Group incentives or individual incentives? A real-effort weak-link experiment

Motivated by previous research on coordination problems and incentive design in organizations, we compare group incentives and individual incentives in a new experimental test-bed: a real-effort task embedding a weak-link technology. Comparing group-incentive and individual-incentive treatments, we find that the observed dynamics of both individual errors and worst performances within firms, after a phase of learning, are largely indistinguishable. Importantly, and possibly explaining our finding, more than 80% of our laboratory firms, notwithstanding initially widespread inefficiency, were eventually able to achieve and sustain efficient coordination despite the presence of an unforgiving payoff structure. This result, which may be due to the reduced strategic uncertainty so characteristic of chosen-effort weak-link game experiments, stands in stark contrast to standard results in the coordination game literature and provides an interesting challenge to test-beds currently used.