RESEARCH DOSSIER
Social Trust, Reciprocity, and Reputation
Abstract. Social trust, reciprocity, and reputation organize how people accept interpersonal vulnerability, respond to others' conduct, and infer reliability from direct experience and socially transmitted information. This fundamental research monograph offers a critical narrative synthesis of social-psychological, behavioral-economic, cross-cultural, organizational, and formal research. It traces the development of trust measurement, investment games, indirect reciprocity and reputation models; compares competing explanations based on beliefs, fairness, social preferences and intentions; and evaluates experimental manipulations, meta-analyses, network models and naturalistic evidence. The review differentiates expectation, behavioral reliance, trustworthiness, reciprocal return and reputational influence, emphasizing the causal and measurement limits of any single task. Evidence supports conditional trust learning from experienced reciprocation, reputational effects on partner selection, and context-dependent responses to sanctioning, while revealing important heterogeneity across incentives, societies and information systems. The synthesis identifies research priorities for longitudinal learning, reliable reputational transmission, psychometric validation, field generalization and calibrated human–artificial reliance.
Keywords: social trust; reciprocity; reputation; interpersonal trust; trustworthiness; trust game; investment game; generalized trust; direct reciprocity; indirect reciprocity; reputational learning; trust calibration; betrayal aversion; gossip; partner choice; social exchange; cooperation; social norms; cross-cultural psychology
Scientific Scope and Research Questions
Social trust is an expectation that another person will act in ways sufficiently reliable, competent, or benevolent to justify accepting vulnerability in a particular interaction. Reciprocity is the patterned response to another's earlier action, including the return of help, the withdrawal of cooperation, and negative responses to perceived exploitation. Reputation is socially available information about a person's prior conduct or character that affects subsequent judgments and opportunities. Together these processes regulate how people rely on others, interpret exchange, and choose partners under incomplete information. The central scientific problem concerns the transformation of uncertain expectations into observable social behavior through interaction and information. The integrative organizational framework of Mayer, Davis, and Schoorman distinguishes perceived trustworthiness from willingness to accept vulnerability; the cross-disciplinary account by Rousseau and colleagues locates trust in expectations under conditions of risk and interdependence Mayer et al. (1995) Rousseau et al. (1998).
This monograph asks how trust beliefs originate, how they guide the choice to rely on an identifiable partner, how reciprocal acts change expectations, how reputational reports travel through social networks, and when these processes sustain or destabilize cooperation. It examines interpersonal and generalized trust, direct and indirect reciprocity, third-party observation, sanctions, repeated experience, and the measurement of each construct. Its evidentiary basis is a critical narrative synthesis of foundational theory, behavioral experiments, observational and cross-cultural studies, experimental games, methodological research, quantitative reviews, and research available through October 2026. The account preserves a distinction between causal experimental manipulations and descriptive associations. It does not claim a new experiment, an exhaustive systematic search, or an original pooled effect estimate Johnson & Mislin (2011) Duncan et al. (2026).
The designated research object is the psychological regulation of reliance through beliefs about others' conduct and socially transmitted histories. Strategic public-goods allocations, the psychology of distributive inequality, political legitimacy, organizational contracts, intimate attachment, and calibration of trust in artificial agents each have broader independent research objects. Evidence from those fields is examined here when it tests mechanisms of interpersonal trust, reciprocity, or reputational learning. The neighboring dossier on social preferences explains outcome valuation, while this study follows the temporal chain from prior conduct to expectation, vulnerability, response, and future partner choice. That organization enables a single canonical scientific identity across Social Psychology and Consumer and Economic Psychology Fehr & Schmidt (1999) Balliet & Van Lange (2013).
Historical Formation of the Research Tradition
The experimental study of interpersonal trust developed through social psychology, social learning theory, economic exchange, and formal game theory. Rotter's 1967 interpersonal trust scale established a broad psychometric approach to expectations about the reliability of other people and social promises. Later theoretical syntheses treated trust as a decision concerning another actor in a situation of consequential vulnerability. The historical importance of the Rotter scale lies in the measurement problem it created: a general expectation can be reported in a questionnaire, but the performance of a specific trust decision depends on the partner, incentives, stakes, and information provided. Research must therefore establish which level of expectation a particular instrument actually captures Rotter (1967) Mayer et al. (1995).
A pivotal experimental development was the investment or trust game presented by Berg, Dickhaut, and McCabe in 1995. A first mover could transfer resources to an unknown second mover; the experiment increased the amount received, allowing the second mover to return some of the proceeds. Transfers represented an observable decision to expose resources to another person's discretion, while returns operationalized reciprocation. The investigators designed the setting to limit contractual enforcement and repeated-game reputation. Their contribution was a tractable experimental architecture for studying two interdependent actions rather than a direct test of every psychological motive underlying either one Berg et al. (1995).
The late twentieth century also brought formal accounts of reputation in indirect reciprocity. Nowak and Sigmund's 1998 image-scoring model showed how helping someone could be favored when observers later directed cooperation toward those seen as generous. Experimental work by Wedekind and Milinski then investigated whether people actually favor recipients with more generous histories. These developments established an important shift from bilateral exchange to networked social information: a person's future treatment could depend on how they had behaved toward third parties they might never encounter again. Subsequent studies of gossip and group selection of partners examined the transmission mechanisms through which this information becomes socially consequential Nowak & Sigmund (1998) Wedekind & Milinski (2000) Sommerfeld et al. (2007).
Trust, Trustworthiness, Reliance, and Cooperation
A trust belief is a judgment about the likely behavior of another actor. Trusting behavior is the decision to become vulnerable on the basis of that judgment and the value of the available alternatives. Trustworthiness refers to the partner's actual disposition or demonstrated pattern of fulfilling relevant expectations. Cooperation refers to behavior that advances another's interest or a joint outcome in a specified incentive structure. These variables may be related without being identical. An individual can believe a stranger is honest and nevertheless decline to transfer money because the stakes are high; another can make a transfer under pressure despite low confidence. Competence, integrity, and benevolence are distinguishable sources of perceived trustworthiness in Mayer and colleagues' model Mayer et al. (1995) Rousseau et al. (1998).
A generalized-trust measure addresses expectations about people across social settings, whereas a partner-specific measure asks whether a particular other will return a benefit, tell the truth, or fulfill a promise. State trust may change rapidly with a counterpart's behavior; dispositional trust is intended to characterize relatively stable expectations. Glanville and Paxton examined competing accounts of generalized trust, including extrapolation from localized experiences and broader predispositions, illustrating that measurement models already imply theories of origin. McAllister distinguished cognition-based and affect-based interpersonal trust in organizational relationships; this division identifies different bases of confidence without requiring that every observed act be assigned to a single pure category Glanville & Paxton (2007) McAllister (1995).
