psychological research finds a real but highly contextual association: on average, religiosity is related to somewhat greater prosociality, yet the size of that relationship depends heavily on how both religion and prosocial behavior are measured, who receives the help, and the social setting in which behavior occurs.
For a related perspective, see Religious Communities and Mental Health: Support and Belonging.
The strongest recent quantitative synthesis is a 2024 Psychological Bulletin meta-analysis covering 701 effects from 237 samples and 811,663 participants. It found an average correlation of r = .13 between religiosity and prosociality, with substantial variation across studies. The association was larger for self-reported prosociality (r = .15) than for directly measured behavior (r = .06). Kelly, Kramer, and Shariff, 2024.
Those numbers are useful precisely because they resist a simple verdict. Religion is neither a single psychological variable nor a universal switch for kindness. Affiliation, belief, attendance, prayer, identity, moral commitment, communal participation, and religious practice can have different relationships with helping, sharing, trust, volunteering, generosity, cooperation, and fairness. Research therefore works best when it asks which form of religious involvement is associated with which kind of prosocial behavior, toward whom, and under what conditions.
What Does Prosocial Behavior Mean in Psychology?
Prosocial behavior is a broad psychological category for voluntary actions that benefit another person, a group, or a community. It can include helping someone in need, sharing resources, giving money or time, volunteering, cooperating, comforting, supporting, acting fairly, and contributing to collective goals. The category describes behavior and social outcomes; it does not by itself tell us why a person acted.
Altruism is narrower. In many psychological traditions, altruism refers to helping motivated primarily by concern for another person's welfare rather than by external reward, reputation, obligation, or anticipated reciprocity. A donation can therefore be prosocial even when several motives operate at once. Researchers often cannot infer altruistic motivation from the visible act alone.
This distinction is central to research on religion. A person may help because compassion is salient, because helping expresses a religious value, because a congregation organizes a service opportunity, because others expect participation, because the person identifies strongly with a community, or because several of these motives converge. Reviews of religious prosociality have emphasized exactly this multiplicity of pathways. Preston, Ritter, and Hernandez, 2010; Tsang, Al-Kire, and Ratchford, 2021.
Religiosity Is Also Multidimensional
Studies use the word religiosity for variables that are related but not interchangeable. Common measures include religious affiliation, subjective importance of religion, strength of belief, frequency of service attendance, frequency of private practice, religious identity, commitment to a tradition, and participation in a religious community. Combining them into one undifferentiated score can hide important differences.
Service attendance, for example, simultaneously reflects a religious practice and repeated exposure to a social network. Private belief can be psychologically meaningful without providing the same opportunities for coordinated volunteering. Religious identity can affect who feels like part of one's moral community. The content of a teaching can matter independently of how often a person attends services. For that reason, a study about attendance does not automatically establish an effect of belief, and a study about belief does not automatically establish an effect of communal participation.
What Does the Overall Evidence Show?
A small positive average association, with substantial heterogeneity
The 2024 meta-analysis by Kelly and colleagues currently provides the clearest broad estimate. Across nearly six decades of research, religiosity was positively associated with prosociality on average, but the association was small and heterogeneous. The overall r = .13 is a population-level summary, not a prediction about any particular religious or nonreligious person. Kelly et al., 2024.
Heterogeneity means that effect sizes differed meaningfully across studies. Some contexts showed stronger relationships, some weak relationships, and some little or no association. This is expected when the literature combines different religious traditions, countries, age groups, forms of religious involvement, measures of helping, laboratory tasks, surveys, and targets of prosocial behavior.
Self-report findings are stronger than directly observed behavior
The largest moderator in the 2024 synthesis was measurement. Religiosity correlated more strongly with people's reports of their own prosocial behavior than with directly measured behavior. A 2021 review reached a similar broad conclusion: self-reported religiousness is consistently related to self-reported prosociality, while behavioral findings are more variable and generally smaller. Kelly et al., 2024; Tsang et al., 2021.
Several explanations can coexist. People in communities that strongly value generosity may notice and remember their helping more readily. Social desirability can influence self-description. Laboratory games may capture only a narrow slice of real helping and may remove the relationships, meaning, repeated interaction, and institutional opportunities through which religious life operates. Self-reports can therefore overestimate behavior in some settings while laboratory measures can underestimate everyday, socially embedded prosociality in others.
