Counts
Use HUD PIT/HIC for the official one-night estimate; use HMIS/HDIS for service-system flow. Publish the universe and period beside every number.
A source map for answering different questions without collapsing a one-night count, a year of service records, a representative survey, an advocacy claim, a provider report, and an investigation into the same kind of evidence.
The strongest reading of the field begins with a measurement decision. The Point-in-Time count estimates a stock on one night. HMIS and California's Homeless Data Integration System describe people who touched participating systems across a longer period. School data cover many children and families living doubled up or in motels. Census housing tables describe the housed population and the pressure surrounding the homelessness system. None is a complete census of everyone who lacks stable housing.
These sources nevertheless converge on a central structural finding: homelessness is geographically concentrated where low-income households face high housing costs, low vacancy, and an inadequate supply of deeply affordable homes. Representative California survey research adds the individual pathways, health conditions, histories, and local ties that administrative series cannot supply. The two levels are complements. Structural scarcity explains why similar personal crises produce different geographic risk; individual evidence explains how that risk is experienced and which supports are needed.
Program and fiscal evidence require a separate chain. An appropriation is not an expenditure. An expenditure is not a completed unit. A completed unit is not necessarily occupied. A placement is not twelve-month retention. A falling street count may reflect housing, shelter, enforcement, movement, method, weather, or several at once. Bay Evidence therefore treats durable exits, fewer entries, shorter episodes, fewer returns, safety, and cost per retained result as the relevant accountability outcomes.
The landscape also contains organized disagreement. Housing First advocates, civil-rights litigators, treatment-centered policy groups, provider networks, and public officials do not occupy the same evidentiary position. Their materials are included because they shape law, funding, practice, and public argument. They are labeled by source class so a policy position is not presented as peer-reviewed research and a press release is not presented as an independent evaluation.
| System | What is counted | Best use | Central limit |
|---|---|---|---|
| HUD PIT / HIC | People observed or sheltered on one January night; beds and units available that night | Comparable point-in-time stock and capacity | Misses hidden homelessness and is sensitive to field method, weather, participation, and coverage |
| HMIS / California HDIS | People recorded using participating homelessness-response programs over a period | Annual service flow, exits, returns, duration, and client characteristics | Does not include every person experiencing homelessness and depends on provider participation and record quality |
| McKinney–Vento education data | Students lacking a fixed, regular, and adequate nighttime residence, including many doubled-up households | Family and student housing instability that PIT frequently misses | A broader statutory universe than HUD homelessness; not directly comparable to PIT totals |
| Census / ACS housing indicators | Housed people and households, income, rent burden, tenure, vacancies, poverty, and demographic context | Housing-market pressure and risk context | Does not adequately enumerate unsheltered people and cannot be presented as a street count |
Use HUD PIT/HIC for the official one-night estimate; use HMIS/HDIS for service-system flow. Publish the universe and period beside every number.
Use research designs capable of testing causal or longitudinal claims. A count, provider dashboard, or press release does not establish why a number changed.
Separate award, obligation, expenditure, opening, occupancy, placement, retention, return, safety, and cost per durable result.
Use the court opinion, statute, regulation, docket, or official guidance as primary authority. Label advocacy organizations' legal interpretation.
Use beat reporting to find documents, affected people, disputes, and implementation failures. Preserve the underlying record whenever possible.
Each entry states the source's proper role and its main evidentiary limit. Links go to the canonical landing page and selected high-value sections reviewed for this update.
Use: Official PIT and Housing Inventory Count series, with national, state, and Continuum of Care estimates and subgroup tables.
Caution: A one-night estimate; unsheltered enumeration is method-sensitive and is not an annual service-use total.
Use: Practitioner hub for PIT/HIC, HMIS, System Performance Measures, Stella, and AHAR materials.
Caution: Tools use different universes and time windows; do not combine them as one series.
Use: Federal strategy, interagency coordination, implementation guidance, and evidence-to-action materials.
Caution: Administration priorities and staffing can change. Archived plans should be identified as historical rather than current policy.
Use: Grant notices, program rules, waivers, press releases, and policy changes affecting CoC and ESG systems.
Caution: Agency announcements are authoritative for policy action, not independent proof of program outcomes.
Use: Entry point to SAMHSA, HRSA, youth, family, Medicaid, health-center, SOAR, and behavioral-health resources.
Caution: A directory across agencies; evaluate the underlying program or study rather than citing the directory as outcome evidence.
Use: Cited primers on federal statutes, appropriations, housing programs, and congressional policy choices.
Caution: Use publication and update dates; CRS explains the federal landscape but does not replace current statutory or regulatory text.
Use: State action plan, HHAP guidance, encampment policy, tribal materials, and statewide coordination.
Caution: Targets are not outcomes. Label baselines, targets, reported results, and time periods separately.
