Gizmo Warehouseby Joe Eichenbaum | 17A

Medicaid work requirements · Oklahoma

Medicaid work requirements in Oklahoma: who is subject and who loses coverage (2027)

Data snapshot 2026-06-03 · part of These 5 Million People Are About to Lose Their Medicaid — But They Don't Have To

Oklahoma expanded Medicaid, so the federal work requirement that takes effect January 1, 2027 applies to its expansion adults. About 224,019 of them (9.5% of an expansion pool of 239,479) will have to prove 80 hours a month of work, school, or service, or an exemption, to keep coverage.

This model's bottom-up estimate is that 126,302 people in Oklahoma lose coverage by 2034: 119,581 of them still eligible but lost to paperwork and reporting failures, and 6,721 not meeting the requirement. Nationally the same model gives 5,461,400, against CBO's 5,200,000.

Whether those eligible people keep coverage depends on the state's ability to renew them automatically (ex parte) from data it already holds. Oklahoma scores 50 of 100 on this model's ex parte capability index, a mid band; its eligibility system is Soonercare IMS.

Key figures

Model v2-2026-06-03. "Strict" counts adults whose ACS record shows them subject with no observable exemption; "permissive" allows the broader exemption reading.
MeasureOklahoma
Expansion stateYes
Subject to the requirement via waiverNo
Expansion pool (adults 19 to 64)239,479
Adults subject, strict definition224,019
Adults subject, permissive definition219,538
Share of expansion pool subject9.5%
Projected coverage loss by 2034126,302
Hardship exception statusNot adopting
Early implementerNo
Ex parte capability score (0 to 100)50 (mid)

Who is subject in Oklahoma

Of the 224,018 subject adults, 38% already work 80 or more hours a month and 57% are not working. 40% have a child under 14 at home. Among those not working, 55% report a disability and 17% are caregivers.

ACS PUMS 5-year 2020 to 2024, subject adults only.
CharacteristicShare of subject adults
Age 19-2415%
Age 25-3424%
Age 35-4422%
Age 45-5417%
Age 55-6422%
Not working57%
1 to 19 hours a month1%
20 to 79 hours a month4%
80 or more hours a month (meets the bar)38%
Not working: Disabled55%
Not working: Caregiver17%
Not working: Postpartum0%
Not working: Student2%
Not working: Other26%

Why eligible people lose coverage

Of the 126,302 projected to lose coverage in Oklahoma, 119,581 meet the requirement or qualify for an exemption but fail to document it. 38,852 of those are working people who cannot prove their hours through data the state can see, and 80,729 qualify for an exemption the state does not detect automatically.

Bottom-up model v8.0-2026-05-30; documentation-failure buckets compound over the 14 six-month renewal cycles through 2034.
GroupProjected to lose coverage
Variable shifts / on-call work10,333
Self-employed (other)7,503
Full-time students not auto-verified7,323
Multiple part-time jobs (aggregation failure)5,887
Volunteering / job training2,479
Cash-paid construction & trades2,396
Seasonal: agriculture, hospitality, food service1,873
Gig / courier (DoorDash, Uber, Lyft, Instacart)1,059
Medically frail without auto-match23,470
American Indian / Alaska Native (tribal exemption)20,990
Parent caregivers of children ≤13 without auto-match12,575
Caregivers of a disabled adult11,203
SUD treatment not flagged in claims data4,114
Kinship caregivers (non-parent) of children ≤134,004
Recent incarceration (data-match gap)2,242
Other categorical exemptions (foster, AYA cancer, SNAP/TANF)1,928
Pregnancy / postpartum data lag203

Can the state renew people automatically?

The ex parte capability index blends the state's observed automatic-renewal rate (51% in CMS eligibility-processing data), the wage, disability, and other data sources it can query, and its history of churn. Oklahoma's composite score is 50 (mid band). A low score means more eligible people will be dropped for paperwork unless the state changes its process before January 1, 2027.

ComponentScore
Composite50
Observed ex parte51
Core capability18
Data sources50
Historical churn50

How subjects overlap across exemption categories

Each subject adult sits in exactly one cell. Categories: Working ≥ qualifying hours / income; Parent caregiver of child ≤13; Medically frail; Full-time student.
CombinationAdultsShare
None of the four74,73133.4%
Working ≥ qualifying hours / income + Parent caregiver of child ≤1336,79716.4%
Medically frail36,31816.2%
Working ≥ qualifying hours / income35,25615.7%
Parent caregiver of child ≤1318,2858.2%
Full-time student4,9932.2%
Working ≥ qualifying hours / income + Full-time student4,3711.9%
Working ≥ qualifying hours / income + Medically frail3,5671.6%

Counties in Oklahoma

Subject adults concentrate: Oklahoma County (42,600), Tulsa County (36,586), and Cleveland County (18,034) account for 43% of the state's subject population.

