Orateur
Description
Motivated by a housing allocation problem faced by our public-sector partner, the Los Angeles Homeless Services Authority (LAHSA), we study sequential resource allocation in high-stakes social settings, where fairness is critical. Each month, we must allocate limited housing capacity among individuals experiencing homelessness. We model this as a Markov Decision Process (MDP), where individuals transition between housing and homelessness states based on stochasticity. To manage complexity, we adopt a grouped weakly coupled MDP framework, clustering individuals by covariates (via robust ML techniques) and solving for an optimal policy using a fluid linear program. To enforce fairness, we introduce demographic parity constraints on both resource allocation and long-term outcomes across protected groups. By refining our clustering approach, we balance equity while maintaining distinctions based on other covariates. We also analyze the price of fairness, characterizing trade-offs between equity and housing stability. The resulting policies outperform historical allocations while mitigating disparities in access and outcomes.
