Insurance coverage
When 25 million people lost Medicaid, was it because they were ineligible, or because of paperwork?
The uninsured rate fell to a record 7.6% in 2023, but coverage is deeply uneven. 16.8% of Texans are uninsured versus 2.8% of people in Massachusetts, and Hispanic working-age adults are uninsured at 24.8%, more than three times the White rate of 6.8%.
The problem
Coverage policy shapes the entire care delivery landscape, but enrollment alone does not guarantee access to hospitals, physicians, medications, or affordable follow-up. The national failure mode is churn plus card-to-care leakage: eligible people lose coverage through administrative process, and insured people can still face cost barriers or provider-acceptance gaps.
The recommendation
Run coverage as an end-to-end access program. The recommended strategy is to automate renewals, reduce procedural disenrollment, monitor whether Medicaid and marketplace coverage convert into accepted appointments, and hold states and plans accountable for continuity of care after enrollment.
The gap: who and where
How wide the uninsured gap is, who it lands on, where it concentrates, and whether it has begun to reverse for children.
Uninsured rate, 2019 to 2023
NHIS estimates for all ages and for working-age adults.
Read it this way Both the all-ages line (10.3% to 7.6%) and the working-age-adult line (14.7% to 10.9%) fall every year with no reversal, confirming the record-low uninsured rate is a sustained multi-year trend. The gap between the two lines, adults consistently higher than the all-ages figure, reflects that children have separate coverage pathways this chart doesn't break out. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Year | All ages (%) | Adults 18 to 64 (%) |
|---|---|---|
| 2019 | 10.3 | 14.7 |
| 2020 | 9.7 | 13.9 |
| 2021 | 9.2 | 13.5 |
| 2022 | 8.4 | 12.2 |
| 2023 | 7.6 | 10.9 |
NCHS, NHIS Health Insurance Coverage Early Release, 2023 · 2023 · source
Child uninsured rate reversed after the unwinding, 2022 vs 2024
Child uninsured rate rose after the Medicaid unwinding, reversing a decade of gains (Census ACS via Georgetown CCF).
Read it this way Both age groups moved the wrong direction from 2022 to 2024, with children under 6 rising slightly more (4.3 to 5.3%) than children under 19 overall (5.1 to 6.0%). Two data points can't confirm a new multi-year trend by themselves, only that the direction reversed between these two years. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Age group | 2022 uninsured (%) | 2024 uninsured (%) |
|---|---|---|
| Children under 19 | 5.1 | 6 |
| Children under 6 | 4.3 | 5.3 |
Georgetown University Center for Children and Families (analysis of Census ACS) · 2024 · source
Uninsured rate by state (2024)
ACS-based estimate. US overall is 8.2%. Texas is highest at 16.8%, Massachusetts lowest at 2.8%.
Read it this way States shaded darkest, led by Texas at 16.8%, run more than double the 8.2% national benchmark, while Massachusetts at 2.8% is roughly a sixth of Texas's rate. This is a single-year snapshot, so it can't show whether a state's uninsured rate is improving or worsening over time. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| State | Uninsured (%) |
|---|---|
| Texas | 16.8 |
| Georgia | 12 |
| Oklahoma | 11.7 |
| Nevada | 11.4 |
| Alaska | 11.1 |
| Florida | 10.9 |
| Arizona | 10.2 |
| New Mexico | 10.2 |
| Mississippi | 9.7 |
| Tennessee | 9.7 |
| Wyoming | 9.7 |
| Arkansas | 9.3 |
| South Carolina | 9.1 |
| Idaho | 9 |
| Kansas | 8.6 |
| Montana | 8.6 |
| North Carolina | 8.6 |
| Utah | 8.3 |
| Alabama | 8.2 |
| South Dakota | 7.9 |
| Colorado | 7.8 |
| New Jersey | 7.7 |
| Louisiana | 7.6 |
| Missouri | 7.6 |
| Nebraska | 7.6 |
| Indiana | 7.4 |
| Delaware | 6.9 |
| Kentucky | 6.9 |
| Illinois | 6.8 |
| Virginia | 6.8 |
| Ohio | 6.7 |
| Washington | 6.5 |
| Maryland | 6.2 |
| North Dakota | 6 |
| Pennsylvania | 6 |
| California | 5.9 |
| West Virginia | 5.9 |
| Connecticut | 5.7 |
| Iowa | 5.7 |
| Maine | 5.6 |
| Wisconsin | 5.4 |
| Minnesota | 5.3 |
| Oregon | 5.2 |
| Michigan | 5 |
| New York | 4.9 |
| Rhode Island | 4.8 |
| District of Columbia | 4.7 |
| New Hampshire | 4.7 |
| Vermont | 4.3 |
| Hawaii | 3.3 |
| Massachusetts | 2.8 |
KFF State Health Facts, Health Insurance Coverage of the Total Population · 2024 · source
Uninsured rate by race and ethnicity, adults 18 to 64 (2023)
Share of working-age adults with no health coverage at time of interview.
