Homelessness and housing insecurity present fundamental challenges for the delivery of social services. The absence of stable accommodations increases the risks of poor outcomes for all members of the household, including children, experiencing this insecurity. This project is an exploratory study in the use of techniques from artificial intelligence (AI) to improve early screening and the delivery of targeted assistance to households that are at risk of future homelessness and child maltreatment. The team will seek to develop novel methods for allocation of scarce housing-support resources to at-risk households, taking into account considerations of both overall efficiency and fairness. This work will necessitate novel problem formulation and algorithm development in AI as well as creating new ethical methods for deciding on how to effectively deliver social services taking into account the vast complexity of human behavior. Moreover, reducing the risks of homelessness and child maltreatment are critical societal goals with the potential to substantively improve the lives of many of our most vulnerable citizens.
This project will explore the feasibility of using novel algorithmic techniques to inform societal decision-making on the allocation of scarce resources, with the specific goal of improving service system outcomes for both homelessness and child welfare. The team's focus will be on homelessness prevention interventions that offer timely, non-reoccurring resources to stabilize families at risk of experiencing housing crises; examples of such resources include landlord mediation, one-time rent or utility payments, and moving expenses. They will leverage unique datasets on child welfare and homelessness in the research, and use these to inform the design of machine learning approaches to prediction of outcomes (specifically, repeat episodes of homelessness and future interactions with child protective services), and optimization techniques that leverage these predictions in order to decide on which households to target for prevention interventions. The interplay of prediction and optimization, in a context where the overall allocation must both improve social welfare (measured along multiple dimensions) and satisfy notions of fairness, equity, and local justice in the allocation of scarce resources, is a challenging domain for AI.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.