Who is actually impacted

The dashed rows are deliberately empty — we will not draw a bar for a number we do not have. SNAP is already operating under the new rules, so unlike Medicaid — which starts 1 January 2027 — this is not a forecast; the effect is already in the caseload and nobody has published the cut. We have not established the date the State began applying them.36

Both figures are soft, and the estimate is only as good as the assumption that one plan’s book resembles the state’s. It is published here because the alternative — planning with no number at all — is worse, and because naming it invites correction.4

The age bands above do not match, and the mismatch is in the statutes rather than in our arithmetic. It catches almost everyone the first time.

Medicaid runs 19 through 64.An applicable individual is someone who “has attained the age of 19 and is under 65 years of age,” so eighteen-year-olds and sixty-five-year-olds both fall outside it.2 Under-19s are not merely outside — the statute puts them in mandatory deeming, so the State must treat them as compliant without asking. At the top end, 65 brings Medicare. Hawaiʻi Med-QUEST Division (Med-QUEST) uses the same band in its own member guidance.4

SNAP runs 18 through 65.Different statute, different line: the exception is “under 18, or over 65.”1 So SNAP catches both the eighteen-year-old and the sixty-five-year-old that Medicaid does not.

This is the same asymmetry that makes Hawaiʻi DHS materials — which describe the newly subject band as 55 to 64 — understate it by a year at the top. Both lists in full.

The 124,000 above is labeled before exclusions for a reason. Nine categories of people are excluded from the Medicaid requirement on status alone, decided before anyone asks what they did that month.2

• ·Caregivers of a child 13 or under
• ·Caregivers of a disabled individual
• ·Medically frail (five sub-categories)
• ·Totally disabled veterans
• ·Indian, Urban Indian, IHS-eligible
• ·Pregnant and postpartum
• ·In SUD treatment
• ·Inmates of a public institution
• ·TANF-compliant adults
• ·SNAP-subject household members

We cannot size the exclusions from any public source, which is why no bar is drawn here. The one empirical check available points the same direction as the ~51 percent above: a QUEST plan working its own book found roughly half of its subject members fell away once it filtered on data it already held.

The largest single exclusion is likely the SNAP bridge. Anyone in a SNAP household who is subject to SNAP’s work requirement is outside Medicaid’s entirely — and Hawaiʻi has 161,300 people on SNAP. The full lists are here.

Set the two programs side by side and count each person once. Roughly 72,479 people in Hawaiʻi face a work requirement under one or the other, and after exclusions, earnings and existing activity, about 21,519 of them need hours built for them.

At the 80-hour threshold that is 1,721,520 hours a month — and that is the floor.

The ceiling is about 2.06 million.The survey base above undercounts Medicaid enrolment against Med-QUEST’s own August 2026 figures.5 Correcting the Medicaid side alone — the one place we have a like-for-like administrative comparison — raises the group needing hours to roughly 25,800 and the demand to 2,062,400 hours a month. We do not scale the SNAP side, because our survey figure there runs aboveUSDA’s administrative caseload rather than below it, and that gap looks like a definition difference — we counted everyone in a SNAP household, USDA counts people receiving benefits — not a real overcount. Correcting one side and not the other is the conservative choice, so the band is 1.7 to 2.1 million.

Both ends exclude top-up hours. A further 3,516 people are already working or in school but short of the threshold. They need the gap closed, not the whole 80, and no figure on this page counts them. Even the ceiling is conservative.

That figure is the case for role redesign rather than recruitment. No organization absorbs it, and it will not be met by more one-off beach cleanups; it needs many organizations each designing a handful of sustained, recurring positions where the same person returns.

Same microdata as the bar above, tabulated across both programs and re-derived from the raw file. Every component reproduces exactly: SNAP 38,096 subject (n=416) and Medicaid 34,383 (n=372). The 21,519 carries a 90% range of 18,502 to 24,536, so the sampling range on the floor alone runs roughly 1.48 to 1.96 million hours a month. Both hours figures are arithmetic on a headcount, not separate estimates. Note the two uncertainties are different in kind and should not be added: the 18,502–24,536 range is sampling error within the survey, while the 1.7-to-2.1 band is the distance between the survey and Med-QUEST’s administrative count.8

The Medicaid community engagement requirement begins 1 January 2027.2

ʻOhana Health Plan does not intend to renew its QUEST Integration contract beginning 1 January 2027.7 Med-QUEST has already removed it from auto-assignment, taking the plans available statewide from five to four in Honolulu and Maui counties, and four to three in Hawaiʻi and Kauaʻi counties.

So a large group of people change health plan on the same day a new requirement starts. What decides who gets reached is whether their plan holds current contact details — and a member who has just been moved is the member least likely to have them on file.

Source: Med-QUEST Division memo QI-2608, 4 March 2026, Auto Assignment Algorithm for QI Members, published to Med-QUEST’s public provider-memo library and linked in the references below. Med-QUEST’s public health-plans page still lists all five.