Housing Continuum Planning

When to Use This Skill

Use this skill to map a community's housing ecosystem, quantify the gap between housing need and supply by AMI band and tenure, and turn that map into "where we play" portfolio recommendations for a nonprofit board. Typical triggers: "we're writing a housing strategy — where is the gap in our county?", "how many of our renters are cost-burdened at 30% AMI?", "should our org get into permanent supportive housing?", "there's no path to ownership here — is that a gap we should fill?", "should we hand our transitional program to another agency?"

Boundary with siblings: - If the board process is a full org-wide strategic plan where housing is one pillar, run nonprofit-strategic-planning and use this skill only for the housing-portfolio analysis inside it. - If the user needs general community needs-assessment methodology (surveys, focus groups, indicator dashboards across health/education/housing), use nonprofit-needs-assessment. This skill is specifically about housing need/supply gaps on the continuum. - Once a decision is made to pursue a specific deal, feasibility, market study, site control, and capital stack are nonprofit-housing-development-finance. - Designing or operating the programs at a chosen stage: street outreach/shelter/diversion/coordinated entry is nonprofit-housing-homelessness-services; rapid rehousing and transitional housing operations are nonprofit-housing-rapid-rehousing-transitional; PSH design and Housing First fidelity are nonprofit-housing-permanent-supportive-housing; affordable rental operations are nonprofit-housing-affordable-rental-operations; ownership programs are nonprofit-housing-homeownership-programs; repair/preservation is nonprofit-housing-repair-preservation; CLT/shared-equity stewardship is nonprofit-housing-community-ownership. - Housing-market advocacy and zoning reform that emerges from the gap map is nonprofit-housing-advocacy-land-use.

Framework 1: The Housing Continuum as an Organizing Map

The continuum orders every housing intervention by household stability and permanence, from crisis response to market participation:

1. Homelessness services — street outreach, diversion/prevention, coordinated entry (no beds).
2. Emergency shelter — crisis beds, 30-90 day stays, low or no barrier.
3. Transitional housing / rapid rehousing (RRH) — time-limited housing with services (TH) or short-term rental assistance without a project bed (RRH, typically ESG- or CoC-funded, 1-2 year subsidy taper).
4. Permanent supportive housing (PSH) — permanent affordable housing with voluntary wraparound services, targeting chronically homeless and disabled households, Housing First model.
5. Subsidized affordable rental — LIHTC units, HOME/CDBG-assisted units, public housing, Housing Choice Voucher (HCV) tenancies, mission-driven nonprofit rental.
6. Market rental — unsubsidized, rent set by the market.
7. Affordable homeownership — shared equity, down-payment-assisted, or full-market ownership paths for households priced out of market ownership.

Two rules make the continuum a planning tool rather than a diagram:

• People flow both directions. A household at 25% AMI does not "graduate" to 60% AMI units in a tight market; the map must show choke points, not a one-way ladder. The most common structural failure is a community rich in shelter and rich in 60-80% AMI product with almost nothing in between — households exit shelter into homelessness, not into PSH or deeply affordable rental.
• Count every stage, including ones you don't operate. A gap map with an org-shaped hole (stages where the org has no visibility) is advocacy, not analysis. Inventory the whole ecosystem: public agencies, other nonprofits, for-profit LIHTC developers, PHAs, and informal supply (family doubling-up, which CHAS and PIT both miss).

Practitioner vs. advisor: a practitioner (housing org staff) builds the map to steer their own portfolio and to speak credibly in the local Consolidated Plan process. An advisor/consultant builds the same map but must also pressure-test the client's self-definition — boards describe what they do by program name, not continuum stage, and half the engagement is translating.

Framework 2: AMI Bands and the 30%-of-Income Standard

AMI (Area Median Income) is HUD's estimate of median family income for a metro area or non-metropolitan county, adjusted by household size and updated annually. Every affordability conversation in US housing resolves to AMI bands:

• ≤30% AMI — extremely low income (ELI). HUD defines ELI as the greater of 30% AMI or the federal poverty guideline.
• 31-50% AMI — very low income (VLI); the HCV/public-housing eligibility world.
• 51-80% AMI — low income; the LIHTC standard (the Section 60%-of-40%-at-50% test makes LIHTC buildings effectively serve this band).
• 81-120% AMI — "workforce"/middle income; served by market product or modest subsidy.
• Above 120% AMI — market.

