Welfare-optimal trade-off between allocative efficiency and waiting costs in NTU models

Determine the welfare-optimal policy in the NTU fluid waitlist model with heterogeneous buyer values that trades off allocative efficiency (assigning goods to higher-value buyers) against the social costs of waiting, beyond the decreasing inverse hazard rate case; explicitly characterize when complete pooling with random allocation versus screening via waiting is optimal under general value distributions.

Background

In the NTU setting with heterogeneous values, allocative efficiency (serving higher-value buyers) conflicts with the social costs generated by waiting. The authors show that under a decreasing inverse hazard rate, welfare maximization leads to complete pooling and random allocation (eliminating waiting costs), even at the expense of allocative efficiency.

However, they explicitly state that, in principle, the general trade-off between efficiency and waiting costs is unclear. This leaves open the broader characterization of welfare-optimal policies beyond specific regularity conditions.