Date: 2026-08-06. This late pass adds a prospective criterion; it does not re-score the board.
Criterion
A targeted rebalancing package is supported if credible quasi-experimental evidence finds HCBS workforce growth or less institutional use without merely shifting institutional workers. It is not a test of wage floors, generic HCBS spending, or individual programme components.
Result: narrowly supported for BIP-style packages
- A 2026 Health Services Research stacked DID using ACS 2005–21 found that Balancing Incentive Program (BIP) adoption increased HCBS workforce by 13.24%, with no significant institutional-workforce effect. It does not identify a wage-floor mechanism and target-state evidence remains subgroup, not random-assignment evidence. The other half of the same paper, added on Phase 2 verification (2026-08-10, steelman-log): its title is Evaluating the Affordable Care Act's LTSS Rebalancing Programs — plural — and it evaluates two. Alongside BIP's 13.24% it finds no statistically significant effect for Community First Choice (§1915(k)): 1.51%, 95% CI −12.77% to 15.79%. CFC is the statutory, self-direction-shaped attendant-services expansion carrying a 6-percentage-point FMAP bonus (42 U.S.C. §1396n(k)(2)), which makes it the closest enacted analogue to both the HCBS de-capping row and the Cash & Counseling row. This file and the scorecard previously carried only the positive arm. The paper's own conclusion is the design point, not the spending point: "additional funds [should be] tied to measurable performance targets."
- A 2025 BMJ Open generalized DID comparing BIP completer states with eligible nonparticipants found about five percentage points fewer long nursing-home stays for Medicaid beneficiaries living alone. HRS does not directly observe HCBS use, and concurrent MLTSS adoption remains a confound.
These studies support the bundled, target-bound BIP design, not a generic claim that more HCBS money automatically creates staff or saves money. A 2025 state-year study found no significant wage association with HCBS spending, while a 2023 DID found only partial institutional offset: each additional dollar of HCBS spending was associated with $0.74 of added total LTSS spending. The self-financing claim is therefore not supported.
The rest of that study, added on Phase 2 verification (2026-08-10, steelman-log). McGarry & Grabowski, J Am Geriatr Soc 71(10):3143–3151, is titled Medicaid home and community-based services spending for older adults: Is there a "woodwork" effect? — and answers no: "We did not find evidence of a woodwork effect in those states that expanded Medicaid HCBS more aggressively." It reports the $0.74 as an offset, not a loss ("each dollar directed to HCBS was offset by $0.26 savings from decreased nursing home use"), plus reductions of 47.1 nursing-home residents and $7.3M institutional LTSS spending per 1% HCBS increase, and "more older adults receiving LTSS at a lower cost per beneficiary served." Both readings are true of the same coefficient; this file previously carried only the one that qualifies the architecture downward. The woodwork effect is the standard forty-year objection to HCBS expansion and the word appears nowhere else in this filing — the record holds its best available refutation and never named it.
A 2026 IV/border study found HCBS use reduced Medicare community-initiated home-health use by 1.02 percentage points. This is relevant cross-payer substitution, not evidence of total Medicare-plus-Medicaid savings. MACPAC's 2025 rate-setting review explains an important measurement gap: states use inconsistent compensation inputs and a public wage series is not yet required.
Scorecard boundary
The HCBS row has stronger capacity/autonomy evidence than an ARPA-only record, but it must not be re-scored without the required independent re-score. Preserve the partial-offset cost caveat and do not infer that paid-family-caregiver policy expands the agency workforce.
Discharged 2026-08-10. The structurally blinded re-score (ws11-rescore-log.md) supplied the required independent pass. The HCBS row's workforce cell moved 2→4 on the BIP stacked-DID finding above, held below 5 on this section's own boundary — the evidence is for the bundled, target-bound BIP design, not for generic HCBS money creating staff, and the scorecard's basis column now carries that scope. The partial-offset cost caveat is preserved (CC held at 2), and the paid-family-caregiver row's workforce cell was held at neutral for exactly the reason stated here.