The 2017 Senate health bill would have capped federal Medicaid funds from eight quarters of each state’s data
On June 26, 2017, CBO described the Senate’s Medicaid cap: federal funding per enrollee, grown by price indexes from eight quarters each state chose. In our reading, that made state eligibility and claims data a budget input, as the 2025 law later did for eligibility.
Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.
Two-minute summaryThe dek and the first sentence of each section, quoted from the article
On June 26, 2017, CBO described the Senate’s Medicaid cap: federal funding per enrollee, grown by price indexes from eight quarters each state chose. In our reading, that made state eligibility and claims data a budget input, as the 2025 law later did for eligibility.
- The week The headline figures came first.
- The base: eight quarters chosen by the state Every cap needs a starting point.
- The growth: inflation indexes in place of spending trends The second input is the growth rate.
- The count: enrollees as the multiplier The third input is the count.
- The response: who absorbs costs above the cap The fourth part is the consequence.
- How the cap would have reached different public bodies and providers The cap was a federal formula with state consequences.
- Objections, in order The bill failed in the Senate in July 2017. [3] [4] [5] The design matters because its core idea, federal funding tied to state data about each enrollee, did not go away.
- What a state Medicaid or budget office could check The four parts translate into checks a state can run on its data, whatever federal financing rules apply.
- With hindsight The Senate did not pass the bill.
The week
On June 26, 2017, the Congressional Budget Office published its cost estimate for the Better Care Reconciliation Act of 2017. [1] The bill was the Senate’s amendment in the nature of a substitute to H.R. 1628, the health bill the House had passed on May 4, 2017. [1] [2] CBO estimated the version posted that day on the website of the Senate Committee on the Budget. [1]
The headline figures came first. CBO and the staff of the Joint Committee on Taxation estimated the bill would reduce the cumulative federal deficit over 2017 to 2026 by $321 billion. [1] They estimated it would increase the number of uninsured people by 22 million in 2026 relative to current law. [1] By 2026, an estimated 49 million people would be uninsured, compared with 28 million under current law. [1]
CBO said the largest savings would come from Medicaid. It estimated that spending on the program would decline in 2026 by 26 percent compared with its current-law projection. [1] It projected 15 million fewer Medicaid enrollees by 2026 than in its March 2016 baseline. [1]
Two Medicaid changes drove those figures. Starting in 2021, the bill would reduce the federal matching rate for adults made eligible by the Affordable Care Act, by 5 percentage points a year through 2023. [1] After that, the rate would equal the state’s rate for other enrollees. [1] And beginning in fiscal year 2020, the bill would cap federal Medicaid payments per enrollee. [1]
CBO set out how the cap would work. A state’s limit would start from its average per-enrollee cost of medical services for most full-benefit enrollees. The average would be taken over eight consecutive quarters of the state’s choosing. Those quarters fell between the first quarter of federal fiscal year 2014 and the third quarter of 2017. [1]
The enrollees fell into five categories: the elderly, disabled adults, nondisabled children, adults made eligible by the Affordable Care Act, and all other adults. [1] Disabled children would be excluded from the caps. [1] The Secretary of Health and Human Services would inflate each state’s averages forward. Children and nondisabled adults would grow with the medical care component of the consumer price index, CPI-M. Most disabled and elderly enrollees would grow at CPI-M plus 1 percentage point. [1] From 2025, all groups would grow with the consumer price index for all urban consumers, CPI-U. [1]
The limit for each year would be the inflated average cost for each group, multiplied by the number of enrollees in that group. [1] If a state spent more, the federal government would not pay its share of the excess. [1] CBO projected that per-enrollee spending for children and nondisabled adults would grow at 4.9 percent a year under current law, faster than CPI-M at 3.7 percent. [1] It projected CPI-U growth of 2.4 percent in 2025 and beyond. [1]
The bill also offered states a block grant for nondisabled adults, other than the expansion adults, from 2020. [1] CBO expected that option to appeal mainly to a few states expecting population decline, and to have little effect on enrollment. [1] CBO said states would not have substantial additional flexibility under the per capita caps. [1]
The House-passed bill had used a different base. It started from each state’s fiscal year 2016 spending and enrollment, reported on the quarterly CMS-64 expense report. [2] It required states to report spending and enrollee counts by category from October 2018, and required federal audits of those figures. [2] The Senate version, as CBO described it, let each state choose its base quarters. [1]
A per capita cap is a formula. Its output, a ceiling on federal Medicaid payments, comes from three inputs: a base cost per enrollee, a growth rate and a count of enrollees. [1] The Senate’s 2017 version made a fourth thing matter too: how a state would respond to the gap between its costs and the ceiling. [1] Each input is a number someone has to produce.
