An official website of the United States government
Abstract ObjectiveThe study tests a community- and data-driven approach to homelessness prevention. Federal policies call for efficient and equitable local responses to homelessness. However, the overwhelming demand for limited homeless assistance is challenging without empirically supported decision-making tools and raises questions of whom to serve with scarce resources. Materials and MethodsSystem-wide administrative records capture the delivery of an array of homeless services (prevention, shelter, short-term housing, supportive housing) and whether households reenter the system within 2 years. Counterfactual machine learning identifies which service most likely prevents reentry for each household. Based on community input, predictions are aggregated for subpopulations of interest (race/ethnicity, gender, families, youth, and health conditions) to generate transparent prioritization rules for whom to serve first. Simulations of households entering the system during the study period evaluate whether reallocating services based on prioritization rules compared with services-as-usual. ResultsHomelessness prevention benefited households who could access it, while differential effects exist for homeless households that partially align with community interests. Households with comorbid health conditions avoid homelessness most when provided longer-term supportive housing, and families with children fare best in short-term rentals. No additional differential effects existed for intersectional subgroups. Prioritization rules reduce community-wide homelessness in simulations. Moreover, prioritization mitigated observed reentry disparities for female and unaccompanied youth without excluding Black and families with children. DiscussionLeveraging administrative records with machine learning supplements local decision-making and enables ongoing evaluation of data- and equity-driven homeless services. ConclusionsCommunity- and data-driven prioritization rules more equitably target scarce homeless resources.
more »
« less
Predicting the Slide to Long-Term Homelessness: Model and Validation
In spite of numerous programs and interventions, homelessness remains a significant societal concern. Long-term homelessness is particularly problematic because it can be increasingly difficult to escape from, and because it represents a continuous drain on societal resources. This paper develops a model for predicting long-term homelessness in response to a simple question: if an individual becomes homeless, what influences the individual's slide to long-term homelessness? The data we analyze to answer the question comes from the City of Boston. The model points to race, veteran status, disability, and age as key factors that predict this slide. The paper describes and illustrates the model along with problems encountered in data preparation and cleansing, prior scholarly work that helped to shape our decisions, and collaboration with participants in the ecosystem for homeless care that complemented the model-building effort. The results are important because they point to possible policy interventions (programs and funding) and process improvements (at homeless shelters) to mitigate this slide.
more »
« less
- Award ID(s):
- 1951896
- PAR ID:
- 10297068
- Date Published:
- Journal Name:
- 2019 IEEE 21st Conference on Business Informatics (CBI)
- Page Range / eLocation ID:
- 31 to 40
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
This study initiated an exploration into how community members, specialists in housing issues, and social scientists might collaborate to address homelessness in Alaska. Through interviews and participant observation of planning meetings and related activities, the researchers are gathering insights from design experts, community organizers, and experts working on urban-rural homelessness in Alaska. This includes gathering information about cold weather design processes and issues facing urban-rural homelessness in Alaska, as well as the identification of possible research questions that can inform the development of a grant application for a multi-year research study. The study includes in-person as well as virtual research activities. Because of geographic distances, the majority of initial research activities were conducted virtually, but in-person field site visits began to take place June 15, 2021, and subsequent trips have taken place from August 2021-onward. These research trips involve site visits, participation in meetings, and in-person interviews when possible. Phase 1: 24 initial interviews were conducted with a range of stakeholders about housing insecurity in Alaska and the impacts of the COVID-19 pandemic. Includes interviewees from remote villages, from the Association of Alaskan Housing Authorities (AAHA), homeless advocates, designers, social scientists, engineers, and builders. Topics included myths about homelessness, homeless versus houseless terminology, research organizations, policies, impacts of pandemic, housing needs, and contrasting strategies. Analysis and synthesis with subsequent data is ongoing. 01: policy 02: interview with researcher 03: homelessness - Anchorage - rural communities - data sharing 04: design in rural communities 05: housing shortages in rural communities 06: technical issues in housing - collaborating with rural communities 07: homeless community in Fairbanks 08: history of Cold Climate Housing Research Center 09: design - homelessness - Anchorage 10: homelessness - rural/hub/urban - need for housing design repository 11: homelessness - Nome - Savoonga - designers need to visit villages 12: reverse interview - designer interviews researchers 13: homelessness - Anchorage - Bethel - housing costs 14: homelessness - rural/hub/urban spectrum - subsistence - houseless term 15: homelessness data and Bethel - impacts of pandemic - myths 16: homelessness data and Bethel - impacts of pandemic 17: ISERC (Integrated Security Education and Research Center) research 18: homelessness data and Bethel - CARES Act 19: homelessness data (gaps) and Bethel - CARES Act 20: homelessness data and Bethel 21: designer - public awareness and museum exhibits 22: veterans and community organizer 23: AAHA staff member 24: homelessness - Fairbanks - pandemic impacts on rescue missions Phase 2: 49 additional interviews were conducted with support from NSF funding (NSF 2103356: RAPID: COVID-19, Remote Ethnography, and the Rural Alaskan Housing Crisis). A meta-data description of the participants and topics are attached ('RAPID_interview_list___Descriptions').more » « less
