What We Still Don't Know

UNK-01Critical

How many people placed into housing remain housed at 12 and 24 months, measured consistently across all nine counties?

Why it matters: Placement counts without retention cannot establish durable resolution.
Likely record holders: Continuums of Care · HMIS leads · county homelessness departments · housing providers
Methodological obstacle: Incompatible definitions, privacy restrictions, and weak cross-county longitudinal linkage.
What would resolve it: Harmonized cohort definitions and privacy-preserving cross-system longitudinal analysis.

UNK-02Critical

Why it matters: Site closure is not equivalent to ending homelessness.
Likely record holders: cities · counties · outreach contractors · shelter operators · HMIS leads
Methodological obstacle: Operations are usually measured as site events rather than person-level longitudinal outcomes.
What would resolve it: Person- and site-level follow-up with documented destinations and recurrence checks.

UNK-03Critical

Why it matters: Regional displacement may be produced in one jurisdiction while fiscal and service burdens are concentrated elsewhere.
Likely record holders: CoCs · counties · cities · HMIS leads · court systems · housing agencies
Methodological obstacle: Last stable address, displacement location, first unsheltered location, and service geography are rarely linked.
What would resolve it: Regional longitudinal mobility dataset with harmonized geographic fields.

UNK-04High

Why it matters: Vehicle homelessness is strategically important in suburban geographies, working households, and families, but is poorly measured.
Likely record holders: PIT teams · parking agencies · police departments · tow contractors · safe parking providers
Methodological obstacle: Enumeration difficulty and fragmented administrative records.
What would resolve it: Linked vehicle-residency, towing, safe-parking, and housing-outcome records.

UNK-05High

Why it matters: Street-centered measurement misses substantial housing precarity and prevention opportunity.
Likely record holders: schools · Census products · tenant organizations · social services · motel programs
Methodological obstacle: No single operational definition or complete administrative universe.
What would resolve it: Mixed-method estimation using survey, education, service, and local administrative data.

UNK-06Critical

Why it matters: Housing placement is an important outcome but not the endpoint of health analysis.
Likely record holders: public health departments · medical examiners · managed care plans · PSH providers
Methodological obstacle: Housing and health systems are analyzed in separate silos.
What would resolve it: Privacy-preserving linkage of placement, retention, health utilization, and mortality data.

UNK-07Critical

Why it matters: Appropriation and activity counts do not establish cost-effectiveness.
Likely record holders: contracting agencies · auditors · providers · HMIS leads
Methodological obstacle: Expenditures, unique persons, exits, retention, and returns are rarely joined at contract level.
What would resolve it: Contract-to-outcome data model with common retention intervals and transparent denominators.

UNK-08Critical

Why it matters: Administrative boundaries can create false narratives about migration, duplication, and responsibility.
Likely record holders: HDIS · CoCs · HMIS leads · counties
Methodological obstacle: Privacy, identifier matching, and fragmented governance.
What would resolve it: State-supported privacy-preserving linkage with transparent governance and research access controls.