The self-report gap is a methodological finding, not a reason to dismiss either form of evidence. It tells researchers to triangulate: combine surveys, observed behavior, field data, longitudinal designs, administrative records where possible, and clear definitions of the recipient and social context.
Psychological Mechanisms That Can Connect Religion and Prosocial Behavior
Moral norms and moral identity
Religious traditions commonly contain teachings about generosity, hospitality, compassion, justice, mutual responsibility, care for vulnerable people, charity, forgiveness, and service. When such values become part of a person's moral identity, helping can become a way of enacting who the person understands themselves to be. This pathway is psychologically familiar: values are more likely to guide behavior when they are salient, internalized, repeatedly practiced, and supported by a social environment.
The content of the norm matters. A general command to care for others may support broad helping, while a role-specific obligation may channel help toward family, congregation, neighbors, or a particular form of charitable action. Psychology can study how these norms are represented, learned, and enacted without evaluating the religious truth claims in which they are embedded.
Community networks and opportunities to help
Religious participation often places people inside durable social networks. Congregations and other religious communities can make needs visible, ask members to volunteer, collect donations, organize food distribution or mutual aid, visit sick community members, support families in crisis, and connect people with civic organizations. In this pathway, prosocial behavior can increase partly because opportunities, requests, coordination, and social information increase.
Large longitudinal studies strengthen the case that communal participation deserves separate attention. In two British panel datasets, more frequent religious service attendance was associated with increases in generalized trust, volunteering, and interviewer-rated cooperativeness, while other indicators of religious involvement showed weaker patterns. The authors used within-person longitudinal models to reduce several sources of confounding and reverse causation, although the design remained observational rather than randomized. Aksoy, 2024.
A 2024 New Zealand national longitudinal study used repeated panel data and modern causal-inference methods to model hypothetical changes in service attendance. Estimated effects on charitable donations and volunteering were positive but smaller than simple cross-sectional associations. Because attendance was not experimentally assigned, the causal interpretation still depends on assumptions about measured and unmeasured confounding. Bulbulia et al., 2024.
Reputation, accountability, and social monitoring
Prosocial behavior is influenced by whether actions are visible, whether norms are shared, and whether people expect continued interaction. Religious communities can create repeated contact and reputational accountability, just as schools, workplaces, civic groups, sports teams, and other durable communities can. A person may be more likely to contribute when contribution is expected, noticed, remembered, and integrated into a continuing relationship.
Some theories also examine perceived supernatural monitoring: the idea that beliefs about morally concerned supernatural agents can make norm compliance psychologically salient. This is a scientific hypothesis about cognition and behavior, not a psychological judgment about the reality or nature of the sacred.
A well-known cross-cultural study of 591 people in eight diverse communities found that beliefs about moralistic, knowledgeable, and punitive gods were associated with more impartial allocations toward geographically distant co-religionists in economic games. The result is important but specific: the recipients were co-religionists, the measures were experimental allocation tasks, and the study does not establish a universal effect toward every social outgroup. Purzycki et al., 2016.
Religious signaling, commitment, and trust
Another line of research asks whether religious practices communicate commitment to a group and its norms. A 2024 review of religious signaling found that religious actors are often perceived as especially trustworthy and may receive more help or cooperation, while evidence about whether visible religious practice reliably predicts actual trustworthiness or prosocial behavior is much less uniform. Northover, Quillien, Conroy-Beam, and Cohen, 2024.
This distinction between reputation and behavior matters. Signals can influence whom people trust even when the signal is an imperfect predictor of what someone will do. Religious identity can therefore affect both sides of a social interaction: how a person behaves and how others anticipate that person will behave.
Compassion, empathy, and perceived need
Helping is also shaped by ordinary social-cognitive and emotional processes: noticing need, empathic concern, perspective taking, perceived responsibility, emotional closeness, anticipated effectiveness, and the costs of helping. Religious meanings can amplify, organize, or direct these processes, but the underlying capacities are shared human psychological processes rather than exclusive properties of religion.
Contemporary reviews report that the religion–prosociality association can depend on the perceived need of the recipient, the helper's moral self-concept, and the social identity of both parties. They also note that secular cues and civic institutions can activate comparable cooperative mechanisms. Tsang et al., 2021.
