How each state pays for disability services

Each jurisdiction is scored out of 100 across five weighted components. Residential funding and spending intensity together carry 60 points, which is what "comprehensive spend" means in practice.

• 35Residential depth. Whether 24-hour provider-operated residential is funded at all (14 points), the breadth of the residential array from group home through supported living, host home, shared living, and family settings (up to 12), whether room and board is supported beyond SSI (up to 5), and whether payment is differentiated by acuity rather than paid flat (4).
• 25Spending intensity. Half from spending per participant, half from total IDD spending per state resident. The second half is denominator-free, which keeps states that count their caseload broadly from being punished for it.
• 20Access. People served per 100,000 residents (12 points) and waiting list pressure, measured as the list against the number actually served (8 points).
• 10Service array and authority. Number of waivers and HCBS state plan authorities, self-direction, and whether state law creates an entitlement.
• 10Supplemental funding. State-only programs with no federal match (up to 6) and county, regional, or local funding streams (4).

Three limits worth naming. First, nine states publish no all-funds IDD total: Michigan, Minnesota, Mississippi, North Carolina, North Dakota, Oklahoma, South Carolina, Virginia, and Wisconsin. Their spending components use the 51-jurisdiction median, and their profiles say so. Their ranks rest on structure and access, not measured dollars.

Second, states count participants differently. California reports its entire Lanterman Act caseload including people who receive only case management. Minnesota, Oregon, Washington, and Wisconsin count across populations broader than IDD alone. Michigan reports waiver slots rather than people served. Every affected profile carries a comparability note.

Third, fiscal years do not line up. Figures range from FY2021 to FY2027 projections depending on what each state publishes. Every number carries its year.

Each agency is scored out of 100 on whether it turns its federal entitlement into services. The score is built only from data that exists for every agency, which is why it is narrower than it could be.

• 60Federal drawdown. Whether the agency relinquished federal funds, drew its own grant and no more, or asked for a share of what other states gave back. Relinquishing is scored against the share of its own two-year allotment handed back, so Guam, which returned 55 percent of two years of funding, scores far worse than Michigan, which returned under 2 percent. Asking for more is treated as evidence of capacity to spend, and an agency whose request went badly underfunded scores highest, because the shortfall is the federal pool's limit rather than the state's.
• 20Access. Order of selection status. An agency serving all eligible people scores full marks; one with categories closed is scored down by how much of the eligible population is shut out. Where an agency does not publish its status, access is dropped from its score entirely and the remaining components are rescaled, rather than imputed. Those agencies are marked "rescaled" in the table. Access is scored for 39 of 78 agencies.
• 20Rate transparency. How many of the seven comparable purchased services the agency publishes an actual price for. This is a proxy for whether providers and families can see what the system pays.

What this score is not. It is not a measure of service quality or employment outcomes. RSA's own site blocks automated access, and the most recent nationally comparable outcome data is program year 2023, reported per jurisdiction rather than per agency, so it could not be scored fairly against the 78-agency structure. Where outcome figures were found on an agency's own site they appear in its profile, dated, and are not comparable across states.

Two structural cautions. First, twenty-one states run a separate agency for blind services with its own award, its own order of selection and often its own rates; those are profiled separately, so a state can appear twice with very different scores. Washington is the clearest case, where the general agency closed all five priority categories in the same year the blind agency drew its full reallotment request. Second, relinquishment is a lagging signal. It reflects whether a prior year's grant was fully obligated, so an agency can end a waiting list and still hand money back, as Indiana did.