Our argument is that the cap would have turned state Medicaid data into federal budget authority. The base came from eight quarters of a state’s own claims and enrollment records. The count came from its own enrollment systems. [1] [2] Errors, gaps or classification choices in those systems would have flowed straight into federal dollars, and stayed there. The policy did not pass. The data lesson returned in another form in the 2025 reconciliation law. [6]
We take the cap in four parts:
- The base: which period set each state’s starting cost per enrollee.
- The growth: how that cost would be inflated each year.
- The count: how enrollees would be counted and classified.
- The response: what states, providers and plans could do when costs ran above the cap.
The first three parts are the formula’s inputs. The fourth is the consequence of its output. Together they cover the cap from its first calculation to its effect on care. After the four parts, we look at how the cap would have reached different public bodies and providers, answer objections in order, and list what a state Medicaid or budget office could check.
Why would a state chief information officer or a health system’s finance leader care about a bill that failed? Because the formula would have run on their systems. The cap’s inputs were claims, capitation payments, eligibility categories and monthly counts. [1] [2] Each is produced by a system with an owner, a vendor and a control environment. A financing debate in Washington would have become a data quality program in every state capital.
A note on sources and our reading. We describe the Senate version as CBO described it on June 26, 2017, because CBO’s estimate is an official, published description of that text. [1] Where we describe the House version, we quote its engrossed text. [2] Statements about how the cap would have affected a particular state or organization are our reading. Agencies should confirm their obligations with counsel.
The base: eight quarters chosen by the state
Every cap needs a starting point. CBO described the Senate’s: a state’s average per-enrollee cost of medical services for most full-benefit enrollees, over eight consecutive quarters of the state’s choosing. The window ran from the first quarter of federal fiscal year 2014 to the third quarter of 2017. [1]
A choice with consequences
Letting a state choose its quarters sounds like a concession, and in one sense it was. A state could pick the period that best reflected its costs. In our reading, the choice also made the base a matter of state analysis. A state would need clean, consistent data across every quarter in the window to know which eight to pick.
In our reading, the quarters in that window need not have been uniform. Enrollment, prices and the mix of enrollees within a category can shift over a stretch of years. A state’s average cost per enrollee in a category could differ from one run of quarters to another. Picking well required modeling each possible window.
The House’s fixed year
The House text took a different approach. It set fiscal year 2016 as the base year. [2] The Secretary would calculate each state’s adjusted total medical assistance spending and its number of enrollees for that year, and notify the state by April 1, 2018. [2] The average was then inflated by CPI-M from September 2016 to September 2019. [2]
The House text also built in an audit. The Secretary would audit the number of individuals and expenditures each state reported through the CMS-64 for fiscal year 2016, fiscal year 2019 and each later year. [2] Audits could use a representative sample. [2] In our reading, the House treated the base as a number to be verified. The Senate version, as CBO described it, treated the base as a number to be chosen.
A penalty for missing data
The House text went further. If a state failed to satisfactorily submit the required spending and enrollee data for any quarter, from fiscal year 2019, two things followed. [2] The Secretary would treat all the affected enrollee categories as a single category. And the growth factor applied to the state would be decreased by 1 percentage point. [2]
In our reading, that provision priced data quality directly. A state that could not report by category would lose a point of annual growth on its federal limit, for every year the problem continued. The provision tied a reporting failure to a permanent funding loss.
What counted as cost
The base included costs of medical services for most enrollees with full benefits. [1] The House text spelled out exclusions. It removed disproportionate share hospital payments and certain other categories from the measured spending. [2] It defined non-DSH supplemental payments and treated them separately in the fiscal year 2019 calculation by category. [2]
Supplemental payments matter to hospitals. In our reading, the treatment of each payment type in the base would decide how much room the cap left for them later. A state that had reported supplemental payments inconsistently across quarters would carry that inconsistency into its permanent base.
Late claims and adjustments
A quarter’s spending is not final when the quarter ends. Claims arrive late, payments are adjusted and recoveries come in. The House text defined spending as reported on the CMS-64, including enrollment data and subsequent adjustments to the report. [2] In our reading, a base quarter would keep changing for some time after it closed.
