Why test the theory of incentives in a dynamic framework

The recognition that contracts have a time dimension has given rise to an abundant literature since the end of the 1980s. In such a dynamic context, the contract may take place over several periods, and agents develop repeated interactions. Surprisingly, few papers have tried to apply the predictions of the dynamic theory incentives to data. However, taking a dynamic context into account can improve static empirical approaches by introducing new tools to distinguish between adverse selection and moral hazard, two asymmetrical contexts representative of the Principal–Agent model, or by solving the problem of endogeneity. A dynamic empirical approach also allows to renew the theoretical contract conception, and reveals some new features of contracts: memory, learning, and commitment.