Trust also has a distinct relation to risk. Bohnet and Zeckhauser experimentally compared decisions involving another person's discretionary conduct with structurally similar uncertain outcomes decided by chance. Their findings indicated that participants demanded more favorable prospects when the bad outcome could be produced by another person's betrayal. The psychological cost of intentional betrayal can therefore affect decisions beyond the monetary probability of loss. In interpreting a trust game, one must consider the first mover's beliefs, betrayal aversion, inequality concerns, altruistic preferences, and the social meaning of choosing. The same monetary transfer cannot uniquely identify all these processes Bohnet & Zeckhauser (2004) Fehr & Schmidt (1999).
Reciprocity as a Psychological Response to Social Conduct
Direct reciprocity occurs when an actor responds to how the same partner previously treated them. Positive reciprocity returns help, accommodation, or generosity; negative reciprocity responds to a perceived violation with withdrawal or sanction. Reciprocity may follow a conditional rule, an emotion of gratitude or anger, an expectation of future reward, or a norm about what one ought to return. Theoretical accounts differ over whether people evaluate outcomes, infer intentions, or integrate both. Falk and Fischbacher's formal theory treats assessments of kindness as dependent on the actor's intentions and available alternatives as well as material consequences Falk & Fischbacher (2006) Fehr & Gächter (2002).
A repeated transaction can sustain cooperation because each actor expects present behavior to influence the other's future choices. Yet reciprocity is also observed in designs intended to minimize future material return. Berg and colleagues found returns in anonymous exchanges; Fehr and Gächter found that some participants incurred costs to sanction low contributors even when such costs provided no direct material benefit. These results expand the explanatory space beyond narrow immediate material self-interest. They do not establish that all participants acted from identical motives, and costly punishment should be interpreted separately from generosity. Norm enforcement, anger, welfare concerns, strategic signaling, and experimental framing can each affect observed action Berg et al. (1995) Fehr & Gächter (2002) Jordan et al. (2016).
Conditional cooperation is particularly important for understanding reciprocal expectations in groups. Fischbacher, Gächter, and Fehr used a strategy-method public-goods design and reported pronounced heterogeneity, including conditional cooperators and freeriders, in the sampled participants. Keser and van Winden compared relatively stable and changing partners in repeated public-goods experiments, finding different trajectories of contribution. These results imply that an average cooperation rate can conceal distinct response functions. A scientifically adequate description estimates how a person's contribution responds to expected or observed contributions of others, and how that response changes when future interaction becomes more or less plausible Fischbacher et al. (2001) Keser & van Winden (2000).
Reputation as Information and Social Evaluation
Reputation is an inference about a socially identifiable actor built from observations, statements, records, and the judgments of others. It can summarize past generosity, fairness, honesty, competence, or group-norm conformity, and its evidentiary value depends on the behavior being predicted. A reputation for generosity in a donation task may predict helping under particular incentives while providing limited information about competence in a technical assignment. Reputational information also differs in provenance: direct observation, a trusted witness, an anonymous rating, and an aggregated score invite different estimates of credibility. The key mechanism is transfer of information about previous conduct into current partner choice, with the possibility of both accurate learning and systematic distortion Nowak & Sigmund (1998) Sommerfeld et al. (2007).
Indirect reciprocity becomes possible when help given to one individual changes how third parties treat the helper. Nowak and Sigmund modeled conditions under which observed generosity could be favored through image scoring. Wedekind and Milinski provided experimental evidence that donations were more frequent to recipients with generous prior histories. Milinski, Semmann, and Krambeck showed that linking public-goods interactions to an environment where reputation mattered could sustain contributions. The evidence concerns specific institutional and informational conditions; reputational incentives may fail where behavior is unobservable, records are unreliable, or reputations become disconnected from actual conduct Nowak & Sigmund (1998) Wedekind & Milinski (2000) Milinski et al. (2002).
Reputation is both an informational and a motivational variable. When observers use it to select partners, the reputational record changes the incentives for future behavior; actors may begin cooperating because they expect their present choice to affect access to valuable relationships. When reports travel through gossip, a third person can influence both the audience's beliefs and the target's opportunities. This creates a feedback process in which observations, interpretations, dissemination, responses, and later conduct alter one another. It also introduces risks of strategic exaggeration, selective disclosure, and social exclusion, which demand explicit analysis of the reliability of reputational channels Sommerfeld et al. (2007) Feinberg et al. (2014) Kawakatsu et al. (2024).
Measurement, Operationalization, and Construct Validity
Measurement begins by matching an instrument to a particular scientific question. A generalized trust scale measures a person's stated expectations about others. A vignette measures intended reliance given a description of a partner. An incentivized trust game records behavior under an experimental payoff structure. A repeated trust game measures how allocation changes with feedback across rounds. Each output has its own construct validity, temporal resolution, and potential confounds. Rotter's scale remains historically foundational, while the meta-analysis by Johnson and Mislin documents the sensitivity of trust-game behavior to game protocol and participant characteristics Rotter (1967) Johnson & Mislin (2011).
In the first-mover phase of an investment game, transfers can reflect expected reciprocity, risk attitudes, concern for the other, and beliefs about the game itself. Second-mover returns can reflect reciprocal motivation, distributional fairness, strategic anticipation, or experimental norms. The two measures should be named behavioral trust investment and behavioral return rather than treated as unambiguous psychometric observations of a single mental property. Johnson and Mislin synthesized 162 replications involving more than 23,000 participants and identified moderator effects associated with design choices. These findings make protocol reporting, payoff details, counterpart description, anonymity, and sample composition essential elements of scientific interpretation Johnson & Mislin (2011).
The 2026 meta-analysis by Duncan and colleagues specifically examined learning in repeated trust games, synthesizing 404 effect sizes from 68 studies and more than 8,000 participants. Partner reciprocation rate was a strong predictor of changing trust behavior. It also reported conditional effects of prior moral or trustworthiness information. Repeated games add a distinct longitudinal element: a person's response to observed returns can be separated, in principle, from an initial willingness to trust. Robust estimation requires attention to nonindependence of rounds, partner-specific history, feedback presentation, experimenter-controlled behavior, endgame expectations, and the reliability of the learning parameter Duncan et al. (2026).
The Investment Game and the Experimental Study of Reliance
The investment game is most informative when its decision sequence is described explicitly. An investor controls an initial endowment and may send an amount to a trustee; the trustee receives a transformed amount and may return part of it. Because the trustee can retain the gain, the investor must choose under another actor's discretionary control. The observed first transfer is a behavioral operationalization of reliance under uncertainty, and the return is an operationalization of repayment within the experimental incentive system. Neither variable alone reveals a subject's subjective probability of repayment. Berg, Dickhaut, and McCabe's original investigation isolated this problem in one-shot anonymous interaction, making trust and reciprocation accessible to experimental comparison Berg et al. (1995).
