Task Allocation and On-the-job Training
We study dynamic task allocation when providers' expertise evolves endogenously through training. We characterize optimal assignment protocols and compare them to discretionary procedures, where it is the clients who select their service providers. Our results indicate that welfare gains from centralization are greater when tasks arrive more rapidly, and when training technologies improve. Monitoring seniors' backlog of clients always increases welfare but may decrease training. Methodologically, we explore a matching setting with endogenous types, and illustrate useful adaptations of queueing theory techniques for such environments.
We thank Heski Bar-Isaac and Francisco Buera for helpful comments. We gratefully acknowledge financial support from the National Science Foundation, grant SES 1629613. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.