Discussion Papers 745
1 Introduction
tary contribution to the public good team effort. We have modeled the situation such that
through cooperation the team can realize higher profits than the maximum sum of profits that
the two players can realize under private remuneration. This is based on the assumption of a
higher payment per effort unit under team remuneration than under private remuneration,
reflecting that in the case of team remuneration the firm has a cost advantage relative to the
case of private remuneration due to reduced monitoring costs. In real life, effort can hardly be
measured so that incentive payment has to be based on a surrogate, typically an output vari-
able. In a work team it is often very hard to measure an individual’s contribution to the team
output.
While there is a large theoretical literature on team incentives when individual effort is not
observable (e.g., Holmstrom 1982, Rasmusen 1987, Itoh 1991, 1992, McAfee and McMillan
1991), only few empirical studies have examined the impact of teams on output (e.g., Weitz-
man and Kruse 1990 and Hamilton, Nickerson and Owan 2003, and the references therein).
While data at the firm level may be available, they might not allow for the comparison of
outcomes with and without team incentives, or of voluntary and enforced teaming. Compari-
sons across firms are likely to suffer from issues of worker self-selection and distorting fac-
tors due to additional differences in human resource and production processes. Thus, con-
trolled laboratory experiments are a valuable method to collect data.
Nalbantian and Schotter (1997), for example, present an experimental analysis of the func-
tionality of various team incentive schemes. Their revenue sharing scheme is very similar to
our team remuneration scheme with profit sharing, except for a stochastic element in their
experiment. More concretely, their participants face a moral hazard problem, while in our
experiment effort is observable and verifiable. Both experiments, however, involve a free-
riding incentive. Nalbantian and Schotter observe declining effort over time, i.e., an increase
in free riding behavior.
In the experimental economics literature, we find a large number of studies on voluntary con-
tributions to finance a public good to which our experiment relates (e.g., Isaac, Walker and
Williams 1994, Andreoni 1988, 1995). These studies provide a good picture of the kind of
environments that are needed to facilitate cooperation and of the behavioral dynamics. The
most important finding is that people do make voluntary contributions above those that are in
their egoistic individual interest and that their contribution decisions tend to be based on the