Treatment effects in trust games depend on what the experimenter changes. A researcher may vary information about the partner, the multiplier applied to transfers, the social proximity of the dyad, the source of initial wealth, the potential for punishment, or the visibility of decisions. Each manipulation tests a more specific causal proposition than the general statement that people trust others. For example, introducing a verifiable record of past reciprocation tests information-based updating; removing all future contact tests whether current behavior depends on anticipated reencounter. Johnson and Mislin's meta-analysis showed that seemingly technical decisions about protocol can alter the measured outcome, underscoring the need to preregister treatment contrasts and publish complete instructions Johnson & Mislin (2011).
An experimental transfer may also communicate generosity. If a first mover values the second mover's welfare, the transfer may occur even when the first mover expects little return. If a trustee dislikes unequal payoffs, a return may compensate for an inequitable distribution rather than express a felt obligation. Formal inequity aversion provides a plausible alternative account for some exchange outcomes. The resulting identification challenge is resolved by comparing game variants, eliciting expectations and preferences separately, and testing whether model parameters predict out-of-sample behavior. Behavioral games become more informative when deployed as linked instruments rather than treated as a singular transparent window into the psychology of trust Fehr & Schmidt (1999) Johnson & Mislin (2011) Thielmann et al. (2020).
Learning to Trust Across Repeated Interaction
Repeated encounters transform a decision under a static description into a learning process. The observer sees what a particular partner actually does, compares that outcome with an expectation, and adjusts future willingness to rely on them. Initial priors may come from generalized expectations, labels, recommendations, or a partner's group membership; subsequent direct feedback can confirm or overturn those priors. A useful description separates the initial probability assigned to reciprocation from the strength and direction of updating after each encounter. Because the actual sequence of returns is observable, experimenters can test learning against known contingencies and distinguish volatility from persistent partner reliability Duncan et al. (2026) Glanville & Paxton (2007).
Duncan, Sganzerla, Kaltwasser, and Dziobek's recent meta-analytic study is unusually direct evidence for this dynamic research object. Their synthesis of 68 studies and more than 8,000 participants identified a pronounced association between reciprocation rates and behavioral trust learning. Their moderator analyses also found that negative priors about trustworthiness or morality could affect updating, while partner group membership did not explain learning in the same way within the examined designs. Such findings suggest that observed behavioral reliability has particular importance when participants receive informative feedback. The inference remains bounded by task structure, the operational definition of learning, and the characteristics of the included experiments Duncan et al. (2026).
Observed investment may decrease late in a finitely repeated game even after a series of successful returns, because the expected future benefits of maintaining a relationship become smaller. Such endgame behavior is a change in strategic opportunity as well as a possible change in belief. A design that ignores the number of rounds remaining can misattribute a strategic decline to a loss of trust. Likewise, participants may adapt to a programmed partner differently than to a human who might infer their intentions. A strong learning design therefore models anticipated reciprocity, horizon, prior information, actual feedback, and the participant's understanding of the interaction as separable predictors Duncan et al. (2026) Keser & van Winden (2000).
Intentions, Fairness, and the Meaning of Return
Reciprocity is sensitive to the attribution of intention. The return of a benefit can express gratitude for voluntary generosity, routine compliance with a role, or exchange according to explicit obligation. When a partner had no reasonable alternative to the observed action, its social meaning differs from a freely chosen costly sacrifice. Falk and Fischbacher constructed a model in which kindness incorporates the consequences of an action and the intentions inferred from the options that were available. A psychological experiment that varies the partner's choice set while holding the outcome fixed can therefore test whether observers respond to intention rather than simply to their own payoff Falk & Fischbacher (2006).
Fairness models provide an intersecting but distinct interpretation. Fehr and Schmidt proposed that disadvantageous and advantageous inequality could shape decisions in bargaining and cooperation. Under such preferences, a participant may sacrifice some money to reduce a distribution considered unfair, independently of a belief that this will strengthen the relationship. Conversely, a person may repay an initially generous partner because they expect future mutual advantage. Distinguishing these explanations matters whenever researchers interpret a return as evidence of a stable prosocial disposition. The appropriate question concerns which changes to information, choice sets, incentives, and partner identity are necessary to distinguish the proposed mechanisms Fehr & Schmidt (1999) Falk & Fischbacher (2006).
Emotion contributes to exchange through appraisals of fairness, intention, gratitude, resentment, fear, and the anticipated consequences of betrayal. The experiment by Bohnet and Zeckhauser suggests that intentionally caused loss carries a psychological cost beyond an equally likely impersonal adverse outcome. Punishment studies likewise show that some people pay to sanction perceived free riding. These findings invite models in which trust and reciprocity involve both expectations about future behavior and the present meaning of being helped or wronged. The evidence warrants a multi-mechanism account rather than assuming that a single emotional impulse explains every return or sanction Bohnet & Zeckhauser (2004) Fehr & Gächter (2002).
Punishment, Sanctions, and Reputation for Norm Enforcement
Punishment can maintain cooperation by changing the cost of exploitation. Fehr and Gächter experimentally demonstrated that allowing costly punishment altered contributions in a public-goods setting. Their work establishes the causal relevance of an available sanction under specified experimental conditions; it does not establish that sanctioning has uniform effects across cultures, communities, or institutions. Meta-analytic work by Balliet, Mulder, and Van Lange on reward and punishment confirms that incentive interventions must be evaluated in relation to the tasks, their parameters, and the motives they recruit. A sustainable cooperative system can rely on mutual expectation, partner selection, sanctions, or several mechanisms acting together Fehr & Gächter (2002) Balliet et al. (2011).
Costly punishment can also function as a reputational signal. Jordan, Hoffman, Bloom, and Rand investigated when observers regard third-party punishers as trustworthy and whether those signals change when a more informative opportunity to help becomes available. Their findings indicate that a sanction can affect reputation because observers infer willingness to enforce a norm, but that inference depends on competing information. The same visible act may be interpreted as principled, aggressive, strategic, or normatively excessive in another setting. A complete explanation should therefore model the observer's inference as well as the punisher's behavior, the target's response, and the surrounding social rules Jordan et al. (2016).
Cross-societal synthesis complicates simple prescriptions about enforcement. Balliet and Van Lange examined trust, punishment, and cooperation across eighteen societies and identified contextual variation in how sanctioning influences contributions. Partner expectations and prevailing institutional conditions change the meaning of a punishment opportunity. An intervention that induces short-term compliance may also alter trust in the group, willingness to cooperate without enforcement, or perceptions of procedural justice. Therefore long-term outcome assessment should include changes in voluntary exchange and later reputation in addition to the immediate amount contributed after a sanction Balliet & Van Lange (2013) Balliet et al. (2011).