All 77 counties in Oklahoma with subject adults, sorted by count. County figures are model allocations of the state estimate by ACS tract characteristics.
CountyAdults subjectSubject rateProjected lossPoverty rateWorking-age population
Oklahoma County42,6009.0%24,01816.0%474,998
Tulsa County36,5869.2%20,62714.0%398,454
Cleveland County18,0349.6%10,16813.0%187,945
Canadian County8,4928.4%4,7888.0%100,745
Payne County6,27811.9%3,53924.0%52,806
Rogers County5,6259.8%3,17210.0%57,558
Comanche County5,6037.6%3,15917.0%73,864
Wagoner County4,8439.6%2,7319.0%50,352
Creek County4,1089.9%2,31614.0%41,603
Pottawatomie County4,1069.5%2,31516.0%43,118
Muskogee County3,6519.6%2,05918.0%37,967
Garfield County3,3469.6%1,88614.0%34,742
Logan County3,25010.6%1,83214.0%30,609
Grady County3,2389.8%1,82511.0%33,112
Washington County3,06710.5%1,72914.0%29,057
Cherokee County3,03410.7%1,71021.0%28,310
Le Flore County2,7209.8%1,53420.0%27,694
Osage County2,69710.4%1,52113.0%26,026
Carter County2,6539.6%1,49615.0%27,604
Bryan County2,6299.4%1,48216.0%27,842
Kay County2,58210.9%1,45616.0%23,736
Delaware County2,50711.3%1,41419.0%22,119
Stephens County2,49110.3%1,40417.0%24,140
McClain County2,4829.4%1,3997.0%26,336
Pittsburg County2,3819.8%1,34319.0%24,411
Mayes County2,30610.2%1,30017.0%22,506
Sequoyah County2,24810.0%1,26820.0%22,438
Pontotoc County2,1199.8%1,19513.0%21,737
Okmulgee County2,07910.0%1,17218.0%20,857
Lincoln County2,02210.5%1,14014.0%19,186
Custer County1,78610.8%1,00716.0%16,592
Ottawa County1,77910.7%1,00320.0%16,677
McCurtain County1,70910.1%96321.0%16,852
Caddo County1,4589.2%82219.0%15,885
Garvin County1,4299.9%80615.0%14,446
Seminole County1,33310.2%75223.0%13,118
Jackson County1,3119.2%73915.0%14,333
Texas County1,23710.3%69720.0%11,960
McIntosh County1,21511.7%68522.0%10,369
Woodward County1,1109.5%62615.0%11,726
Beckham County1,0878.1%61322.0%13,365
Adair County1,0869.8%61222.0%11,137
Marshall County97111.6%54819.0%8,389
Pawnee County94710.8%53415.0%8,781
Kingfisher County91010.7%51312.0%8,545
Murray County83110.9%46815.0%7,624
Craig County8079.9%45519.0%8,172
Choctaw County79610.4%44923.0%7,646
Atoka County7789.2%43918.0%8,453
Hughes County7049.0%39721.0%7,856
Haskell County68810.6%38821.0%6,466
Noble County64210.8%36211.0%5,941
Johnston County63711.1%35923.0%5,758
Pushmataha County63210.7%35622.0%5,921
Love County59610.3%33614.0%5,770
Washita County59110.0%33313.0%5,928
Woods County58611.4%33019.0%5,148
Latimer County57311.1%32317.0%5,180
Okfuskee County5678.6%32027.0%6,566
Nowata County5169.8%29116.0%5,282
Blaine County51511.0%29014.0%4,701
Kiowa County46210.1%26023.0%4,577
Tillman County41110.9%23217.0%3,762
Major County3899.7%21912.0%4,012
Coal County32811.6%18523.0%2,835
Jefferson County32311.4%18223.0%2,843
Cotton County31010.1%17522.0%3,058
Beaver County28810.7%1627.0%2,686
Dewey County26211.6%14814.0%2,256
Alfalfa County2577.4%14512.0%3,452
Grant County25211.6%14213.0%2,172
Greer County2417.1%13617.0%3,399
Ellis County22411.5%12615.0%1,947
Roger Mills County20011.3%11314.0%1,770
Harper County1659.4%9314.0%1,756
Cimarron County15613.9%8811.0%1,124
Harmon County14811.6%8418.0%1,279

State brief and methodology

Estimates, not counts. The model rakes to CBO's national subject total and allocates by ACS microdata; state agencies hold the administrative data that would replace these figures.

Sources

Back to These 5 Million People Are About to Lose Their Medicaid — But They Don't Have To