-
Artificial intelligence, machine learning, and algorithmic techniques in general, provide two crucial abilities with the potential to improve decision-making in the context of allocation of scarce societal resources. They have the ability to flexibly and accurately model treatment response at the individual level, potentially allowing us to better match available resources to individuals. In addition, they have the ability to reason simultaneously about the effects of matching sets of scarce resources to populations of individuals. In this work, we leverage these abilities to study algorithmic allocation of scarce societal resources in the context of homelessness. In communities throughout the United States, there is constant demand for an array of homeless services intended to address different levels of need. Allocations of housing services must match households to appropriate services that continuously fluctuate in availability, while inefficiencies in allocation could “waste” scarce resources as households will remain in-need and re-enter the homeless system, increasing the overall demand for homeless services. This complex allocation problem introduces novel technical and ethical challenges. Using administrative data from a regional homeless system, we formulate the problem of “optimal” allocation of resources given data on households with need for homeless services. The optimization problem aims to allocate available resources such that predicted probabilities of household re-entry are minimized. The key element of this work is its use of a counterfactual prediction approach that predicts household probabilities of re-entry into homeless services if assigned to each service. Through these counterfactual predictions, we find that this approach has the potential to improve the efficiency of the homeless system by reducing re-entry, and, therefore, system-wide demand. However, efficiency comes with trade-offs - a significant fraction of households are assigned to services that increase probability of re-entry. To address this issue as well as the inherent fairness considerations present in any context where there are insufficient resources to meet demand, we discuss the efficiency, equity, and fairness issues that arise in our work and consider potential implications for homeless policies.more » « less
-
null (Ed.)In spite of significant prior work, few scholars have addressed the unique problems of business process modelling in humanitarian settings. We surface and explore these challenges by treating the ‘guest intake process’ at homeless shelters as an exemplar. In most cities, the homeless shelters are a part of the ecosystem of care that includes hospitals, training agencies, housing agencies and others to support homeless individuals. The exemplar we have chosen represents the first line of defence for the homeless guests: the Intake practice at a homeless shelter, a process that remains fairly knowledge-intensive. Our work follows a collaborative research approach, working with the largest homeless shelter in the region. We conduct observations and interviews at the shelter to construct process models with the Petri Net formalism, followed by analytical experimentation. Lessons learned from the work are compared against prior research in modelling of knowledge-intensive processes. We find that process modelling in humanitarian settings such as homelessness remains fraught with problems. We identify new directions, where prior work in knowledge-intensive processes may require extensions in response to the unique nature of work in humanitarian settings. The paper elaborates these findings, illustrates with authentic examples, and outlines possibilities for future research and implications for practicemore » « less
-
Previous homelessness research examined common pathways into homelessness, yet not much is known about how people navigate through services while experiencing homelessness. This study explored the service pathways of homeless individuals in the U.S. context, which show their connection with multiple organizations and their lived experiences of using services over time. We conducted 12 semi-structured in-depth interviews to grasp the history of service pathways, including the number of organizations, time gaps between services, and referral patterns. We also conducted participant observation shadowing with a subset of the study participants to understand how they interact with caseworkers. The length of service pathways varied, from less than five years to more than two decades. On average, participants went through at least three and up to eight organizations. Regarding service experiences, systemic- and individual-level themes were drawn for negative or positive experiences, such as strict organizational policies and the caseworker’s demeaning attitudes (negative), or supportive organizational culture and strong employee competencies (positive). The findings of this study provide deeper insights into homeless populations’ service trajectories and their experiences throughout the service-navigating process.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/CBI.2019.00011
