Tête-à-tête. The brain activity of an investor (left) and trustee in an economic exchange game shed light on the neural basis of trust.

Building Trust in the Brain

Economists have long puzzled over why people sometimes trust complete strangers with their hard-earned dough. A new study suggests that this seemingly irrational behavior is wired into our brains.

Researchers at Caltech in Pasadena, California, and Baylor College of Medicine in Houston, Texas, recruited volunteers to play a trust game. In each of 10 rounds, one player, designated the "investor", received $20. The investor then had the option of sending some, all, or none of the $20 to the other player, the "trustee." Both sides knew the money given to the trustee would triple, so trust boiled down to two questions: How much money would the investor give to the trustee, knowing that the trustee could choose to give nothing back? And how much money would the trustee return to the investor, knowing the investor could pocket the cash and stiff the trustee in the future?

The best indicator of whether a player trusted another with his or her money turned out to be a feature of the exchange that the researchers termed "reciprocity." Reciprocity could be benevolent, neutral, or malevolent. An investor could demonstrate benevolent reciprocity, for example, by increasing the contribution to the trustee immediately following a round in which the trustee had reduced payback to the investor. Trustees generally rewarded such behavior with a greater return on the subsequent round. But if an investor demonstrated malevolent reciprocity by repaying the trustee's generosity with stinginess, the trustee usually gave less the next time around.

Functional magnetic resonance imaging (fMRI) scans of the trustees' brains during the transactions revealed that activity in a region called the caudate nucleus was greatest for rounds in which the investor showed benevolent reciprocity and most subdued when the investor showed malevolent reciprocity. Moreover, caudate activity rose and fell with changes in the amount of money trustees returned to their investors on the subsequent round. In the 1 April issue of Science, the team concludes that activity in a trustee's caudate reflects both the fairness of the investor's decisions and the trustee's intention to repay those decisions with trust (or not). The caudate nucleus is well connected to the brain's reward pathways, and previous work has shown that it revs up when subjects expect a reward. The researchers therefore speculate that trust, admirable trait that it is, boils down to a cold calculation of expected rewards--in this case, reciprocity from the investor.

"I think it's a very important paper," says Paul Zak, who directs the Center for Neuroeconomic Studies at Claremont Graduate University in California. "It's going to change the way we think of social interactions."

Related sites
A recent Science article on the field of neuroeconomics
Co-author Read Montague's site
Co-author Steven Quartz's site
Co-author Colin Camerer's site
Paul Zak's site