Read it this way The Hispanic uninsured rate (24.8%) is more than three times the White rate (6.8%) and more than five times the Asian rate (4.4%), so coverage gaps by race and ethnicity are large even after the overall uninsured rate hit a record low. This is a national aggregate, so it can't show how much of the gap traces to income or state of residence versus race itself. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Group | Uninsured (%) |
|---|---|
| Hispanic | 24.8 |
| Black, non-Hispanic | 10.4 |
| White, non-Hispanic | 6.8 |
| Asian, non-Hispanic | 4.4 |
NCHS, NHIS Health Insurance Coverage Early Release, 2023 · 2023 · source
Uninsured rate by income, adults 18 to 64 (2023)
By household income as a percent of the federal poverty level (FPL).
Read it this way Uninsured rates fall sharply as income rises, from 20.2% below the poverty line to 4.3% at 400% of it or more, nearly a fivefold difference between the bottom and top bands. Income bands here are self-contained categories, so this chart can't show how income and the race and ethnicity gap shown alongside it interact for any individual. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Income band | Uninsured (%) |
|---|---|
| Below 100% FPL | 20.2 |
| 100 to 199% FPL | 19.1 |
| 200 to 399% FPL | 11.5 |
| 400% FPL or more | 4.3 |
NCHS, NHIS Health Insurance Coverage Early Release, 2023 · 2023 · source
Losing coverage: the unwinding
Why 25 million lost Medicaid, paperwork versus true ineligibility, and how children bore a large share of the loss.
How 25.2 million Medicaid disenrollments split: paperwork vs ineligibility
Procedural means a paperwork or administrative reason, not a finding of ineligibility.
Read it this way More than two in three of the 25.2 million people disenrolled during the unwinding lost coverage for a paperwork reason, not because anyone determined they were ineligible. That split is the direct evidence behind the recommendation to expand automatic renewals, though it doesn't show how many of the 69% procedurally disenrolled were actually still eligible. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Reason | Share (%) |
|---|---|
| Procedural (paperwork) | 69 |
| Ineligibility | 31 |
KFF, As Medicaid Unwinding Concludes, 25 Million Lost Coverage · 2024 · source
Medicaid disenrollment rate, highest states vs lowest
Share of completed Medicaid renewals that ended in disenrollment. Only states with a source-confirmed rate are shown.
Read it this way Montana's disenrollment rate (57%) is nearly five times North Carolina's (12%), showing the unwinding's impact varied enormously by state administration, not just by national eligibility rules. As the caveat notes, only these six states had a source-confirmed rate at collection, so this can't be read as a ranking of all 50 states. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
Caveat KFF's full 50-state tracker is a live table. Only these six states had a discrete, source-confirmed rate at collection. Others are omitted rather than estimated.