Use: Annualized HMIS records from California CoCs: people served, demographics, first-time homelessness, and exits.
Caution: Counts people who interact with participating systems. Coverage, deduplication, HMIS participation, and missingness differ from PIT.
Use: Downloadable demographic tables for statewide and CoC analysis.
Caution: Inspect data dictionaries, suppression, update dates, and the population universe before comparison.
Use: State-defined measures for exits, returns, length of time homeless, and related system performance.
Caution: Performance measures can change with coverage, record quality, and client mix; they do not by themselves establish causal program impact.
Use: Reported Homeless Housing, Assistance and Prevention allocations, activities, and outcomes.
Caution: Award, obligation, expenditure, service, placement, and retained housing are separate fiscal and outcome stages.
Use: Housing production, planning law, RHNA, funding notices, Homekey, awards, and monitoring materials.
Caution: A funded or announced unit is not necessarily complete, occupied, or retained as affordable housing.
Use: Parent-agency policy, implementation records, dashboards, and Cal ICH materials.
Caution: Cabinet-level summaries should be checked against program records and underlying data.
Use: Official announcements about state actions, preliminary counts, awards, reorganizations, and policy priorities.
Caution: Administration framing is a claim source. Verify count universes, implementation dates, and outcomes with program records.
Use: Representative survey research and policy translation on pathways, health, aging, race, encampments, and services.
Caution: CASPEH is cross-sectional. It supports population estimates and associations, not individual longitudinal causal claims.
Use: State-level empirical analysis of housing, homelessness, drug use, disability benefits, demographics, and public opinion.
Caution: Separate cross-sectional associations from changes over time and from causal identification.
Use: Research on housing supply, land use, construction cost, vouchers, supportive housing, conversions, and subsidy design.
Caution: Policy briefs vary in design and evidence strength; cite the specific study, methods, geography, and period.
Use: Affordable-housing needs dashboards, county reports, preservation risk, subsidy gaps, maps, and finance analysis.
Caution: Institutional policy analysis, not a state agency. Review assumptions behind shortfall and financing estimates.
Use: Research, events, and policy analysis spanning housing, health, governance, and homelessness.
Caution: The topic page aggregates multiple authors and methods; cite the underlying study rather than the institutional landing page.
Use: Research catalog, practitioner–researcher coordination, policy briefs, and Los Angeles system studies.
Caution: Products range from convening materials to original research; label the design and funder of each item.
Use: Housing, health, behavioral health, service delivery, and equity research.
Caution: Distinguish peer-reviewed findings from center news, project descriptions, and training materials.
Use: Economic and policy syntheses on California homelessness, housing markets, and policy responses.
Caution: Policy briefs synthesize selected evidence; use their cited studies for precise causal or quantitative claims.
Use: Poverty, labor-market, safety-net, and inequality research relevant to housing insecurity and prevention.
Caution: Homelessness is one part of a broader portfolio; select only directly relevant studies.
Use: Program evaluation, system research, housing policy, unsheltered counts, and implementation studies.
Caution: RAND reports differ in sponsor, geography, and design. Cite the specific report and funding disclosure.
Use: Evaluations, data tools, housing program trends, supportive-housing research, and state/local policy analysis.
Caution: Tools and reports use different source data and assumptions; avoid treating modeled estimates as administrative counts.
Use: Foundational research and practice resources on families, children, trauma, and service systems.
Caution: Some flagship materials are older; publication date and current applicability must be explicit.
Use: Family and student homelessness data, GIS, shelter-system analysis, methodology, and policy research.
Caution: Many analyses are New York-specific and school definitions are broader than HUD PIT definitions.
Use: Housing-market, place, poverty, zoning, safety-net, and federal policy analysis relevant to homelessness.
Caution: Use specific author, methods, date, and disclosed perspective; the topic page is not a dataset.
Use: Housing affordability, state policy, and explanatory analysis connecting housing costs with homelessness.
Caution: Pew Research Center and Pew Charitable Trusts are distinct entities; identify the actual publisher and methods.
Use: Federal research, technical assistance, and contracted production of homelessness and housing evidence, including AHAR-related work.
Caution: A contractor's work may be official when published by an agency, but the agency release and methods remain the controlling record.
Use: Annual state dashboards, policy analysis, unsheltered research, workforce studies, and research translation.
Caution: A research-and-advocacy hybrid. Use HUD or local primary data for official counts and label Alliance interpretation.
Use: Lived-experience leadership, civil-rights advocacy, grassroots networks, and public education.
Caution: Use as advocacy and lived-experience evidence, not as a substitute for official administrative totals.
Use: Litigation, legal briefs, criminalization research, education rights, model policies, and case tracking.
Caution: Legal analysis reflects an advocacy position; cite court opinions, statutes, and dockets as primary legal authority.
Use: Clinical guidance, medical respite, health-center practice, research, and health policy.
