Discussion Papers 745
4 Experimental design
both players choose team remuneration and an effort of 50 they can make a profit as high as
250 each. This solution maximizes the sum of profits of both players in the group.
In the ASYM parameterization, the player with the lower effort cost, player 1, makes a maxi-
mum profit in the case of private remuneration that is larger than his equilibrium profit in the
case of team remuneration, while the opposite is true for the player with the higher effort cost,
player 2. Thus, in the subgame perfect equilibrium solution private remuneration will be real-
ized. The low-cost player chooses an effort of 35 and makes a profit of 122.5, while the high-
cost player chooses an effort of 28 and makes a profit of 98. Each of the two players can do
better by choosing team remuneration and an effort of 40 by the low-cost player and 32 by the
high-cost player. In this Pareto optimal solution, the low-cost player’s profit is 128 while the
high-cost player’s profit is 160, which implies that the high-cost player makes a higher profit
than the low-cost player. As each of the two players has an incentive to individually deviate,
this Pareto optimal solution does not describe an equilibrium of the second stage subgame.
4.2 Organization
In the experiments, participants played 30 repetitions, called periods, of the effort game in
two-player groups that remained unchanged during the entire experiment (“partners” design).
Based again on the backward induction principle, the subgame perfect equilibrium prediction
of the finitely repeated effort game prescribes in each period the subgame perfect equilibrium
prediction presented above. Behaviorally, however, the repetition opens up room for recipro-
cal behavior or conditional cooperation.
In the beginning of the second decision stage, in the VOLUNTARY treatments, each partici-
pant was informed of the actual remuneration mode, depending on both players’ decisions in
the first stage. At the end of each period, each participant was informed about his or her indi-
vidual profit. In the case of team remuneration also the average effort level (and, thus, impli-
citly the effort level chosen by the other player) was communicated to each participant. A
participant’s sum of profits in all 30 periods determined his or her individual earnings in the
experiment.
The computerized experiments were run at the experimental economics laboratory at
CIRANO-LUB (Centre Interuniversitaire de Recherche en Analyse des Organisations - Labo-
ratoires Universitaires Bell) in Montréal. We conducted sessions with eight participants each.
11