Point-in-Time homelessness estimates
U.S. Department of Housing and Urban Development
In a RepVector build, this dataset describes the built environment and residential stability of a territory.
| Geography | CoC Number | CoC Name | Year | Overall homeless | Sheltered |
|---|---|---|---|---|---|
| Pennsylvania (state) | — | — | 2025 | 14,834 | 11,952 |
| Philadelphia CoC (CoC) | PA-500 | Philadelphia CoC | 2025 | 5,516 | 4,338 |
| Harrisburg/Dauphin County CoC (CoC) | PA-501 | Harrisburg/Dauphin County CoC | 2025 | 347 | 207 |
| Upper Darby, Chester, Haverford/Delaware County CoC (CoC) | PA-502 | Upper Darby, Chester, Haverford/Delaware County CoC | 2025 | 472 | 402 |
| Wilkes-Barre, Hazleton/Luzerne County CoC (CoC) | PA-503 | Wilkes-Barre, Hazleton/Luzerne County CoC | 2025 | 195 | 173 |
| Lower Merion, Norristown, Abington/Montgomery County CoC (Co | PA-504 | Lower Merion, Norristown, Abington/Montgomery County CoC | 2025 | 534 | 475 |
| Chester County CoC (CoC) | PA-505 | Chester County CoC | 2025 | 313 | 287 |
| Reading/Berks County CoC (CoC) | PA-506 | Reading/Berks County CoC | 2025 | 682 | 503 |
What this dataset is
Annual HUD Point-in-Time (PIT) count of people experiencing homelessness, published at the Continuum of Care (CoC) and state levels. Covers total homeless, sheltered, unsheltered, age, veteran, and chronic homelessness subpopulations where reported by HUD. It is published by U.S. Department of Housing and Urban Development and can be queried at state level, refreshed annual (january count, released mid-year). Each row is an aggregate statistic for a geography, so results describe a market as a whole rather than a single account.
Why it is useful
- Signals growth corridors before service capacity catches up with them.
- Adds cost-of-living and tenure context that shapes access to services.
Questions it helps answer
- Which parts of this market are adding households fastest?
- Does housing cost pressure explain access gaps here?
How different verticals use it
RepVector is vertical-agnostic: this dataset is selected by what your question asks for, not by industry. These are common ways teams put it to work.
Quantifies unstable-housing pressure that drives service demand.
Reads supply and cost pressure at the geography you are underwriting.
Explains social drivers behind avoidable utilisation in a market.
Grounds affordability assumptions in published housing data.
How RepVector uses it
- You describe a market in plain language. The geography resolver turns the place you named into the exact identifiers this dataset needs — city to county to state, plus DMA where relevant — so a query at state runs without you knowing any codes.
- The planner selects Point-in-Time homelessness estimates only when your question matches what it actually reports. It is never included to pad a result.
- Returned rows are normalised into the shared metric layer, so figures from U.S. Department of Housing and Urban Development line up with every other source in the same dashboard instead of sitting in an isolated table.
- Quality checks score coverage and recency. If U.S. Department of Housing and Urban Development suppresses or omits a value for your geography, the gap is reported as a gap — nothing is estimated or filled in.
- The result renders as charts, KPIs and, where the rows are entities, prioritised targets you can export to PDF, Excel, CSV or JSON.
What it does not tell you
- Describes structures and tenure, not service demand directly.
| Field | Key | Type |
|---|---|---|
| Geography | geography | string |
| CoC Number | cocNumber | string |
| CoC Name | cocName | string |
| Year | year | string |
| Overall homeless | overall | number |
| Sheltered | sheltered | number |
| Unsheltered | unsheltered | number |
| Individuals | individuals | number |
| People in families | families | number |
| Under 18 | under18 | number |
| Age 18-24 | age18to24 | number |
| Over 24 | over24 | number |
| Veterans | veterans | number |
| Chronically homeless | chronic | number |
How it fits with the rest of the catalog
Housing growth is an early indicator; population and facility data confirm whether service followed. RepVector queries every selected dataset against the same resolved geography and time window, then normalises the results into one metric layer so they can be charted side by side. These datasets are the most common companions to Point-in-Time homelessness estimates:
Other housing datasets
Common questions
What does Point-in-Time homelessness estimates contain?
Annual HUD Point-in-Time (PIT) count of people experiencing homelessness, published at the Continuum of Care (CoC) and state levels. Covers total homeless, sheltered, unsheltered, age, veteran, and chronic homelessness subpopulations where reported by HUD. It is published by U.S. Department of Housing and Urban Development at state grain.
How often is it refreshed?
U.S. Department of Housing and Urban Development publishes on a annual (january count, released mid-year) cycle. RepVector re-reads the source on that cadence and keeps every prior capture, so you can see what changed.
Do I need an account to use it?
No. Browsing the catalog and the free lookup tools needs no account. Running a full market build across every county in a territory is the paid product.
Build the list for one county and see whether the names are right.
Ask in plain language. RepVector finds the universe, enriches each target from its own website, scores it 0–100, and hands you a ranked call list with phones, addresses and named contacts.
3 free runs — no card required.