🤖 AI Summary
This work addresses the dynamic task allocation problem under stochastic task arrivals and an average-effort fairness constraint by proposing a novel dynamic pseudomarket mechanism. For the first time, individual preferences are explicitly incorporated into the equilibrium task assignment framework, achieving both asymptotic balance and Pareto efficiency. The theoretical analysis integrates stochastic matching models with asymptotic equilibrium theory and yields a closed-form expression for productivity gains that can be estimated using only aggregate data. Simulation results demonstrate that the proposed mechanism substantially improves average productivity compared to conventional rotation schemes, and its allocations are Pareto superior to those of current practical approaches.
📝 Abstract
I model balanced task allocation where tasks stochastically arrive and must be matched to a fixed set of agents; the novel constraint is that agents must receive allocations that require the same level of average effort. Social work supervisors, call center managers, and courts all rotate allocation across workers to satisfy this constraint, but the Rotation mechanism is not Pareto efficient. I design the Dynamic Pseudomarket (DPM) mechanism, and it satisfies Pareto efficiency and asymptotic balance. I derive an explicit equation characterizing DPM's expected productivity gain over Rotation that can be estimated only from aggregate statistics in firm-level data. Simulation results indicate large average productivity gains. These results indicate that preference-based allocation can Pareto dominate the status quo.