Gossip, Testimony, and Reputational Transmission
Most social decisions are made with incomplete direct observation of others. Gossip conveys information about third parties, and its psychological significance depends on the credibility of its source, the audience's prior beliefs, and the opportunities for verification. Sommerfeld and colleagues experimentally studied the role of gossip in indirect reciprocity games; their study with groups of students showed that communicated reputational assessments affected choices even when direct information was available. This is evidence that social testimony can have substantial behavioral force. It is also evidence that reputational information can be manipulated, making the accurate transfer of conduct a different question from the social effectiveness of a report Sommerfeld et al. (2007).
Feinberg and colleagues explored prosocial functions of transmitting reputational information. Experiments on gossip and ostracism showed that participants used communicated information to avoid uncooperative partners and that changes in partner selection could support contributions in mixed-motive settings. The central mechanism is selective association: reliable cooperators gain future opportunities, whereas exploitative behavior limits access to beneficial interactions. Such results should not be read as a general endorsement of exclusion. A reputational system may misclassify individuals, amplify initial errors, or penalize those who lack social standing, and the conditions of laboratory information exchange cannot represent every real network Feinberg et al. (2012) Feinberg et al. (2014).
Theoretical developments make information fidelity itself a research variable. Kawakatsu, Kessinger, and Plotkin modeled the conditions under which gossip can produce sufficient consensus about reputations to stabilize cooperation. Their analysis shows that errors, norms, and biases in communication affect the dynamics of indirect reciprocity. This complements direct experimental findings by specifying which assumptions about information transmission matter for the predicted outcome. It also distinguishes theoretical possibility from a field demonstration: a mathematical threshold for consensus is a model result until the relevant information and behavioral parameters have been observed or estimated in actual networks Kawakatsu et al. (2024) Sommerfeld et al. (2007).
Partner Choice and the Social Structure of Trust
Trust is expressed through the choice of whom to approach as well as how much to transfer. A reputation for repayment or helping increases a potential partner's attractiveness under conditions where exchange is voluntary and information can be shared. Selective partner choice changes the reward structure: the opportunity to form future beneficial relationships becomes contingent on earlier conduct. This mechanism explains why reputation can support exchange even when a specific dyad is unlikely to meet again. The prediction, however, depends on observers having alternatives, receiving informative histories, and assigning those histories to the correct individuals Nowak & Sigmund (1998) Wedekind & Milinski (2000) Feinberg et al. (2014).
Social networks distribute observation unevenly. Some actors are highly visible and others are encountered through a chain of intermediaries; reports may reach different audiences at different times. The reputation that guides a decision is therefore not a single public score but a locally available set of claims, observations, and inferences. Information can be redundant because several people repeat the same source, or apparently consistent because a rumor spreads through a tightly connected group. Contemporary mechanistic models of gossip illustrate that agreement across observers may depend on transmission dynamics as much as on the accuracy of the underlying observation. Research needs to track source independence, network position, and whether reported conduct occurred under comparable circumstances Kawakatsu et al. (2024) Sommerfeld et al. (2007).
Partner choice provides an explanatory bridge from interpersonal learning to social institutions. A market, team, or neighborhood may sustain repeated transactions because its members can evaluate prior conduct and prefer reliable partners. The available evidence establishes several behavioral and formal mechanisms through which this can happen. It leaves open how well laboratory reputational signals predict costly, durable trust in complex settings. Naturalistic records of who selects whom, under what informational conditions, and with what subsequent outcomes are essential complements to controlled games. The theoretical objective is to connect individual expectation updating to network opportunities without treating every cooperative aggregate as direct evidence of trust Milinski et al. (2002) Henrich et al. (2010).
Individual Differences and the Stability of Trust
People differ in their initial expectations, preferred level of vulnerability, sensitivity to betrayal, and inclination to return help. Generalized interpersonal trust captures part of this variation, while personality research identifies broader dispositions relevant to prosocial choices. Thielmann, Spadaro, and Balliet's meta-analysis organized personality predictors by the structure of social interdependence, emphasizing that the same trait may predict behavior differently when exploitation is possible, joint gains are accessible, or another person's need is salient. Thus a reliable measure of individual difference should explain variation across specified social situations rather than stand as a context-free psychological essence Rotter (1967) Thielmann et al. (2020).
Experimental evidence suggests some cross-task consistency. Peysakhovich, Nowak, and Rand studied decisions across different cooperative games and across time, reporting correlations compatible with a relatively stable propensity to incur costs that benefit others. Their investigation also separated cooperation from willingness to punish, showing the importance of avoiding an undifferentiated prosociality score. A person's cooperative disposition, baseline trust, and expectations about a specific partner are related possible causes of exchange but must be estimated separately. The empirical magnitude and persistence of any one disposition depend on the selection of tasks, measurement reliability, and repeated sampling of contexts Peysakhovich et al. (2014).
The origins of generalized trust remain difficult to establish from cross-sectional data. Stable family experiences, socialization, community institutions, repeated cooperative encounters, and personal expectations can influence one another over time. Glanville and Paxton explicitly compared different accounts of the relation between generalized trust and localized experience. A prospective design can observe the transition from childhood expectations into adolescent and adult exchange; a cross-cultural longitudinal design can further test whether the same questionnaire items carry comparable meaning across populations. Correlation between trusting beliefs and frequent cooperative activity is informative, but only a design addressing selection and reciprocal causation can show which tends to produce which Glanville & Paxton (2007) Rousseau et al. (1998).
Cultural Variation and the Limits of Generalization
Trust research began with many university-based convenience samples, making the transport of findings an empirical question. Henrich and colleagues conducted bargaining and allocation experiments across fifteen small-scale societies and found substantial variation in behavior that was partly associated with everyday economic organization and cooperative practices. The results challenge the assumption that identical economic incentives produce identical social decisions everywhere. However, ultimatum, dictator, and public-goods games often measure allocation, fairness, and willingness to cooperate rather than partner-specific trust itself. Their greatest value here is as evidence that the norms and expectations surrounding reciprocal exchange vary across social environments Henrich et al. (2005).
Henrich and colleagues' later work across diverse populations examined the relation between market integration, community size, religion, fairness, and punishment. These findings place experimental behavior within institutional and cultural settings, demonstrating that the meaning of an exchange depends partly on the organized forms of interaction people encounter in ordinary life. The observed covariation does not by itself establish the causal effect of market integration or religious affiliation. It does indicate that trust research seeking broad psychological explanations should measure everyday opportunities to cooperate, networks of accountability, and culturally specific expectations about strangers and familiar partners Henrich et al. (2010).
Differences across societies can reflect both real behavioral variation and differences in what a task communicates. The same anonymous transfer may signal altruism, tactical caution, friendliness, or adherence to a local norm depending on expectations about the situation. Johnson and Mislin's synthesis of investment games identified geographic and protocol variation, but geographical contrasts cannot be interpreted as intrinsic properties of entire populations. Comparative studies should test language equivalence, incentives relative to local conditions, recruitment, comprehension, experimenter effects, and the possibility that laboratory anonymity has different social significance in different settings Johnson & Mislin (2011) Henrich et al. (2005).
