Glossary

Trust game

Published 2 min read

What is the trust game?

The trust game studies decisions about sending and returning resources. One player transfers part of an endowment to another, the transfer is multiplied, and the second player decides how much to return.

How it works

Berg and colleagues introduced the investment game in 1995 to examine trust and reciprocity without repeated interaction. If both players care only about their own money and each knows this about the other, backward induction predicts no return and no initial transfer.

Real decisions can reflect expectations of repayment, risk attitudes, and concern for the other person. Those influences affect how the results should be interpreted.

Example

Player 1 starts with $10 and sends $5. If the transfer triples, Player 2 receives $15 and chooses what to return. Returning $5 leaves Player 1 with the original $10; returning more than $5 leaves Player 1 better off.

What happened in the original experiment?

In Berg and colleagues’ no-history condition, 32 pairs of University of Minnesota undergraduates played without receiving information about earlier participants’ decisions. Thirty of the 32 first players sent a positive amount. Across the pairs, the average amount sent was $5.16 and the average amount returned was $4.66.

Returns varied substantially. Of the 28 second players who received transfers initially exceeding $1, 12 returned only $0 or $1, while 11 returned more than the first player had sent. Thus, the experiment included both positive transfers and repayment, but trust did not reliably pay for every sender. These amounts were observed in one condition; they are not typical results for every trust game. A transfer can also express generosity or other preferences, so it is not a pure measure of trust.

Why it matters

The game makes trust and reciprocity questions testable under specified conditions. Cooperation need not be irrational: rational-choice models can include preferences about other people.

The ultimatum game asks a different question by letting the recipient reject a proposed division.

Sources and further reading