The Only Big Test of Duals Integration Raised Spending. Decks Skip That Slide.
Dual-eligible members are the most discussed growth market in managed care, and Medicare-Medicaid "alignment" is the most repeated word in the pitch. The public record says something more interesting than the pitch. Six things from building an instrument for it.
1. The experiment already ran, and the decks lost. The Financial Alignment Initiative was the real test — passive enrollment, capitated three-way contracts, a decade of runway, official independent evaluations. As synthesized by MedPAC: eight of eleven demonstrations produced statistically significant increases in Medicare spending, 3.1 to 9.8 percent. None showed savings. Beneficiary satisfaction was genuinely high — the record cuts both ways, and honesty means carrying both directions. Integrate for coordination or experience if you choose; do not book it as a cost play.
2. Outcome claims travel far ahead of outcome evidence. MedPAC's words for the integration outcomes literature: limited and inconsistent. Quality scores run fairly similar across plan types. The 83-study inventory kept by MACPAC draws no overall conclusion. An integration business case with an outcomes slide needs a citation under it, and there are very few to lend.
3. The one claim with real evidence is retention — carried with its label. FIDE plans lose about 8% of members in year one against 28% for standard Medicare Advantage. That gradient is the best-evidenced fact in the record — and it is associational: who selects into fully integrated plans is uncontrolled. The honest use is a hypothesis to test on your own book, by tier. The dishonest use is booking the literature's spread as your program's effect.
4. Alignment is a ledger, not an adjective. Exclusively aligned enrollment ran 7 to 8 percent of D-SNP membership nationally; only about a quarter of duals sit in overlapping managed care in the same month. Most books described as "integrated" are, when counted, coordination-only with a fully-integrated garnish. Count before claiming — the tiers are legal definitions with teeth, and since 2025 the fully-integrated one requires capitated long-term services, behavioral health, and exclusively aligned enrollment.
5. The member breaks at the Medicaid end. During the redetermination unwinding, one dual in seven lost all Medicaid for a month or more; fewer than a third returned within six months. The grace period runs 30 days to six months, then the member — and their alignment, and their survival curve — terminates. A duals retention strategy without redetermination support is a strategy to watch members evaporate through a hole in another agency's floor.
6. There is no 2030 mandate. The bill that would have required integration died in committee. The regulatory push is real — monthly integrated-care enrollment, plan caps, look-alike squeezes — and worth planning around. A statutory deadline on a slide is a fabricated forcing function, and every plan priced against it inherits the fabrication.
We built this as an instrument: the alignment ledger against the national record, the Medicaid break funnel, the enrollment mechanics after the 2025 rework, and a veto on integration savings. Browser-only; nothing entered leaves the machine.
In duals, the growth is real, the alignment is countable, and the savings claim is the one thing the record has already tested.
What share of your D-SNP book is actually aligned — and if the answer takes more than an hour to produce, what does that say about the deck?