The Social Psychology of Organizations and Institutions
Organizations transform interpersonal trust by making roles, obligations, and records durable. A colleague's competence, another manager's integrity, and a department's treatment of commitments become evidence for specific forms of reliance. Mayer, Davis, and Schoorman's model distinguishes characteristics of the person who trusts, the attributes of the person trusted, and the role of perceived risk. McAllister's investigation of managers and professionals distinguished cognition-based and affect-based trust and explored their relation to cooperation. Taken together, these accounts show why personal familiarity can support relational confidence while professional competence supplies a different basis for relying on someone in high-stakes tasks Mayer et al. (1995) McAllister (1995).
An organization can also supply institutional assurances through transparent procedures, incentives, accountability, and dispute resolution. These mechanisms affect both actual trustworthiness and perceived willingness to accept risk. Kramer surveyed the organizational trust literature, including the conditions that build trust and the barriers that impede it. An institution's monitoring system may reduce the harm caused by opportunism, but its presence need not automatically increase interpersonal trust; people may comply because a rule is enforced. Researchers should therefore compare performance, expectation, satisfaction, and the willingness to rely when formal guarantees are withdrawn or altered Kramer (1999) Rousseau et al. (1998).
This dossier uses organizational studies as a test bed for general psychological mechanisms. The independent construct of the psychological contract concerns perceived reciprocal obligations between employee and employer; organizational culture concerns shared systems of practice; political legitimacy concerns judgments of governing authority. Social trust intersects with these phenomena through expectations of good faith, information about past fulfillment, and reactions to violations. The separation is analytically useful because institution-level satisfaction cannot be substituted for the probability that a particular partner will honor a specific reliance decision. Valid measurement must define the actor, action, vulnerability, time interval, and enforcement conditions Mayer et al. (1995) Kramer (1999).
Trust Across Markets, Communities, and Public Life
Market exchange places many people in contact with counterparties they have never met. Reliability is conveyed through the remembered history of a firm, reputational information from other buyers, explicit guarantees, and institutions that make claims or commitments verifiable. The basic mechanism is an estimate about future behavior formed from incomplete evidence. Social trust may facilitate mutually beneficial exchange by lowering the perceived danger of depending on another actor, but market activity itself is not a pure measure of trust: regulation, necessity, alternative options, and unequal bargaining power also shape transactions Rousseau et al. (1998) Henrich et al. (2010).
An instructive field experiment studied ordinary civic honesty through the return of apparently lost wallets. Cohn and colleagues implemented the study across countries and documented how the presence of money in a wallet related to return behavior. This design records an observable prosocial act in everyday settings rather than an expressed attitude in a survey. At the same time, returning a wallet is influenced by norms, self-image, perceived consequences, empathy, and situational interpretation. It establishes evidence about honesty under a particular naturalistic task; it should not be equated with a comprehensive measure of the entire country's interpersonal trust Cohn et al. (2019).
Neighborhoods and voluntary organizations add repeated contact and social networks to exchange. Reputation can become a local memory of promises kept, help delivered, or obligations neglected. It can also carry unequal weight across social positions: an already visible actor may receive more opportunities to demonstrate reliability than a newcomer. The resulting selection process complicates a simple association between reputation and ability. Sound community research compares the information available to different participants, the range of actual opportunities to cooperate, and the accuracy of later judgments relative to independently observable outcomes. The same principles apply to online marketplaces where reputational signals can be aggregated, curated, or strategically engineered Sommerfeld et al. (2007) Feinberg et al. (2014) Kawakatsu et al. (2024).
Causal Identification, Replication, and Evidence Quality
A causal claim about reputation requires a manipulation of information or its availability while holding relevant partner behavior constant. A causal claim about reciprocity requires comparing how a person's response changes after different prior acts or different intentions. A causal claim about trust learning requires exogenous variation in feedback, a sufficiently defined sequence, and analysis that distinguishes learning from horizon effects. Many influential game studies meet important parts of these requirements. Their findings are strongest for the precise experimental contrasts implemented and weaker for claims about unmeasured motives or behavior outside the laboratory Berg et al. (1995) Falk & Fischbacher (2006) Duncan et al. (2026).
Replication requires more than reusing a familiar game name. Endowment size, exchange multiplier, participant recruitment, one-shot versus repeated encounter, anonymity, instructions, payment procedure, treatment of disbelief, and availability of outside options may all change behavior. Johnson and Mislin's 2011 meta-analysis identified the importance of protocol variation in trust games. Duncan and colleagues' 2026 review provides a complementary examination of repeated trust learning, using its own synthesis of the relevant published literature. A cumulative evidence base should maintain distinct estimands for initial trusting behavior, conditional returns, learning slopes, reputational influence, and actual long-term reliability Johnson & Mislin (2011) Duncan et al. (2026).
The quality of synthesis depends on sample diversity, documented search and inclusion methods, risk-of-bias assessment, handling of multiple effects from the same study, publication selection, and the precision of moderators. A conceptual narrative review like this monograph synthesizes findings across methods and clarifies theoretical links; it does not justify a pooled quantitative estimate without an explicit reproducible protocol. Future quantitative reviews can improve upon fragmented evidence by combining preregistered inclusion criteria, hierarchical models of dependent observations, open analysis code, and assessment of publication bias. Strong inference also requires reporting contradictory results and design-dependent reversals rather than selecting only the conditions where cooperation appears most pronounced Balliet & Van Lange (2013) Johnson & Mislin (2011).
Neural and Cognitive Accounts of Social Evaluation
A social trust decision demands attention to another actor, memory for the relevant history, inference about intention, evaluation of possible outcomes, and control of an action under uncertainty. These functional processes provide a bridge from social psychology to cognitive and affective neuroscience. Yet a neural response during a trust game is best interpreted as evidence about processing under the specified task, rather than a unique anatomical indicator of trust. The cognitive content of the experiment must first be identified: a person may be calculating expected repayment, anticipating betrayal, or choosing an action that signals generosity Bohnet & Zeckhauser (2004) Mayer et al. (1995).
A psychologically precise account distinguishes the confidence of an expectation from the value attached to its potential violation. A betrayal-sensitive participant may decline to trust despite a substantial estimated probability of repayment; another participant may accept the same uncertainty because the relationship is valuable. The research program therefore benefits from measuring stated probabilities, affective responses, perceived norms, and trial-level investments together. These observations can constrain computational models of belief updating, preference weights, and response selection. Without such constraints, several models may reproduce the same observed transfer distribution while making divergent predictions when intentions or source reliability are manipulated Bohnet & Zeckhauser (2004) Falk & Fischbacher (2006).
