EPIT: Every Point in Time

The subject is visibility

Delaware does not have one homelessness reality and one data source. Schools see attendance changes, doubled-up families and transportation instability. Nonprofits see waitlists, turnaways and emergency requests. Hospitals see unsafe discharge risk. Outreach teams see encampments and people avoiding formal systems. Police and first responders see recurring crisis locations. HMIS sees the people and services entered by participating providers.

Each view is useful. None is the whole community. CCCI does not merge every private record into one surveillance file. It creates a governed way to share lawful, privacy-protected signals so decision-makers can see changing need without exposing the person.

What Delaware counted—and did not

That does not mean no one was sleeping outside. It means the 2026 result cannot answer how many people were unsheltered or whether total homelessness rose or fell. The sheltered result was slightly below 2025’s sheltered count, but the years are not fully comparable because one omitted an entire population the other attempted to count.

The system already acknowledges the blind spots

In a presentation hosted by the Delaware State Housing Authority, Housing Alliance Delaware described PIT as an annual snapshot with an undercount of unsheltered people. It described CMIS/HMIS as real-time but limited to participating agencies and identified exclusions including certain state motel-voucher holders, Sunday Breakfast Mission, domestic-violence programs and shelters, and limited unsheltered data.

The presentation’s own conclusion is the essential one: many more people live in temporary or unstable situations and face homelessness risk than are counted as homeless or served by the formal homeless-response system.

Children make the difference impossible to ignore

This figure must not be subtracted from or compared one-for-one with a PIT result. Education law covers a broader set of living situations and measures students across a school year; HUD PIT uses a narrower definition on one night. The difference is the point: government programs encounter different parts of the same instability, yet public decisions are often made as if the narrowest snapshot were the complete picture.

Nationally, SchoolHouse Connection reports that public schools identified more than 1.5 million students experiencing homelessness in 2023–24. Only 4.4% were unsheltered and 12.7% were in shelters; most were staying temporarily with others or in motels—conditions outside HUD’s ordinary PIT methodology.

What EPIT could measure

Public money should buy public capacity

Public records show a dedicated $156,900 HUD HMIS award to Housing Alliance Delaware for FY2023. Delaware’s Council on Housing later approved a separate $230,000 state grant supporting broader CoC administration, including CMIS. Housing Alliance Delaware’s current FAQ lists a $325 annual fee for each CMIS user.

Those amounts cover different years and scopes and do not, by themselves, establish duplication or misconduct. They do establish a fair decision question: if compliant HMIS software and unlimited authorized participation are available for $0, what additional public capacity could Delaware purchase with money no longer needed for commercial licenses and per-seat access?

Trust is part of the instrument

EPIT is not permission to circulate student names, medical records, police narratives or identifiable HMIS cases. Source organizations retain control of their records. Community governance defines approved measures, purpose, access, retention, suppression and disclosure. The shared signal should ordinarily be de-identified, aggregated and no more precise than the decision requires.

Identifiable information moves only through lawful, purpose-bound authorization and role-based access. Schools, police, hospitals and nonprofits do not become one another’s unrestricted databases merely because they contribute to community understanding.

The conclusion