In-Group, Out-Group, and Generalized Helping
One of the most important questions is not simply whether people help, but whom they help. Human prosociality is often structured by social identity. Family membership, nationality, ethnicity, political identity, school affiliation, neighborhood, occupation, and religion can all create distinctions between in-groups and out-groups. Religious prosociality therefore has to be studied within the broader psychology of group membership.
Research frequently finds stronger prosocial effects toward religious in-groups or socially close recipients. The 2021 review by Tsang and colleagues described religious prosociality as often parochial: helping can be stronger toward people who share a religious identity. Yet parochiality is not equivalent to hostility, and it does not mean that religious people only help co-religionists. Tsang et al., 2021.
A systematic review focused specifically on giving to out-groups and secular organizations examined 77 studies and found that religious involvement was most often positively related to external giving, while negative and null findings also occurred. The review identified religious norms and identity, social networks and attendance, and private practices as recurring pathways, with meaningful variation by context, tradition, measurement, and recipient. Yasin, Adams, and King, 2020.
These findings fit together. In-group favoritism and out-group generosity can coexist. A religious community can strengthen bonds among members while also supporting service beyond the community. The size and direction of either tendency depend on the group's teachings, local norms, intergroup relations, opportunities for contact, social status, and the specific behavior being measured.
Charitable Giving and Volunteering
Charity and volunteering are among the clearest real-world domains in which religious involvement and prosocial behavior intersect. Religious communities routinely create organized channels for giving money and time, and service attendance can expose participants to repeated requests, visible needs, collective projects, and models of helping.
The systematic review of out-group and secular giving found predominantly positive associations across the literature but also substantial heterogeneity. The New Zealand longitudinal study likewise estimated positive effects of service attendance on reported donations and volunteer time, while emphasizing that model-based causal effects were considerably smaller than raw cross-sectional differences. Yasin et al., 2020; Bulbulia et al., 2024.
A key interpretive point is that religious giving and secular giving are not mutually exclusive categories. A religiously affiliated organization can provide food, education, health services, disaster aid, or community support to people outside the faith community. Conversely, a religious individual can donate to a secular organization. Research therefore needs to identify the actual recipient and purpose rather than infer social reach from an organization's label alone.
Cooperation, Trust, and Fairness
Prosocial behavior also includes cooperation, reciprocity, trust, and fairness. Experimental economics has supplied many of the field's common tools, including dictator games, trust games, public-goods games, and allocation tasks. These methods are valuable because they observe choices under controlled conditions, yet each task simplifies real social life.
Anonymous one-shot games remove long-term relationships, shared history, reputational stakes, community institutions, and many of the meanings that accompany everyday generosity. That makes them useful for isolating mechanisms, but it also means that a weak laboratory effect cannot be translated directly into a claim that religion has little role in real-world cooperation. The reverse is equally important: a strong survey association cannot be assumed to represent spontaneous behavior in an anonymous setting.
Recent global research using experimentally validated measures has reported positive associations between religious identification and prosocial preferences such as altruism, trust, and positive reciprocity, while also finding cross-national and cross-tradition heterogeneity. Such studies expand the cultural range of the literature but remain observational at the level of religious identification. Kelly et al., 2024 provides the broader meta-analytic context for interpreting these associations.
Religious Priming: An Important but Contested Evidence Base
Some experiments attempt to test causality by briefly making religious concepts more mentally accessible before measuring behavior. This is called religious priming. The logic is that if exposure to religious words, settings, or cues changes subsequent helping or generosity under random assignment, the cue itself may have a causal effect.
A 2016 meta-analysis of 93 studies and 11,653 participants concluded that religious priming showed effects across several outcomes, including prosocial measures, and that effects were not reliable among nonreligious participants. Shariff, Willard, Andersen, and Norenzayan, 2016.
The literature then became a prominent case study in replication and publication-bias debates. A critical reanalysis found that different bias-correction methods produced conflicting conclusions, and it argued that large preregistered replications were necessary. van Elk et al., 2015 A preregistered replication with 650 participants did not find a statistically significant effect of implicit religious primes on dictator-game allocations. Gomes and McCullough, 2015.
The appropriate scientific conclusion is that religious priming remains methodologically contested. Priming studies are informative about the possible short-term activation of religious concepts, but they should not carry the main burden of claims about religion's effects on everyday generosity, moral character, or long-term social behavior.
Culture and Social Context Change the Relationship
Religion is embedded in culture. The psychological meaning of being religious differs when religion is socially expected, socially optional, a majority identity, a minority identity, inherited mainly as culture, or maintained through intense personal commitment. The same attendance frequency can therefore represent different motivations and social experiences in different places.