Research on spontaneous and reflective cooperative choice further suggests that cognitive processing conditions matter. Rand, Greene, and Nowak experimentally examined the relation between fast decisions and cooperation, contributing to an influential debate about how social heuristics operate. That study concerns a particular family of cooperation tasks, and its findings should not be generalized into a claim that all trusting action is automatic or all deliberation reduces social commitment. Cognitive accounts should state the class of interactions, experimentally manipulated processing conditions, participant learning history, and outcome being predicted. The goal is a mechanism that explains when and why cognition changes reliance, not a single global characterization of human sociability Rand et al. (2012) Kiyonari et al. (2000).
Trust Repair, Betrayal, and the Dynamics of Reassessment
A breach of trust changes the informational basis of reliance. A person who had expected a partner to return a benefit may now assign a lower probability to future cooperation, experience an emotional cost of having been betrayed, and reconsider whether to continue the relationship. These are three distinct outcomes: revised belief, affective response, and behavioral withdrawal. An apology, explanation, restitution, or an observed pattern of future reliability may affect each outcome differently. Experimental work separating betrayal from impersonal risk indicates why a violation by another person may matter beyond its material consequences Bohnet & Zeckhauser (2004) Mayer et al. (1995).
Trust repair can be studied prospectively by observing the same dyads before a violation, immediately afterward, and through later opportunities to cooperate. Such a design must distinguish public information about wrongdoing from direct personal experience and compare actual behavioral change with self-reported forgiveness or confidence. A recipient might report improved feelings while remaining unwilling to delegate meaningful resources; another may cooperate for practical reasons while preserving distrust. Models of trust that combine ability, benevolence, integrity, and risk highlight why a successful corrective action should address the kind of evidence that was undermined by the violation Mayer et al. (1995) McAllister (1995).
Reputational damage creates a further process because different observers may update from incomplete, contradictory, or false reports. Individuals who receive a negative signal through gossip may treat the subject differently even without direct interaction. A public correction can increase the availability of counterevidence without guaranteeing that the original narrative disappears from the network. Mechanistic research on reputational consensus shows why the timing, fidelity, and directional bias of information transfer matter. The central open problem is to identify when corrective experience reliably changes social beliefs, and when network structures preserve earlier errors Sommerfeld et al. (2007) Kawakatsu et al. (2024).
Digital Reputation and Platform-Mediated Social Exchange
Online platforms can make previous conduct visible through feedback scores, transaction histories, verified accounts, and written reviews. These systems alter the material that people use to estimate reliability, but their measurement categories also define what counts as publicly legible reputation. A numerical score may summarize satisfaction, speed, honesty, communication quality, or compliance with a platform's policies. A scientifically responsible analysis asks which underlying behavior produced the measure, who could rate whom, whether negative experiences were reported, and how the interface changed selection among potential partners. The core psychological processes of observation, inference, and reliance continue, while the institutional architecture shapes the evidence people encounter Mayer et al. (1995) Kawakatsu et al. (2024).
Reputation mechanisms can create incentives for reliable exchange when future access to partners depends on past behavior. The experimental literature on image scoring, indirect reciprocity, gossip, and partner choice provides plausible mechanisms through which digital histories might influence cooperation. Transferring those mechanisms to platforms requires additional evidence because ratings may be purchased, manipulated, suppressed, or generated after a selection process that filters out unsuccessful interactions. The relationship between platform score and true conduct is therefore an empirical calibration problem. Valid field analysis should compare score histories with independently observed fulfillment, account for selective review participation, and consider strategic responses to public evaluation Nowak & Sigmund (1998) Feinberg et al. (2014) Kawakatsu et al. (2024).
Digital records also change the persistence and reach of reputational judgments. An ordinary interpersonal evaluation may remain within a small network, while an online account can expose a single disputed event to thousands of observers. Such systems can make previously invisible cooperation recognizable, but they can also spread misleading allegations or amplify preexisting inequality in who is judged credible. A legitimate research program separates the predictive accuracy of a reputation from its social power. It asks whether public metrics improve the calibration of trust, whether correction mechanisms work, and how policy rules influence both cooperation and the possibility of reputational recovery Sommerfeld et al. (2007) Kramer (1999).
Human–Artificial Relations as a Comparative Research Horizon
In the Artificial Era, people increasingly rely on nonbiological systems for information, recommendations, communication, and delegated actions. These encounters create a new comparative setting for established questions about reliability and calibrated reliance. A user may evaluate a system's consistency across conversations, its record of correct predictions, its transparent handling of errors, and the reliability of the organization operating it. Such judgments can be measured with repeated tasks, source-quality manipulations, and prospective outcome assessment. The psychological question concerns what evidence causes a human to rely on a particular system, how confidence changes with performance, and whether the resulting decision is appropriately calibrated to real capabilities Mayer et al. (1995) Duncan et al. (2026).
Trust in an artificial partner requires specifying what plays the role of actor, decision, vulnerability, and reciprocation. An artificial system can provide a useful response, maintain or fail to maintain interactional consistency, and change outputs according to previously observed information. These observable functions make functional comparison possible. Human reciprocal obligation, moral evaluation of intention, and experienced interpersonal betrayal have additional empirical and conceptual requirements; their transfer to artificial systems cannot be assumed from a conversational surface. The comparative study should characterize actual system behavior and human judgments separately, with documented error rates, controllability, and changes over time Bohnet & Zeckhauser (2004) Falk & Fischbacher (2006).
The broader perspective of Artificial Era · From Homo to Artificial makes this comparison relevant to the architecture of psychological research. Social trust can be analyzed in relation to human-human relationships, interactions with artificial agents, and institutionally mediated systems without collapsing the distinct objects. The operational program tests whether feedback about artificial performance generates learning curves comparable in form to human trust learning, whether disclosed information improves calibration, and which features of reputational systems encourage accurate reliance. These are research questions requiring controlled human-system studies and independently verified outcomes. The existing empirical literature on interpersonal trust supplies experimental concepts, not a ready-made conclusion about artificial psychological experience Duncan et al. (2026) Johnson & Mislin (2011).
The Evidentiary Status of Major Explanations
Different theories explain different components of the trust–reciprocity–reputation system. A social-learning perspective emphasizes expectations formed through experience. A relational perspective describes trust as willingness to accept vulnerability. Social-preference theories specify how fairness and concern for others enter a choice. Reciprocity theories incorporate the meaning of another person's intentions and actions. Indirect-reciprocity models establish how reputational information can sustain cooperation without repeated interaction between the same partners. These are complementary explanations only when their constructs and predictions remain distinguishable; disagreement arises when a behavioral measure is asked to identify a mechanism it cannot isolate Rotter (1967) Mayer et al. (1995) Fehr & Schmidt (1999) Falk & Fischbacher (2006) Nowak & Sigmund (1998).
