Why Are so Many People Who Experience Homelessness in a Canadian City from out of Town? a Report on a Preliminary Mixed Methods Study Using Machine Learning Models to Understand Migration and Homelessness

Background and purpose: In 2018, the largest social service delivery organization in Northern Ontario, Canada published a report in which approximately 66% of people experiencing homelessness who had reported migrating to its largest city of Thunder Bay. As a result, the District of Thunder Bay Social Services Administration Board (TBDSSAB) entered into a formal partnership with Lakehead University to understand why. Apart from this empirical purpose, the TBDSSAB, which receives funding from the provincial government to fund homelessness and housing initiatives, and oversees programming in the City, needed an evidence-base to advocate for funding vis-à-vis the province of Ontario and to tailor programming to suit the needs of its homeless population. To achieve the empirical purpose, we developed seven research questions: (1) from where are people migrating?; (2) why do people leave their home communities? (3) why do people choose to migrate to the City; (4) why do people choose to remain in the City once here? (5) what factors predict if someone stays or leaves; (6) if someone stays, how long are they likely to stay; and (7) what factors predict their stay duration.

Methods: To answer these research questions we used a mixed methods research design consisting of four largely independent data sets: qualitative interviews (n = 17), homelessness enumeration (n = 98); surveys administered in shelters (n = 120) and two variables from a national database on homelessness (n = 120). To analyse these data, we used Thematic Analysis for the qualitative interviews, SPSS for the enumeration data, and Python-based machine learning models for the shelter survey and HIFIS.

Results: The six key findings from the study are: (1) social factors including family, friends, and a sense of community drive migration and motivate stay; (2) service factors, such as access to health care, housing, and social services drive migration and motivate stay; (3) economic factors – i.e., employment and education – drive migration; (4) lack of money is a barrier to leaving the city for those who want to return to communities; (5) a majority of people migrated from one of three bordering districts, each with their own social services organizations; (6) migrating from or through one of these districts seems to predict migration to the city and to a lesser extent stay; and (7) a high proportion of people migrated from First Nations communities in those neighbouring districts.

Conclusion and Implications: The findings from this study were presented in November 2022 to a community-university audience in the form of a report. The organization has begun using that report to advocate vis-à-vis the province and federal government. For instance, in January 2023 the TBDSSAB used some of the findings to advocate for increased funding for housing in First Nations communities. Ongoing advocacy work continues. And consultations with neighbouring district service organizations and First Nations organizations for a collaboration to scale the project beyond the City are under way.