Caution: Clinical guidance and policy advocacy answer different questions; note evidence grade and target population.
Use: Runaway and homeless youth policy, youth participation, training, research summaries, and federal advocacy.
Caution: Youth definitions and age ranges differ across HUD, education, child-welfare, and provider systems.
Use: McKinney–Vento implementation, student homelessness data, higher-education access, FAFSA, and practitioner tools.
Caution: School homelessness includes some doubled-up and motel situations outside HUD's PIT universe.
Use: Veteran homelessness policy, provider capacity, technical assistance, and interpretation of federal veteran data.
Caution: Use HUD and VA releases for official counts and label coalition recommendations as advocacy.
Use: LGBTQ+ youth homelessness research translation, toolkits, training, and systems-change advocacy.
Caution: Prevalence estimates depend on youth definition, sampling frame, and whether the measure covers instability beyond HUD homelessness.
Use: Origin, fidelity, implementation, and training materials for the Pathways Housing First clinical model.
Caution: A model-origin source is useful for defining the intervention; independent evaluations are stronger for comparative effectiveness.
Use: Funding strategy, policy frameworks, equity guidance, and philanthropic coordination.
Caution: Funding guidance is not a neutral effectiveness review; disclose philanthropic and policy perspective.
Use: Arguments for treatment, camping restrictions, performance contracts, and alternatives to Housing First dominance.
Caution: Use as a clearly labeled policy position. Verify factual claims against primary data and independent research.
Use: State legislation, coalition policy, Roadmap Home, conference materials, and resident organizing.
Caution: Coalition priorities and legislative positions are advocacy; cite official bill text and fiscal analyses for legal effects.
Use: Regional housing policy, research, developer network, legislation, and production-gap analysis.
Caution: Member and sector perspective should be explicit; verify production and finance claims with primary records.
Use: Youth homelessness policy, crisis-line resources, legislative advocacy, and provider coordination.
Caution: Service and advocacy materials are valuable but do not replace independently evaluated outcomes.
Use: Network-level information on prevention, rental assistance, shelter, immigration, and diocesan services.
Caution: Programs are locally operated and not uniform; verify services and results with the relevant member agency.
Use: Regional housing-security strategy, policy, data framing, and cross-sector coordination.
Caution: Regional strategy and self-reported initiatives require project-level verification for outcome claims.
Use: System strategy, philanthropic investment, policy, and homelessness-prevention and housing initiatives.
Caution: Separate grantmaking, program operation, claimed placements, and independently measured county outcomes.
Use: Outreach, shelter, interim housing, supportive housing, and service-program information.
Caution: Provider reports describe operations and self-reported results; independent evaluation needs a comparison design and retention data.
Use: Housing, shelter, outreach, and support-program descriptions across multiple counties.
Caution: Verify program geography, contract terms, enrollment universe, exits, and housing retention.
Use: Shelter, outreach, interim housing, and housing-program information.
Caution: Operational claims should be reconciled with county contracts, HMIS definitions, and retention measures.
Use: Housing, health, outreach, shelter, domestic violence, and supportive-service information.
Caution: Use for program description and provider perspective; use independent records for comparative effectiveness.
Use: Medical respite, health services, housing, and family-program information.
Caution: Clinical and housing outputs need clear denominators, follow-up periods, and independent validation.
Use: Shelter, navigation, supportive housing, workforce, and behavioral-health program information.
Caution: Provider program pages establish service design, not countywide causal impact.
Use: Outreach, shelter, housing, and service-system program information.
Caution: Distinguish provider outputs from system-wide outcomes and verify retention windows.
Use: Discovery of organizations, filings, and sector lists for follow-up research.
Caution: A lead-generation directory, not a quality rating or evidence source. Verify against IRS filings and organization records.
Use: Discovery of funders, grantees, and nonprofit information.
Caution: Directory inclusion and rankings are not program evaluation; verify financial claims with original filings.
Use: Sustained state-policy reporting, legislative tracking, data explanation, investigations, and public-record reporting.
Caution: Journalism is secondary evidence; follow links to statutes, audits, budgets, and datasets for core factual claims.
Use: Bay Area counts, housing policy, courts, public systems, and reported lived experience.
Caution: Use linked official releases and source documents for series values and legal holdings.
Use: Long-form investigations, local system accountability, count reporting, budgets, and historical analysis.
Caution: Paywall and corrections status should be recorded; underlying documents remain preferable for reproducible claims.
Use: San Francisco and regional housing, homelessness, encampment, budget, and governance reporting.
Caution: Distinguish reported preliminary figures from final agency releases and corrected datasets.
Use: County counts, housing development, regional policy, litigation, and local accountability reporting.
Caution: Verify syndicated copies, dates, updates, and source-document links.
Use: State budget, legislation, executive action, Sacramento systems, and investigative reporting.