The empirical record supports several comparatively secure propositions. Anonymous participants sometimes transfer resources and return benefits; cooperation responds to conditional expectations and opportunities for sanction; others' histories influence charitable allocations and partner choice; communicated reputational assessments affect social decisions; and observed reciprocation changes behavior in repeated exchange. The supporting evidence spans original experiments, computational models, and meta-analytic syntheses. Its strength differs by proposition. A randomized manipulation identifies a task-specific causal contrast, a mathematical model identifies logical possibilities under its assumptions, and a quantitative review assesses patterns and heterogeneity across previously published investigations Berg et al. (1995) Fehr & Gächter (2002) Wedekind & Milinski (2000) Sommerfeld et al. (2007) Duncan et al. (2026).
The same literature also contains clear boundaries. Laboratory tokens are not identical to consequential personal vulnerability; generosity in one allocation game does not identify a permanent moral character; geographic variation in experimental play cannot be interpreted as an innate national trait; and agreement about an individual's reputation may reflect information cascades as well as accurate independent observation. The most defensible synthesis therefore preserves the chain linking information, belief, motivation, decision, observed conduct, reputational transmission, and later opportunity. Each link can be tested with different methods, but the full chain requires longitudinal and network evidence rather than an isolated one-shot game Johnson & Mislin (2011) Henrich et al. (2005) Kawakatsu et al. (2024).
An Integrative Framework for Research on Reliance and Reputation
A cumulative research program can describe any episode of social trust through seven observable or estimable elements: the actor who may rely, the target of reliance, the task or promise at stake, the actor's evidence about the target, the cost of a mistaken decision, the behavioral action taken, and the feedback subsequently observed. Where a reputation exists, researchers additionally record the origin, content, transmission path, and credibility of social information. This framework clarifies what a questionnaire, game, field experiment, or network observation measures. It is a synthesis of existing construct distinctions and methodological requirements rather than a newly validated psychological scale Mayer et al. (1995) Rousseau et al. (1998) Johnson & Mislin (2011).
The temporal organization is crucial. Prior direct or indirect information informs an expectation; the actor considers the value and risk of relying; an action creates exposure to the target's behavior; an outcome generates cognitive and emotional consequences; the observer may revise the expectation; the action can become reputational evidence for third parties. These events may occur simultaneously in complex relationships, but their conceptual distinction enables targeted experiments. The researcher can manipulate the reliability of a report, the target's actual reciprocation, the actor's ability to exit, or the prospect of subsequent interaction while observing how each affects reliance and later cooperation Duncan et al. (2026) Sommerfeld et al. (2007) Milinski et al. (2002).
At higher levels of analysis, network structure and institutions determine how many opportunities an actor has to observe others, how costly dishonesty becomes, and how widely performance is known. Cultural practices shape what different exchanges mean; organizational rules define what counts as a fulfilled commitment; platforms transform events into reputational scores. These conditions form part of the mechanism rather than mere background variables. The strongest research integrates behavioral experiments with longitudinal exchange records, qualitative reconstruction of local expectations, and formal models whose parameters can be connected to measured conduct Henrich et al. (2010) Kramer (1999) Kawakatsu et al. (2024).
Unresolved Questions and Priorities for Future Research
The first major problem is construct identification. A trusting transfer, a stated general expectation, an allocation regarded as fair, and an act of benevolence can coincide, but they measure different aspects of social behavior. Research should collect converging evidence without assuming that convergence is guaranteed. Pre-registered multi-method studies can test whether the same person's reported expectations predict costly reliance across repeated contexts, whether behavioral trust and trustee reliability have stable but separable latent structures, and which payoff manipulations distinguish belief from preference. Hierarchical modeling can accommodate variation within people, across partners, and between environments Rotter (1967) Johnson & Mislin (2011) Thielmann et al. (2020).
A second challenge is reputational accuracy. Researchers need designs where actual conduct is recorded independently of the reputation later communicated about it. Network experiments can vary the precision, latency, and independence of reports, while field studies can examine whether ratings predict future commitments after accounting for selective exposure. Models of gossip predict that biased information transmission can either stabilize or undermine cooperation; empirical work should estimate the circumstances under which those model mechanisms occur. Information correction, the evolution of reputations after false allegations, and inequalities in who receives a credible hearing are central unanswered problems Sommerfeld et al. (2007) Kawakatsu et al. (2024).
A third challenge is external validity across culture, time, and technological mediation. Replicated tasks should be adapted and validated locally, while preserving transparent records of incentives, translation, and social context. Longitudinal studies are necessary to determine whether repeated reliable interaction builds generalized trust or whether people with higher baseline expectations preferentially enter cooperative networks. Human–artificial interaction research should use verified performance records to study reliance and updating, with clear separation between people's attributions and the observed capacities of systems. These steps allow psychological theory to expand its explanatory reach without losing its empirical anchors Henrich et al. (2005) Glanville & Paxton (2007) Duncan et al. (2026).
Conclusion
Social trust, reciprocity, and reputation form an interconnected system of psychological and social processes through which people manage vulnerability under incomplete information. Trust concerns expectations and the choice to rely; reciprocity concerns how actions respond to earlier treatment; reputation concerns how information about conduct becomes available to others and shapes future interactions. The system can encourage cooperation when the information is reasonably accurate, reciprocal expectations are credible, and reliable behavior creates opportunities for continued exchange. It can also amplify errors when information is distorted, incentives change, or reputational judgments persist beyond their evidentiary basis Mayer et al. (1995) Nowak & Sigmund (1998) Kawakatsu et al. (2024).
The experimental literature establishes important mechanisms through controlled investment games, sanctions, image scoring, gossip interventions, and repeated exchange. Quantitative synthesis shows that experimental protocols affect the behavioral measures often called trust and that observed partner reciprocation is central to how trust changes with experience. Cross-cultural research shows that everyday institutions and norms shape behavior in economic games, placing limits on context-free generalization. These results warrant a connected, multi-level explanation in which beliefs, preferences, social histories, information transmission, and incentives are measured separately and related through explicit causal hypotheses Berg et al. (1995) Johnson & Mislin (2011) Duncan et al. (2026) Henrich et al. (2010).
The durable scientific agenda is to observe the full life history of reliance: what an actor knew, how a decision exposed them to another, what the partner did, how the actor updated, and how that conduct subsequently circulated through the social environment. Combining experimental control with naturalistic outcomes, longitudinal learning measures, transparent psychometrics, and formal network models is the most promising route to cumulative knowledge. The expanding Artificial Era adds new forms of mediated and human–artificial reliance to this program while making accurate operationalization and accountable evidence even more important. In every context, a valid explanation begins with identifiable actors, meaningful uncertainty, observable conduct, and traceable information Duncan et al. (2026) Sommerfeld et al. (2007) Rousseau et al. (1998).