⊞ data table⬇ CSV
| State | Disenrollment rate (%) |
|---|---|
| Montana | 57 |
| Utah | 56 |
| Idaho | 55 |
| Oklahoma | 54 |
| Texas | 54 |
| North Carolina | 12 |
KFF, Medicaid Enrollment and Unwinding Tracker · 2024 · source
Insurance isn't access
Even with a card, cost still blocks care, and a Medicaid card buys second-class access because fewer doctors accept it.
Insurance is not access: barriers by insurance status, 2023
NHIS. The insured comparator is all insured for the first measure and privately insured for the other two.
Read it this way The uninsured bar is several times taller than the insured bar on every measure, most sharply for having no usual source of care (40.6% vs 9.1%), showing lacking any coverage is a much bigger barrier than the type of coverage a person has. The insured comparator switches definitions across measures (all insured versus privately insured only), so read the exact multiples as directional, not precise. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Barrier | Uninsured (%) | Insured (%) |
|---|---|---|
| Delayed or went without care (cost) | 26.3 | 5.8 |
| No usual source of care | 40.6 | 9.1 |
| Skipped a prescription (cost) | 13 | 4.9 |
NCHS NHIS, via Healthy People 2030 (AHS-04, AHS-06, AHS-07) · 2023 · source
Cost barrier to needed care, 2019 to 2023
Share who delayed or went without needed care because of cost, all ages (NHIS, Healthy People 2030 AHS-04).
Read it this way The cost barrier eased from 8.5% to 6.8% between 2019 and 2022 but has been flat since, showing the improvement stalled rather than continuing. This tracks all ages combined and doesn't break out whether the plateau looks different for insured versus uninsured people. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| Year | Cost barrier (%) |
|---|---|
| 2019 | 8.5 |
| 2020 | 7.2 |
| 2021 | 7 |
| 2022 | 6.8 |
| 2023 | 6.8 |
NCHS NHIS, via Healthy People 2030 (AHS-04) · 2023 · source
Physicians accepting new patients, by payer
Share of office-based physicians accepting new patients, by payer. Medicaid is 2014 to 2017, Medicare and private are 2019.
Read it this way Medicaid acceptance (73.7%) trails Medicare (89%) and private insurance (91%) by 15 to 17 points, so a Medicaid card still buys measurably less access to a new physician than other coverage types. As the caveat notes, the Medicaid figure is from 2014 to 2017 while Medicare and private are 2019, so the true gap today could be narrower or wider. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
Caveat Payer figures come from different survey years (Medicaid 2014 to 2017, Medicare and private 2019).
⊞ data table⬇ CSV
| Payer | Accepting new patients (%) | Data years |
|---|---|---|
| Private / commercial | 91 | 2019 |
| Medicare | 89 | 2019 |
| Medicaid | 73.7 | 2014 to 2017 |
MACPAC and KFF, Physician Acceptance of New Patients by Payer · 2019 · source
Medicaid-to-Medicare fee index by state (all services, 2024)
Medicaid physician fee as a ratio of the Medicare fee for the same services. Below 1.0 means Medicaid pays less. Tennessee runs Medicaid fully through managed care, so it has no fee-for-service index.
Read it this way States shaded darkest, like Rhode Island at 0.52 and New Jersey at 0.61, pay Medicaid physicians roughly half the Medicare rate, while Montana (1.32) and Alaska (1.3) pay Medicaid physicians more than Medicare. Tennessee has no value because it runs Medicaid fully through managed care, so its absence from the map isn't a data gap. Use this chart to see where coverage breaks down, enrollment, affordability, or provider access, and why the recommendation treats card-to-care conversion as the outcome rather than insured status alone.
