# Patch Notes #002: Housing / Homelessness Evidence Matrix

Version: v0.1 public pre-memo matrix
Date: 2026-07-03
Status: Source-anchor matrix for reviewer critique. This is not a final evidence review.

## Working Standard

Every major claim should separate affordability, homelessness, population segment, local market, intervention type, evidence strength, and implementation risk.

| Claim | Evidence Strength | Initial Source Anchor | Reviewer Ask |
| --- | --- | --- | --- |
| U.S. homelessness rose sharply in the 2024 PIT count, but PIT is a point-in-time estimate with known limitations. | Strong for PIT count; limited for hidden homelessness | HUD AHAR / PIT | How should PIT limitations be explained without minimizing visible homelessness? |
| Housing affordability pressure affects renters and homebuyers. | Strong | Harvard Joint Center for Housing Studies | Which affordability metrics are least misleading? |
| The United States remains undersupplied by millions of housing units. | Moderate/strong nationally; must be localized | Freddie Mac housing supply estimate | How should national shortage estimates be translated into local plans? |
| Low-wage renters cannot afford modest rental housing in many markets. | Strong for wage-rent gap framing | National Low Income Housing Coalition, Out of Reach | What does this measure capture and miss? |
| Zoning and permitting constraints can limit supply, but supply alone does not solve homelessness for the lowest-income and highest-need groups. | Directionally strong; local causal strength varies | JCHS, Brookings, local planning literature | Where are zoning claims strongest, weakest, or too broad? |
| Homelessness is not one population. Families, chronically homeless adults, youth, veterans, domestic-violence survivors, and people exiting systems need different interventions. | Strong/moderate | HUD AHAR categories and homelessness service literature | Which segmentation helps implementation, and which creates stigma or blind spots? |
| Housing First and permanent supportive housing can be effective for some high-need populations, but implementation capacity and local context matter. | Moderate/strong for housing stability; varies by outcome and population | HUD Exchange, USICH, supportive housing evaluations | Where should the paper be more precise about outcomes? |
| Eviction prevention can reduce displacement risk in some cases, but evidence varies by targeting, timing, legal context, and rental market. | Moderate; program-specific | Eviction Lab, Urban Institute, local evaluations | Which program designs have the strongest evidence? |
| Local land-use rules, permitting delays, infrastructure constraints, and neighborhood politics can shape production. | Strong directionally; local impact varies | Brookings, White House housing supply materials, local planning research | What local blockers are most often underestimated? |
| Supply, subsidy, services, prevention, and operations solve different parts of the housing/homelessness problem. | Synthesis claim; requires reviewer validation | Cross-source synthesis | Does this five-lane frame improve clarity or oversimplify? |

## Research Gaps Before Memo v0.2

1. Zoning/permitting causal estimates and where they are strongest.
2. Housing First and permanent supportive housing evidence by population and outcome.
3. Shelter, transitional housing, rapid rehousing, permanent supportive housing, and prevention comparisons.
4. Construction cost, land cost, labor, materials, interest-rate, and financing breakdown.
5. Local fiscal incentives, infrastructure constraints, neighborhood political constraints, and permitting timelines.
6. Domestic violence, youth, family, veteran, chronic homelessness, and system-exit segmentation.
7. Voucher utilization, landlord participation, inspection timelines, and lease-up failure points.
8. First geography choice for deeper policy memo: national overview, California, New York, Los Angeles, a high-growth Sun Belt metro, or smaller-city case study.