References
- Balliet, D., Mulder, L. B., & Van Lange, P. A. M. (2011). Reward, punishment, and cooperation: A meta-analysis. Psychological Bulletin, 137(4), 594–615. https://doi.org/10.1037/a0023489
- Balliet, D., & Van Lange, P. A. M. (2013). Trust, conflict, and cooperation: A meta-analysis. Psychological Bulletin, 139(5), 1090–1112. https://doi.org/10.1037/a0030939
- Balliet, D., & Van Lange, P. A. M. (2013). Trust, punishment, and cooperation across 18 societies: A meta-analysis. Perspectives on Psychological Science, 8(4), 363–379. https://doi.org/10.1177/1745691613488533
- Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social history. Games and Economic Behavior, 10(1), 122–142. https://doi.org/10.1006/game.1995.1027
- Bohnet, I., & Zeckhauser, R. (2004). Trust, risk and betrayal. Journal of Economic Behavior & Organization, 55(4), 467–484. https://doi.org/10.1016/j.jebo.2003.11.004
- Cohn, A., Maréchal, M. A., Tannenbaum, D., & Zünd, C. L. (2019). Civic honesty around the globe. Science, 365(6448), 70–73. https://doi.org/10.1126/science.aau8712
- Duncan, C., Sganzerla, L., Kaltwasser, L., & Dziobek, I. (2026). Trust learning in the repeated trust game: A meta-analytic study. British Journal of Psychology, 117(3), 1026–1055. https://doi.org/10.1111/bjop.70045
- Falk, A., & Fischbacher, U. (2006). A theory of reciprocity. Games and Economic Behavior, 54(2), 293–315. https://doi.org/10.1016/j.geb.2005.03.001
- Fehr, E., & Gächter, S. (2002). Altruistic punishment in humans. Nature, 415, 137–140. https://doi.org/10.1038/415137a
- Fehr, E., & Schmidt, K. M. (1999). A theory of fairness, competition, and cooperation. The Quarterly Journal of Economics, 114(3), 817–868. https://doi.org/10.1162/003355399556151
- Feinberg, M., Willer, R., Stellar, J., & Keltner, D. (2012). The virtues of gossip: Reputational information sharing as prosocial behavior. Journal of Personality and Social Psychology, 102(5), 1015–1030. https://doi.org/10.1037/a0026650
- Feinberg, M., Willer, R., & Schultz, M. (2014). Gossip and ostracism promote cooperation in groups. Psychological Science, 25(3), 656–664. https://doi.org/10.1177/0956797613510184
- Fischbacher, U., Gächter, S., & Fehr, E. (2001). Are people conditionally cooperative? Evidence from a public goods experiment. Economics Letters, 71(3), 397–404. https://doi.org/10.1016/S0165-1765(01)00394-9
- Glanville, J. L., & Paxton, P. (2007). How do we learn to trust? A confirmatory tetrad analysis of the sources of generalized trust. Social Psychology Quarterly, 70(3), 230–242. https://doi.org/10.1177/019027250707000303
- Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., McElreath, R., Alvard, M., Barr, A., Ensminger, J., Henrich, N. S., Hill, K., Gil-White, F., Gurven, M., Marlowe, F. W., Patton, J. Q., & Tracer, D. (2005). “Economic man” in cross-cultural perspective: Behavioral experiments in 15 small-scale societies. Behavioral and Brain Sciences, 28(6), 795–815. https://doi.org/10.1017/S0140525X05000142
- Henrich, J., Ensminger, J., McElreath, R., Barr, A., Barrett, C., Bolyanatz, A., Cardenas, J. C., Gurven, M., Gwako, E., Henrich, N., Lesorogol, C., Marlowe, F., Tracer, D., & Ziker, J. (2010). Markets, religion, community size, and the evolution of fairness and punishment. Science, 327(5972), 1480–1484. https://doi.org/10.1126/science.1182238
- Johnson, N. D., & Mislin, A. A. (2011). Trust games: A meta-analysis. Journal of Economic Psychology, 32(5), 865–889. https://doi.org/10.1016/j.joep.2011.05.007
- Jordan, J. J., Hoffman, M., Bloom, P., & Rand, D. G. (2016). Third-party punishment as a costly signal of trustworthiness. Nature, 530, 473–476. https://doi.org/10.1038/nature16981
- Kawakatsu, M., Kessinger, T. A., & Plotkin, J. B. (2024). A mechanistic model of gossip, reputations, and cooperation. Proceedings of the National Academy of Sciences, 121(20), e2400689121. https://doi.org/10.1073/pnas.2400689121
- Keser, C., & van Winden, F. (2000). Conditional cooperation and voluntary contributions to public goods. Scandinavian Journal of Economics, 102(1), 23–39. https://doi.org/10.1111/1467-9442.00182
- Kiyonari, T., Tanida, S., & Yamagishi, T. (2000). Social exchange and reciprocity: Confusion or a heuristic? Evolution and Human Behavior, 21(6), 411–427. https://doi.org/10.1016/S1090-5138(00)00055-6
- Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50, 569–598. https://doi.org/10.1146/annurev.psych.50.1.569
- Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi.org/10.5465/amr.1995.9508080335
- McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. https://doi.org/10.2307/256727
- Milinski, M., Semmann, D., & Krambeck, H.-J. (2002). Reputation helps solve the ‘tragedy of the commons’. Nature, 415, 424–426. https://doi.org/10.1038/415424a
- Nowak, M. A., & Sigmund, K. (1998). Evolution of indirect reciprocity by image scoring. Nature, 393, 573–577. https://doi.org/10.1038/31225
- Peysakhovich, A., Nowak, M. A., & Rand, D. G. (2014). Humans display a ‘cooperative phenotype’ that is domain general and temporally stable. Nature Communications, 5, 4939. https://doi.org/10.1038/ncomms5939
- Rand, D. G., Greene, J. D., & Nowak, M. A. (2012). Spontaneous giving and calculated greed. Nature, 489, 427–430. https://doi.org/10.1038/nature11467
- Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35(4), 651–665. https://doi.org/10.1111/j.1467-6494.1967.tb01454.x
- Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. https://doi.org/10.5465/amr.1998.926617
- Sommerfeld, R. D., Krambeck, H.-J., Semmann, D., & Milinski, M. (2007). Gossip as an alternative for direct observation in games of indirect reciprocity. Proceedings of the National Academy of Sciences, 104(44), 17435–17440. https://doi.org/10.1073/pnas.0704598104
- Thielmann, I., Spadaro, G., & Balliet, D. (2020). Personality and prosocial behavior: A theoretical framework and meta-analysis. Psychological Bulletin, 146(1), 30–90. https://doi.org/10.1037/bul0000217
- Wedekind, C., & Milinski, M. (2000). Cooperation through image scoring in humans. Science, 288(5467), 850–852. https://doi.org/10.1126/science.288.5467.850