Discussion Papers 745
5 Experimental results
Reciprocity can show both in the choice of the team-effort level and in the choice of the re-
muneration mode. For the choice of team effort, we define reciprocity as in the public-goods
literature. Specifically, we interpret participants' voluntary contributions in terms of condi-
tional cooperation, following Keser and van Winden (2000): when a player changes his effort
from one period to the next, he increases the effort if in the previous period he contributed a
lower effort than the other player but decreases it if in the previous period he contributed a
higher effort than the other player.
In the VOLUNTARY SYM treatment, considering all the observations where the realization of
team remuneration followed team remuneration and where the two players’ efforts were dif-
ferent in the previous period, we identify seven groups that obeyed this reciprocity rule in the
majority of cases; only one group did not follow this rule of reciprocity in the majority of
cases. Similarly, in the ENFORCED SYM treatment, we identify seven of the eight groups
that follow the rule of reciprocity in the majority of cases. Thus, participants show in both
SYM treatments a significant tendency to follow this specific rule of reciprocity (binomial
tests, 10 -percent significance).
In the VOLUNTARY ASYM treatment, considering all cases where team remuneration fol-
lowed team remuneration and where the two players’ efforts were different in the previous
period, we observe that the low-cost players, if they change their effort from one period to the
next, show a significant tendency to increase their effort if they contributed less in the previ-
ous period and to decrease it if they contributed more in the previous period (binomial test, 5 -
percent significance). The high-cost players do not show this kind of reciprocal behavior. In
the ENFORCED ASYM treatment, we find that neither the low nor the high-cost players show
a significant tendency for reciprocal behavior under enforced teaming. Thus, only when team-
ing is voluntary, do we observe reciprocity in the low-cost players’ effort choices.
With respect to the remuneration choice, our hypothesis is that participants tend to choose
private remuneration after having chosen a higher effort than the other player in previous
team remuneration. In other words, the choice of private remuneration can be used strategi-
cally as a kind of punishment—by ostracism or a simple rejection to interact—with the inten-
tion to enforce future cooperation. The experimental results by Fehr and Gachter (2000a) and
Masclet et al. (2003) show that participants in public-goods experiments make use of costly
opportunities to punish uncooperative others. Hirshleifer and Rasmusen (1989) show in a
theoretical paper that a way to enforce cooperation in a finitely repeated n-person prisoners’
20