Moral hazard in long-term guaranteed contracts: theory and evidence from the NBA (with Brad Rice) October 2007, Job Market Paper

This paper constructs a model of optimal contract length based on the trade-off between the effort incentives of short-term contracts and risk aversion. The results lend support to the notion that agents exert less effort in a long-term contract, and that the effort levels are increasing within contracts, even in the presence of career concerns. We further show that, in spite of these adverse effort incentives, a multi-period contract between a risk-averse agent and a risk neutral principal may be constrained Pareto efficient. To quantify the degree of shirking in long-term contracts, we use performance and contract data from an unbalanced panel of 654 National Basketball Association players. Because over 90 percent of players in the data set are observed for more than one year, we are able to compare within-contract productivity while accounting for individual-specific heterogeneity. The fixed-effects estimates imply that player effort increases, on average, over 2% per year as he moves towards the contract’s expiration. Furthermore, increases in effort are non-linear, with the largest gains occurring at the end of the contract.