⊞ data table⬇ CSV
| State | Medicaid-to-Medicare fee index |
|---|---|
| Montana | 1.32 |
| Alaska | 1.3 |
| New Mexico | 1.21 |
| North Dakota | 1.06 |
| Nebraska | 1.01 |
| Wyoming | 1 |
| Arizona | 0.98 |
| Delaware | 0.96 |
| Indiana | 0.96 |
| Maryland | 0.95 |
| Idaho | 0.94 |
| Oklahoma | 0.94 |
| Mississippi | 0.93 |
| Alabama | 0.92 |
| South Dakota | 0.91 |
| Nevada | 0.9 |
| South Carolina | 0.89 |
| Oregon | 0.88 |
| Vermont | 0.87 |
| Missouri | 0.86 |
| Hawaii | 0.84 |
| Colorado | 0.83 |
| Georgia | 0.83 |
| Virginia | 0.83 |
| North Carolina | 0.82 |
| West Virginia | 0.82 |
| District of Columbia | 0.81 |
| Utah | 0.8 |
| Connecticut | 0.79 |
| Maine | 0.78 |
| Iowa | 0.77 |
| Arkansas | 0.76 |
| Michigan | 0.76 |
| New York | 0.76 |
| Massachusetts | 0.74 |
| Minnesota | 0.74 |
| New Hampshire | 0.73 |
| Kansas | 0.69 |
| Kentucky | 0.69 |
| Pennsylvania | 0.68 |
| California | 0.67 |
| Wisconsin | 0.66 |
| Florida | 0.64 |
| Louisiana | 0.64 |
| Washington | 0.64 |
| Illinois | 0.63 |
| Ohio | 0.63 |
| Texas | 0.63 |
| New Jersey | 0.61 |
| Rhode Island | 0.52 |
| Tennessee | no fee-for-service index |
KFF State Health Facts, Medicaid-to-Medicare Fee Index (Urban Institute) · 2024 · source
Geography
The same question, state by state and then county by county. Pick a state in the filter above to drill into its counties.
Uninsured rate (all ages)
County · direct countEach tile is a state. Pick a state in the Scope control above to drill into its counties.
Census SAHIE / ACS via HealthPulse · 2022 · source
Adults without health insurance
County · modeled prevalence (95% CI)Each tile is a state. Pick a state in the Scope control above to drill into its counties.
CDC PLACES (model-based small-area estimates) · 2024 · source
Why this matters
Of the 25.2 million people disenrolled from Medicaid during the unwinding, 69% lost coverage for a paperwork or administrative reason, not a finding of ineligibility, and children made up 37% of disenrollments in the 14 states reporting age-specific data. State administration varied enormously, from a 57% disenrollment rate in Montana to 12% in North Carolina among the six states with a source-confirmed rate, showing the paperwork burden itself, not eligibility rules, is what pushes people off coverage. Leadership should care because this is coverage the country already earned and is giving back to a process failure, not a policy reversal.
Recommended actions
- Target automatic (ex parte) Medicaid renewal expansion in states still relying on manual paperwork renewals.
- Prioritize remediation of the child Medicaid enrollment reversal, a 5 million net drop through October 2024, given children's outsized share of procedural disenrollments.
- Monitor the six states with source-confirmed disenrollment rates (Montana, Utah, Idaho, Oklahoma, Texas, North Carolina) as an early-warning cohort for administrative-burden fixes.
- Shift outreach resources toward Hispanic and lower-income populations, where uninsured rates run three to five times the White and high-income baselines.
- Track the cost-barrier-to-care rate, flat at 6.8% since 2022, as the KPI deciding whether coverage gains are translating into actual care access.
The recommendation
Therefore, run coverage as an end-to-end access program. The recommended strategy is to automate renewals, reduce procedural disenrollment, monitor whether Medicaid and marketplace coverage convert into accepted appointments, and hold states and plans accountable for continuity of care after enrollment.
Demographic slice race, income, age, citizenship, native to Census ACS and NHIS. The anchor metric for the Equity panel.
Sources
- NCHS, Health Insurance Coverage: Early Release from the National Health Interview Survey, 2023 · 2023
- KFF State Health Facts, Health Insurance Coverage of the Total Population · 2024
- KFF, As Medicaid Unwinding Concludes, 25 Million Lost Coverage · 2024
- MACPAC, Physician Acceptance of New Medicaid Patients (NEHRS) and Medicaid-to-Medicare Fee Index · 2024
- Georgetown University Center for Children and Families, child uninsurance analysis of Census ACS · 2025