Dynamic delegation of experimentation

I study a dynamic relationship where a principal delegates experimentation to an agent. Experimentation is modeled as a one-armed bandit that yields successes following a Poisson process. Its unknown intensity is high or low. The agent has private information, his type being his prior belief that the intensity is high. The agent values successes more than the principal does, so prefers more experimentation. The optimal mechanism is a cutoff rule in the belief space: the cutoff gives pessimistic types total freedom but curtails optimistic types' behavior. Pessimistic types overexperiment while the most optimistic ones underexperiment. This delegation rule is time consistent.