The 30% standard: a household is cost-burdened when housing (rent or ownership costs plus utilities) exceeds 30% of gross income; severely cost-burdened at 50%. This is the federal benchmark embedded in CHAS data, in the Brooke Amendment rent ceiling for public housing, and in virtually every needs assessment. Compute the affordability line per band: at 30% AMI, an "affordable" rent is 30% of that band's income, roughly the Section 8 standard of rent-plus-utilities equal to 30% of adjusted income (tenant-paid share is income-based, not unit-based).

Two traps: - AMI for a 1-person household is not the 4-person AMI you see in the newspaper. Always cite the income limit for the actual household size, and never mix HUD AMI with state or city income-limit systems (some states and municipal programs publish their own, which differ). - "Affordable" is not "available." A unit affordable to a 30%-AMI household may be occupied by a higher-income household. Gap analysis must count affordable and available units (NLIHC's distinction), not just rents below the threshold.

Framework 3: Gap Analysis — Sources and Method

Build the needs/supply picture from four data layers, in this order:

1. HUD CHAS data — the Comprehensive Housing Affordability Strategy tabulation, HUD's special tabulation of American Community Survey data (5-year estimates), published by HUD PD&R for every state, county, place, and tract. CHAS gives households by income band (0-30%, 30-50%, 50-80%, 80-100%, 100%+ AMI) cross-tabbed by tenure, cost burden, overcrowding, and housing problems — this is the spine of the AMI-band needs table, and the same data the local Consolidated Plan uses, so your numbers reconcile with the jurisdiction's.
2. HUD CPD documents — the jurisdiction's 5-year Consolidated Plan and annual Action Plans (on HUD Exchange) contain the region's own priority needs, plus counts of HOME/CDBG/ESG-funded units. If your target community has a Consolidated Plan, mine it before collecting anything; a gap map that contradicts the ConPlan needs a reason why.
3. PIT and HIC — the Point-in-Time count (one-night census of sheltered and unsheltered homeless persons, each January, required by HUD for each Continuum of Care) and the Housing Inventory Count (beds by program type: ES, TH, RRH, PSH, safe haven). Together they size the homelessness-services end of the continuum. PIT is an undercount with known day-to-day variance — treat it as a floor, compare 3-5 years of trend, and note the CoC geography rarely matches your service-area geography.
4. State/local/regional needs assessments — state HFA housing needs studies, regional housing needs allocations (e.g., RHNA-style regional plans), city housing element gap analyses, university/PPRI studies. Layer these for the 80-120% AMI and ownership bands CHAS underexplains, and for local market dynamics (job growth, rent trends).

Supplementary sources for supply-side counts: HUD's LIHTC database and state HFA allocation plans for subsidized units, the National Housing Preservation Database for expiring-use risk, HUD TRACS/public housing inventory for deep-subsidy units, and NLIHC's The Gap and Out of Reach reports for national/state benchmarks (the 2026 Gap report finds a shortage of 7.2 million affordable and available rental homes for ELI renter households nationwide, ~35 per 100 ELI renters — useful to anchor how bad local numbers are against the national picture).

The AMI-band needs/supply table — the core analytical deliverable. For each AMI band, by tenure (renter/owner):

Read the table diagonally as well as by row: households priced out of their band overflow downward, so a 50-80% AMI shortage shows up as cost burden at 30-50% AMI.

Framework 4: Where-We-Play Decisions

Once the gap map exists, the board question is: enter a stage, expand, exit, or partner? Score each candidate move against four tests, and require all four before recommending entry or expansion:

1. Mission fit — does the population in the gap match the org's mission population and theory of change?
2. Gap evidence — is the gap documented (your table), durable (not a one-year market wobble), and unclaimed (no other provider is competently filling it — duplication is a failure, not competition)?
3. Right to win — credible access to the operating model: for PSH, service capacity and CoC/HMIS standing; for LIHTC rental, a development partner or developer affiliate; for ownership programs, lending partners and stock to buy; for shelter, land and neighborhood acceptance.
4. Financial sustainability — an operating pro forma for the stage's dominant funding stream (CoC grants for PSH/RRH, project-based vouchers for deep rental, buyer-fee models for ownership). Do not recommend entering a stage whose funding stream the org has never raised; flag it as a prerequisite.

Exit and partner are real options, not failures. A shelter provider with full occupancy and strong RRH outcomes but no PSH pipeline should often partner (referral MOU with a PSH operator) rather than build PSH. Exiting a stage: recommend exit when (a) the gap has closed or moved, (b) another provider has clear comparative advantage, or (c) the funding stream has structurally eroded — and pair every exit recommendation with a transition plan for participants, staff, and funders. Never recommend an exit from a stage where the org is the sole provider without naming the successor.

