Implications for Trust and Trustworthiness When Institutions Change

Institutions change—in Iraq, the formerly Communist countries, and many private and public organizations. But do people adapt to the new institutional environment, and if so, how quickly? This paper examines institutional change—how long it takes people to transition from one institutional environment to another or, put differently, whether old institutional regimes have an afterglow. More specifically, we study whether trust and trustworthiness can be fostered by first exposing people to an environment conducive to trust. We are interested in whether (intrinsic) trust and trustworthiness can be induced in the long run by providing extrinsic incentives for trust and trustworthiness in the short run. Reputation systems may provide incentives for trustworthiness and trust. Direct reputationbuilding may occur in repeated games where pairs of subjects play the same stage game repeatedly, but repeat transactions are not necessarily the rule in today’s global economy. In population games where agents are randomly re-matched in every period, indirect reputation systems are a potential substitute for personal interactions—provided information about others’ past behavior is available. On eBay, for example, buyers are willing to pay a premium of 8.1 percent of the selling price to a seller with an established good reputation (Paul Resnick et al., 2003). This paper examines experimentally to what degree indirect reputation-building substitutes for direct reputation-building in repeat interactions in the short run and analyzes the effects these environments have on behavior in the long run. In contrast, most earlier experimental studies focus on one-shot and repeat interactions in the short run. We compare the effects of direct and indirect reputation-building in a binary-choice trust game where a buyer (the trustor) can either interact with the seller (the trustee) or exit. The trustee can either honor or exploit trust. The payoffs are such that a money-maximizing trustee prefers exploiting to honoring trust in a one-shot game, while a money-maximizing trustor prefers not offering trust to being exploited. The unique Nash equilibrium of the single-shot game predicts no trade. Figure 1 presents the game we implemented with the actual payoffs in cents used. In our experiment, subjects participate in the trust game in two blocks of 10 rounds each, which is common knowledge. In phase 1, the first 10 rounds, they are confronted either with a standard, “one-shot” random matching treatment (“stranger” or “S”); a fixed-pairs, finitely repeated game treatment (“partner” or “P”); or * Bohnet: Kennedy School of Government, Harvard University, 79 JFK Street, Cambridge, MA 02139 (e-mail: Iris_Bohnet@Harvard.edu); Huck: Department of Economics and ELSE, University College London, Gower Street, London WC1E 6BT, United Kingdom (e-mail: s.huck@ucl. ac.uk). We thank Rachel Croson and the participants at the 2004 ASSA meetings for their helpful comments and Jeffrey Bielicki for his excellent research assistance. We gratefully acknowledge financial support from the Russell Sage Foundation, the Leverhulme Trust, and the Economic and Social Research Council (U.K.) via ELSE. 1 For a recent survey, see James Andreoni and Rachel Croson (2004). Studies examining the effects of different institutional environments over time include Bohnet et al. (2001) and Ernst Fehr and Simon Gachter (2003). 2 The experiments were computerized using Urs Fischbacher’s (1999) z-tree software. The instructions are available upon request.