An analysis across more than 70 countries found that associations between personal religiosity and several prosocial or moral outcomes depended on the degree to which religiosity was socially enforced in the surrounding country. In more strongly religious social environments, differences between religious and nonreligious individuals on several outcomes weakened. Stavrova and Siegers, 2014.
This kind of moderation cautions against universal rankings of religious traditions or populations. Results from one country, one denomination, one age group, or one religious minority cannot be generalized to 'religion' as a whole. Minority status, discrimination, socioeconomic conditions, political history, institutional trust, and local opportunities for civic participation can all alter observed relationships.
Why Attendance Often Looks More Predictive Than Belief Alone
Across several lines of evidence, communal attendance often has a more consistent relationship with volunteering and social participation than affiliation or belief considered alone. A plausible explanation is that attendance bundles multiple mechanisms together: repeated social contact, exposure to norms, requests for help, observation of others' behavior, reputational accountability, shared projects, access to information, and practical opportunities to contribute.
This does not reduce religious participation to a social club. For a believer, worship and communal practice can carry sacred, theological, moral, and existential meaning. Psychology can recognize that meaning while also studying the social mechanisms that accompany regular gathering. The religious meaning and the social mechanism can operate at the same time.
What the Research Does Not Establish
The evidence does not establish that religious people are uniformly more moral, generous, empathic, or cooperative than nonreligious people. Group averages overlap extensively. A small average correlation cannot classify an individual person's character.
The evidence also does not establish that religion is required for prosociality. Nonreligious people help, cooperate, volunteer, donate, care for family members, participate in mutual aid, and act on moral commitments. Secular institutions, civic identities, ethical commitments, empathy, reciprocity, reputation, social norms, and human relationships can support the same broad class of prosocial behaviors.
Nor does a positive association show that every religious belief or practice has the same psychological effect. Religiosity is multidimensional, and religious traditions contain different moral vocabularies, institutional structures, practices, histories, and community norms. Scientific accuracy requires studying those differences rather than treating religion as a single exposure.
Finally, research on prosocial behavior does not test whether religious teachings are spiritually true. Psychological methods can examine behavior, cognition, emotion, identity, social networks, perceived sacred meaning, and reported experience. Questions of theological truth belong to other domains of inquiry.
Alternative Explanations and Confounding
Religious involvement is not randomly distributed. People who attend services frequently may differ from nonattenders in age, family structure, personality, health, migration history, neighborhood, social support, socioeconomic position, political values, or prior civic involvement. Some of these factors can also predict helping and volunteering.
Reverse causation is another possibility. People who are already community-oriented may be more likely to join or remain active in religious groups. Generous people may receive more invitations into social networks. Families with strong volunteering traditions may transmit both religious participation and civic engagement to children.
Longitudinal designs, fixed-effects models, causal diagrams, weighting, and sensitivity analyses improve inference, but observational studies cannot guarantee that every relevant confounder has been measured. Randomized experiments solve some causal problems but often create others, especially weak realism and narrow outcome measures. The strongest interpretation therefore comes from convergence across methods rather than from treating one design as decisive.
Major Research Limitations
Measurement varies widely. Religiosity can mean affiliation, belief, attendance, identity, commitment, or practice, while prosociality can mean self-reported kindness, donation, volunteering, game behavior, trust, or observed cooperation.
Self-report can be shaped by memory, identity, values, and social desirability. Direct behavioral tasks can be too narrow to represent sustained real-world helping.
Many studies are correlational. Even longitudinal models depend on assumptions about confounding, measurement error, and temporal ordering.
Recipient identity matters. Helping a family member, a co-religionist, a stranger, a member of another religion, or a secular organization can involve different psychological processes.
Cultural generalization is limited. Findings from Western university samples, majority-faith populations, or one country should not be treated as universal.
Religious traditions and local communities are internally diverse. Aggregated labels can obscure theology, practice, socioeconomic conditions, minority status, and institutional structure.
Publication bias and replicability are especially important in the priming literature. Earlier positive experimental findings require interpretation alongside preregistered null replications and bias-sensitive reanalyses.
Practical Implications for Communities, Researchers, and Clinicians
For religious communities, the research suggests that prosocial outcomes are most plausibly supported when moral commitments are connected to concrete opportunities for service, stable social relationships, transparent norms, and concern that extends beyond immediate in-groups. The evidence does not supply a universal recipe, but it consistently points toward the importance of social context and organized participation.
For researchers, the next step is less about asking whether religion is simply 'good' or 'bad' for prosociality and more about specifying mechanisms, recipients, time scales, and cultural settings. Strong studies separate belief from attendance, distinguish self-report from observed behavior, identify whether recipients are in-group or out-group members, and make causal assumptions explicit.
For clinicians and other helping professionals, prosocial religious involvement can be relevant to a person's support network, volunteering, identity, values, and sources of meaning. These are contextual features rather than diagnostic markers. Religious participation, charitable behavior, strong faith, low faith, or nonreligious identity do not by themselves indicate mental health, personality pathology, or moral functioning.
Culturally competent psychological work asks how religion functions for this person in this community. It avoids assuming that religious commitment automatically produces social benefit, and it also avoids treating faith as a psychological defect. The relevant clinical questions concern lived meaning, relationships, functioning, support, conflict, distress, and the person's own values.
A Research-Based Summary
The most defensible current conclusion is contextual. Religiosity is associated with somewhat greater prosociality on average, and the relationship is especially visible in self-report, service attendance, charitable giving, and volunteering. Direct behavioral effects are smaller and less consistent. Religious communities can support prosocial behavior through norms, identity, social networks, opportunities, accountability, and shared practices. At the same time, recipient identity and cultural context influence where helping is directed.
The 2024 meta-analysis places the average relationship in the small range and shows a large measurement gap between self-report and directly observed behavior. Contemporary reviews add that religious prosociality can be in-group focused in some settings, while systematic evidence on external giving also documents substantial helping beyond religious boundaries. Longitudinal studies suggest modest positive effects of service attendance on volunteering and cooperation under specified assumptions. Kelly et al., 2024; Tsang et al., 2021; Yasin et al., 2020.
Psychology therefore treats religion and prosocial behavior as a relationship among values, identities, institutions, social networks, situations, and individual differences. The question is not whether one worldview owns human goodness. The scientific task is to understand when religious life becomes one of the contexts through which people help, cooperate, give, trust, and care for others.
Frequently Asked Questions
Does religion make people more prosocial?
On average, religiosity is associated with a small increase in prosociality, but the effect varies substantially across studies and is much stronger in self-report than in directly measured behavior. The 2024 meta-analysis reported r = .13 overall, r = .15 for self-reported prosociality, and r = .06 for directly measured behavior. Kelly et al., 2024.
Are religious people more altruistic than nonreligious people?
Research cannot support a simple ranking of individuals by worldview. Prosocial behavior and altruistic motivation are related but distinct, group averages overlap widely, and both religious and nonreligious people engage in generosity, volunteering, cooperation, and care. The evidence is about probabilistic associations and mechanisms, not moral classification of persons.
Why do self-report studies find stronger effects than behavioral studies?
Self-reports can reflect identity, moral ideals, memory, and social desirability, while laboratory tasks may strip away the relationships and organized opportunities through which real-world helping occurs. Both measurement types capture part of the phenomenon. Their difference is itself an important research finding.
Does religious prosociality only benefit the in-group?
No. In-group favoritism appears in parts of the literature, but systematic research also documents charitable giving and helping directed toward out-groups and secular organizations. The balance depends on norms, networks, recipient identity, culture, and the specific religious community. Yasin et al., 2020.
Does attending religious services matter more than private belief?
Attendance often shows a more consistent relationship with volunteering and some cooperative outcomes because it includes repeated social contact, opportunities to help, information about needs, and community organization. That pattern does not mean private belief is unimportant; it means the variables capture different psychological and social processes.
Do religious primes reliably increase generosity?
The evidence is contested. An influential meta-analysis reported positive religious-priming effects, while bias-sensitive reanalysis and preregistered replication work raised substantial concerns. Religious priming should therefore be treated as a debated experimental literature rather than settled proof of a general causal effect. Shariff et al., 2016; van Elk et al., 2015; Gomes and McCullough, 2015.
Does psychology evaluate whether religious beliefs are true?
This research does not adjudicate theological truth. It studies human behavior, cognition, emotion, identity, social relationships, institutions, and meaning in religious contexts. The psychological reality of a person's values and experiences can be studied respectfully without turning scientific methods into a test of sacred claims.