who have “cheated” her in the past; and (c) tends to disregard information on the
opponent’s behavior in other matches.
One can identify three main contributions. Our study of indefinite interactions among
strangers complements and extends the experimental literature on indefinitely repeated
games, which has mostly focused on interactions among partners (recent examples
include Palfrey and Rosenthal, 1994; Aoyagi and Frechette, 2003; Dal Bo, 2005; Duffy
and Ochs, 2006). Second, our experimental findings can help define an empirically-
relevant criterion for equilibrium selection, based on behavioral considerations. This is
important from a practical standpoint because random matching models often display
multiple equilibria with various levels of efficiency, but an unambiguous equilibrium
selection criterion is missing (e.g., see the monetary equilibria in Aliprantis et al., 2006,
2007). Our laboratory findings shed light on what type of economic institutions may
facilitate the emergence of norms of cooperation in anonymous societies, complementing
a growing literature devoted to uncover theoretical links between the availability of
enforcement and punishment institutions on one side, and patterns of exchange and
cooperation on the other (e.g., Krasa and Villamil, 2000; Dixit, 2003).
The paper proceeds as follows: Section 2 discusses the related literature; Section 3
presents the experimental design; Section 4 provides a theoretical analysis; results are
reported in Section 5; and Section 6 concludes.
2 Related experimental literature
Our paper builds on the experimental literature on infinitely repeated games
(supergames), whose theoretical foundation can be traced back to Friedman (1971). Roth
and Murnighan (1978) were the first to implement infinitely repeated games in an
experiment by employing a probabilistic continuation rule, which transforms it into an
indefinitely repeated game. For risk-neutral subjects, a constant continuation probability
is theoretically equivalent to assuming a constant discount rate and an infinite horizon.
A number of experiments have adopted probabilistic continuation rules to study the
empirical validity of folk theorems for supergames. A basic result is that subjects