Practitioner vs. advisor: the practitioner presents the decision memo to their own board with a recommendation. The advisor presents the same evidence but should hold the recommendation loosely — the classic consultant failure is recommending entry into the stage the client's ED already wanted. Test alternatives explicitly in the memo.

Deliverable 1: Continuum Gap Map

1. Define the geography (county, city, or CoC) and state why; note where CoC, county, and city boundaries diverge — this changes PIT numbers.
2. Inventory every provider at every stage: programs, unit/bed counts, target population, waitlist if obtainable.
3. Overlay flows: where do households go when they exit each stage (ask providers for exit-destination data; HMIS reports it for CoC-funded programs)?
4. Mark choke points (exits to homelessness, waitlists >1 year, bands with zero product) and surpluses/duplication.
5. Draw the one-page map: continuum across the top, providers and bed counts underneath, gap flags in bold. Done when: a reader can name the community's two worst choke points without asking, and every stage has a number (including zero) with a source.

Deliverable 2: AMI-Band Needs/Supply Table

1. Pull latest CHAS 5-year data for the geography (state the vintage; CHAS lags the ACS by ~2 years).
2. Build the table by band and tenure with the columns above; compute the affordable-and-available line, not just affordability.
3. Add PIT/HIC counts under the ≤30% AMI renter column as a cross-check (CHAS misses sheltered/unsheltered households in group quarters).
4. Add pipeline and expiring-use risk (National Housing Preservation Database) so the gap is net.
5. Write a 3-5 finding narrative: the bands where the gap is structural, the bands that are actually fine. Done when: every number has a named source and year, and the table reconciles with the Consolidated Plan's cited figures (or explains the discrepancy).

Deliverable 3: Where-We-Play Decision Memo

Structure (keep to 5-7 pages):

1. The gap in one page — the map plus the table's headline findings.
2. Our current position — the org's stages, scale, performance, and financial exposure by funding stream.
3. Options considered — enter / expand / exit / partner / hold for each candidate stage, each scored against the four tests. Include at least one option the board will resist.
4. Recommendation — one primary move, with prerequisites (funding, partners, HMIS/coordinated-entry standing, board policy changes) and what it costs to be wrong.
5. What we will not do — explicit stop/de-prioritize list; a play memo with no subtraction has not decided anything.
6. First 12 months — the 3-4 milestones that make the move real or kill it (e.g., "execute MOU with the PSH operator" not "explore PSH"). Done when: the board can vote on a single motion naming the move, the prerequisites, and the stop list.

Advisor Framing

• Scope data work separately from the decision memo — clients can build the gap table in-house with this skill and hire you only for the board process.
• Present the map to the full ecosystem (CoC, city housing staff, other providers) before the client's board sees recommendations; a recommendation built on other providers' data that they haven't seen breeds coalition damage that outlasts the engagement.
• In Consolidated Plan cycles, time the engagement so findings land during the jurisdiction's public-comment window — one engagement can shift HOME/ESG allocations, which is often the real motive behind commissioning the map.

Common Failure Modes

• Org-shaped map: only stages the client operates are analyzed; gaps where the client could never play are ignored, so the board "decides" to keep doing what it does. Fix: inventory every provider before opening any option discussion.
• Affordable ≠ available error: counting units with low rents as gap-closers without checking who occupies them. Fix: always compute the affordable-and-available line.
• CHAS vintage and geography mismatches: comparing a 2023 5-year CHAS against last night's shelter count, or PIT (CoC-wide) against city boundaries. Fix: date-stamp every source in the table; convert to one geography and state the method.
• One-way-ladder thinking: presenting the continuum as a graduation pipeline when the actual flow is households stuck at the bottom. Fix: show exit-destination data and choke points explicitly.
• Enter-bias: the board always wants to add a stage; partner/exit options are never scored. Fix: require every option paper to include a partner and an exit scenario.
• Uncosted recommendation: "we recommend entering PSH" with no operating-funding analysis — PSH without committed services funding becomes an unaffordable rental property. Fix: a stage-entry recommendation without its funding stream analysis is returned, not circulated.
• Funder-grade overreach: the memo quietly becomes a fundraising pitch, dropping the inconvenient bands. Fix: keep the findings narrative and the recommendations in separate documents with separate review.