behavior is inconsistent with personal income maximization, it has been shown to be
remarkably robust (Ostrom et al, 1992; Fehr and Gaechter, 2002; Casari and Plott, 2003).
A third novel feature of our study is to examine how this behavioral trait may be
employed in supporting the cooperative equilibrium in an infinitely repeated game, where
there does already exist a punishment technology. This design may be useful in isolating
possible elements or economic institutions that can facilitate selecting the cooperative
equilibrium in a more general setting.
As noted earlier, the matching protocol across supergames is also important because
of possible contagion effects across supergames. It is therefore helpful to mention the
various protocols adopted in previous experiments. To play a supergame in a session with
N participants, subjects can be partitioned into K economies. The way we ran multiple
supergames is to ensure that any two subjects were never assigned to the same economy
for more than one supergame. A more rigorous partitioning procedure in the experimental
literature is to rule out that anyone may share a common past opponent.3 Both procedures
control for contagion effects.4 This contrasts with randomly matching the same set of
subjects after each period and after each supergame (for instance, Schwartz et al., 1999,
and random pairing in Duffy and Ochs, 2006).
3 Experimental design
This experiment has four treatments (Table 1). While the stage game (Table 2), the
continuation probability, and matching protocols were identical across treatments, we
manipulated the amount of information and the punishment options available to subjects.
The efficient outcome can be supported as an equilibrium in all treatments.
3 In Dal Bo (2005) each subject plays three supergames (treatment). In the “Dice” sessions, in each
supergame participants are partitioned into K=(N/2) two-person economies. The partitioning across
supergames is such that the decisions one subject made in one supergame could not affect, in any way, the
decisions of subjects he or she would meet in the future. Ensuring the absence of contagion effects in this
manner requires very large session sizes. For a theoretical discussion of matching procedures see Aliprantis
et al. (2006, 2007).
4 In our study each subject played for five supergames. Subjects may have shared a common past opponent
in supergames three or later. Aoyagi and Frechette (2003) used a different in between matching protocol;
each agent plays G>10 supergames. In the first 10 supergames they partition agents as in the former way
described in the main text above and in the last (G-10) supergames the randomly rematch participants.