Caution: Use legislative text, fiscal analyses, and agency records for final policy details.
Use: State policy and Sacramento-region housing and homelessness reporting.
Caution: Confirm whether pages are current, archived, corrected, or produced with a partner newsroom.
Use: Local government, homelessness counts, encampments, shelters, housing, and service-system reporting.
Caution: For count and budget claims, retain the underlying LAHSA or government document.
Use: National count coverage, multi-state reporting, breaking policy news, and investigations.
Caution: Wire versions can be republished or shortened; cite the canonical AP version and update timestamp.
Use: National investigations, policy analysis, local case studies, and narrative reporting.
Caution: Use as reporting and synthesis; validate core quantitative claims with linked primary sources.
Use: Federal policy, national investigations, courts, housing, and local-system reporting.
Caution: Topic coverage is intermittent; record access, publication, update, and correction dates.
Use: Long-form argument, policy synthesis, and national context.
Caution: Analysis and reported essays are not administrative data; inspect cited studies and distinguish argument from finding.
Use: Public-record investigations into housing, displacement, health systems, policing, and government accountability.
Caution: Use supporting documents and methodology when available; a single investigation may not generalize beyond its cases.
Use: Cross-state policy comparison, legislation, budgets, and implementation reporting.
Caution: State rules and definitions differ; follow state-specific links before comparing outcomes.
Use: State and local policy, management, finance, housing, and public-administration reporting.
Caution: Trade and analysis coverage should be separated from peer-reviewed evaluation and official records.
Use: Local-government operations, federal grants, state policy, infrastructure, and public management reporting.
Caution: Use as a lead and contextual source; confirm program claims in agency documents.
Use: Litigation and advocacy concerning sweeps, property seizure, disability, due process, and policing.
Caution: For legal holdings, cite the docket and opinion. Complaints state allegations, not adjudicated facts.
Use: Litigation and advocacy on encampments, property destruction, policing, disability, and civil rights.
Caution: Separate allegations, preliminary orders, settlements, and final judgments.
Use: Jumping-off index for advocacy organizations and local networks.
Caution: Uneven and not quality-rated. Verify organizations, links, status, and claims independently.
Use: Training on homelessness, housing, behavioral health, recovery, equity, and cross-system practice.
Caution: Training materials summarize evidence; use cited research and current federal guidance for formal claims.
Use: Medi-Cal policy for housing-related community supports, recuperative care, transitions, and local plan implementation.
Caution: An authorized benefit is not the same as uptake, adequate provider capacity, or stable housing; use plan-level utilization data.
Use: McKinney–Vento implementation, local liaisons, student data, guidance, and educational rights.
Caution: The education definition includes living arrangements not counted in HUD PIT; never merge the series without labels.
Use: McKinney–Vento law, data, briefs, training, and liaison resources.
Caution: Education-system counts answer a different question from HUD sheltered and unsheltered estimates.
Use: Veteran housing, outreach, health, prevention, SSVF, and program-performance information.
Caution: VA program contacts, service-use records, HUD veteran PIT counts, and long-term housing retention are distinct measures.
Use: Plain-language presentation of federal data and trends for public orientation.
Caution: A secondary aggregator. Cite HUD or the underlying agency release for authoritative values and revisions.
Use: Controlling 2024 Supreme Court opinion concerning enforcement of generally applicable public-camping laws under the Eighth Amendment.
Caution: The decision does not answer every due-process, disability, property, state-law, or local-policy question raised by encampment enforcement.
Use: Nine-part framework covering measurement, prevention, housing-led responses, services, governance, finance, and evaluation.
Caution: Institutional arrangements and definitions vary across countries; international lessons require U.S. and California translation.
Use: Global definition framework, data map, count methodologies, city networks, publications, and resource hub.
Caution: Cross-national estimates may not share definitions, enumeration methods, or administrative coverage.
Use: Framework distinguishing rooflessness, houselessness, insecure housing, and inadequate housing.
Caution: ETHOS is not a U.S. legal definition and should not be substituted for HUD or McKinney–Vento eligibility rules.
The expanded stack improves question selection more than it increases the number of defensible outcome claims. It is already sufficient for regional PIT comparisons, stock-versus-flow explainers, family and subgroup briefs, an allocation-versus-outcome audit, and analysis of the cost-burden reservoir around the system. It is not sufficient for provider rankings or causal claims about a single grant, sweep, shelter, or housing pipeline.
The next evidentiary gains are project-level expenditure records, occupancy and opening dates, HMIS exit and return cohorts, shelter turnaways, eviction and displacement series, contract performance definitions, and housing-quality evidence after placement. Those records turn a broad literature review into accountable local reporting.
For current reporting, every claim should preserve its source class, universe, geography, period, retrieval date, and limitation. That discipline is the difference between a directory of reputable institutions and an evidence system.