Trust Game
Also known as: Investment Game, Berg Game, Two-Player Trust Game
The trust game, introduced by Berg, Dickhaut, and McCabe in 1995 (and often called the investment game), is a two-player exchange that operationalizes interpersonal trust and reciprocity in money. An investor receives an endowment and may send any portion to an anonymous trustee; the experimenter multiplies the transfer (typically tripling it); the trustee then decides how much, if any, to return. Standard game theory with purely self-interested players predicts the investor should send nothing because a selfish trustee returns nothing -- yet investors reliably send substantial amounts and trustees reliably return some, contradicting the narrow self-interest prediction. Because the amount sent cleanly measures trust and the amount returned measures trustworthiness, the paradigm became a workhorse in social psychology, behavioral economics, and neuroscience for studying social preferences and cooperation between strangers.
Key highlights
- Provides incentivized, continuous behavioral measures of both trust (amount sent) and trustworthiness (amount returned).
- Has a clear self-interest benchmark, so deviations directly quantify social preferences.
- Highly portable across labs, cultures, and methods, including neuroimaging and pharmacological studies.
- Manipulable: partner identity, information, communication, and stakes can be varied to test causal hypotheses about trust.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use the trust game when you need a behaviorally incentivized, quantitative measure of interpersonal trust and reciprocity rather than self-reported trust attitudes. It is well suited to studying how trust varies with partner characteristics (group membership, reputation, facial appearance), with context (anonymity, communication, oxytocin, stakes), and across populations and cultures, and it pairs naturally with neuroimaging and hormonal studies of social decision making. It is less appropriate when trust toward institutions or generalized social trust is the target, where survey scales fit better, or when the single-shot, monetary framing strips away features of real trust relationships (emotion, history, identity) that matter for the question. Careful control of anonymity and beliefs is essential, since participants' inferences about the partner drive transfers.
Strengths & limitations
- Provides incentivized, continuous behavioral measures of both trust (amount sent) and trustworthiness (amount returned).
- Has a clear self-interest benchmark, so deviations directly quantify social preferences.
- Highly portable across labs, cultures, and methods, including neuroimaging and pharmacological studies.
- Manipulable: partner identity, information, communication, and stakes can be varied to test causal hypotheses about trust.
- Single-shot monetary exchange abstracts away emotion, history, and identity central to real trust.
- Amounts sent reflect not only trust but also risk preferences, altruism, and expectations, requiring control conditions to separate them.
- Behavior is sensitive to framing, stake size, and perceived anonymity, complicating cross-study comparison.
- Generalization from anonymous strangers to known partners and institutions is not automatic.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
Why do investors send money when the self-interested prediction is to send nothing?
The subgame-perfect prediction for purely selfish players is that the trustee returns nothing, so the investor should send nothing. Real investors send substantial amounts because they hold positive beliefs about reciprocity and value cooperative outcomes. The reliable finding that trustees do return meaningful sums vindicates this trust on average, demonstrating that trust and reciprocity are basic features of human exchange rather than irrational mistakes.
How is the trust game different from a simple measure of altruism or risk-taking?
Sending money risks loss for a possible gain, so amount sent can reflect risk preference and altruism as well as trust. To separate these, researchers add control conditions -- for example a risky dictator game where the return is determined by a random device rather than a person -- so that any extra sending in the trust game over the control condition reflects genuine trust in another's reciprocity rather than mere risk tolerance or generosity.
What does the amount returned by the trustee measure?
The trustee's return is the behavioral measure of trustworthiness or positive reciprocity. Because a money-maximizing trustee would keep everything, any return reveals a willingness to honor trust and share the cooperative surplus. Researchers often examine returns as a function of the amount received to study conditional reciprocity -- whether trustees reward larger displays of trust with proportionally larger returns.
Sources
- 1.Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, Reciprocity, and Social History. Games and Economic Behavior, 10(1), 122-142.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Trust Game. ScholarGate. https://scholargate.app/social-psychology/trust-game