Media Briefings


  • Published Date: January 2009

The majority of people are honest in good relationships where all parties have behaved well in the past but dishonest towards people who have let them down in the past. That is the central finding of new experimental research by Professor Tore Ellingsen and colleagues, published in the January 2009 issue of the Economic Journal.

Professor Ellingsen comments:

"People who are honest in a first interaction with another person will not be in future if they experience dishonesty in return. They live by the maxim “fool me once, shame on you – fool me twice, shame on me”.‟

The implication of these findings is that negotiators whose relationships for one reason or another have soured will be unable to reap the benefits that honesty brings. So traders, couples, and diplomats hoping for honest dealing are not entirely naïve, but they must first inspire their partners‟ trust.

Honesty is often beneficial for the group but costly for the individual. For example, the buyer of a car who honestly tells the seller how much she is willing to pay increases the probability that there will be a trade but leaves herself open to exploitation. Yet to the great surprise of economists and other cynics, previous experimental research reveals that many people nonetheless negotiate honestly.

This study describes a new experiment that pinpoints more precisely the nature of such honesty. The main new finding is that unconditional honesty is rare.

The experiment has two stages. In the first stage, pairs of participants play a so- called Prisoners‟ Dilemma game, where each player chooses whether to co-operate or defect. Defection benefits the individual at the expense of the opponent. The participants are not informed that there will be a second stage before the first stage is over.

In the second stage, the same pairs of participants are invited to negotiate a trade of a single unit of a commodity. One is the seller, the other the buyer. Both the seller‟s cost and the buyer‟s valuation are drawn randomly from a known distribution by the experimenter. Before the start of the negotiation, the seller is privately informed about the cost, and the buyer is privately informed about the valuation.

The negotiation itself proceeds in two phases. In the first phase, sellers and buyers communicate freely through an electronic messaging system. In the second phase, they trade by way of a double auction. What happens is that both submit binding bids: if the bids are compatible, there is a trade at the midpoint price; if bids are incompatible, there is no trade.

The key finding concerns the behaviour of the „co-operators‟, that is, of the participants who themselves co-operated in the Prisoners‟ Dilemma. If their opponent also co-operated, 64% of the co-operators are honest in the second phase, and hardly any of them lie. But if their opponent did not co-operate, the co-

operators are hardly ever honest in the second phase, and the prevalence of outright lying goes up to more than 50%.


Notes for editors: „Trust and Truth‟ by Tore Ellingsen, Magnus Johannesson, Jannie Lilja and Henrik Zetterqvist is published in the January 2009 issue of the Economic Journal.

Ellingsen and Johannesson are at the Stockholm School of Economics. Jannie Lilja is at Uppsala University. Henrik Zetterqvist is at Hilti.

For further information: contact Tore Ellingsen on +468 716 2347 (email:; or Romesh Vaitilingam on 07768 661095 (email: