Replacing the visual street census with respondent-driven sampling, a peer-referral method that produces quasi-probability estimates of the unsheltered population. Piloted in King County in 2022, refined in 2023, and used for the formal HUD count from 2024.
Almquist et al. (2025), American Journal of Epidemiology · 2023 Understanding Homelessness project
Measuring and modelling the multi-modal networks and demographics of people experiencing homelessness, and building the statistical methods that make network-based enumeration practical.
Using longitudinal street-outreach records from Seattle to study where unsheltered people actually live and where they go afterwards — encampment siting relative to amenities, and housing and movement outcomes following encampment clearances.
Sutton, Walker, Hagopian & Almquist (2025), Cities · Morande et al., encampment clearances
Extending peer-referral sampling to youth homelessness, with young people themselves recruiting participants rather than being counted by canvassers.
Developing higher-frequency enumeration and needs-assessment methods built around the people and organisations already doing outreach.