That raised a timing question. A state choosing its quarters in 2017 would be choosing among periods with different degrees of completeness. The later quarters in the window would be less settled than the earlier ones. A state would want to know how much its recent figures were likely to move before relying on them.
Managed care in the base
Where a state pays managed care plans a monthly capitation rate, its federal spending report shows the capitation, not the care. The cost per enrollee in the base would then reflect the rates the state had set. [1] [2] In our reading, a state that had set rates low in the base period would carry low rates into its limit. One that had set them high would carry room it might not need.
Capitation also moves the detailed data outside the state. Plans hold the encounter records that show what care was delivered. A state defending its base or its categories would need those records to be complete and consistent with its payments.
Base data as a frozen asset
Once set, the base would not be recalculated. Each year’s limit would grow from the year before. [1] [2] A state whose base understated its true costs, because of late claims or misclassified spending, would carry the shortfall forward indefinitely. A state whose base was clean would carry an accurate starting point.
That makes the base a one-time data event with lasting effect. For a state Medicaid agency, preparing the base would be a reporting exercise with permanent consequences. It would call for the agency’s finance, claims and eligibility teams, its actuaries and its fiscal agent to agree on every quarter.
What this part asked of state systems
The practical work would start with an inventory. A state would list which systems held claims and capitation payments for each quarter, how each assigned enrollees to categories, and whether definitions changed when systems were replaced. The answers would decide whether a state could pick its best eight quarters with confidence.
The growth: inflation indexes in place of spending trends
The second input is the growth rate. Under the Senate version, the base would grow by CPI-M for children and nondisabled adults, and by CPI-M plus 1 percentage point for most disabled and elderly enrollees. [1] From 2025, all groups would grow by CPI-U. [1]
Why the index matters more than the base
A base sets the starting point. A growth rate decides the path. CBO’s projections showed why. Per-enrollee spending for children and nondisabled adults was projected to grow at 4.9 percent a year under current law, against 3.7 percent for CPI-M. [1] Every year, the cap would fall further behind projected costs for those groups.
For elderly and disabled enrollees, the picture differed in the early years. Their projected per-enrollee growth was 3.3 percent, below CPI-M plus 1 percentage point. [1] CBO said the cap would not generally bind for them before 2025, though some cost shifting among groups could still affect spending. [1]
The switch to CPI-U
A bigger change came in 2025. CBO projected CPI-U growth of 2.4 percent in 2025 and beyond. [1] It said the differences between Medicaid spending growth under current law and the caps would be substantial for all groups. [1]
In our reading, the switch made the cap’s long-run effect a matter of index choice. A general consumer price index tracks the cost of living, not the cost of medical care. Medicaid costs, as CBO projected them, grew faster. [1] The gap would widen each year after 2025.
What the index ignores
A price index does not capture changes in what care costs per person for reasons other than prices. New drugs, an aging enrollee population within a category and public health emergencies all raise per-enrollee costs. The formula, as CBO described it, did not adjust for them. [1] A state would bear those costs above the cap.
The House text allowed one route to correction, and it concerned data. A state could appeal for recalculation of its base and later amounts because of data errors. [2] Any adjustment from such an appeal could not raise the state’s target total spending by more than 2 percent. [2] In our reading, the narrowness of that remedy made the index the real policy, and made clean data the only reliable protection.
The index and the rate cycle
States set managed care rates each year, with actuaries certifying that they are sound. Those rates follow expected costs, not a price index. Under the cap, the state would face a federal limit growing with CPI-M or CPI-U, and plan rates growing with costs. [1] In our reading, each annual rate cycle would become a negotiation over who absorbs the difference: the state, the plans or the providers paid by the plans.
Mix within a category
Each category groups enrollees who differ. The elderly category includes people at 66 and at 96. The disabled adult category includes people with very different needs. [1] A price index does not track changes in that mix. If a state’s elderly enrollees grew older on average, their per-person costs could rise faster than prices. The cap would not follow.
In our reading, a state would need to watch the mix within each category as closely as the totals. Its actuaries already study such shifts when they set managed care rates. Under the cap, the same analysis would become a forecast of federal funding risk.
What this part asked of state budget offices
A state budget office would need to forecast the cap alongside its Medicaid spending. The two forecasts would use different drivers: the cap followed published price indexes, while spending followed enrollment, utilization and prices. [1] The gap between them would be the state’s exposure. Budget offices would need a model that tracked both, by category, years ahead.
The count: enrollees as the multiplier
The third input is the count. The limit in each year would be the per-enrollee amount for each category multiplied by the number of enrollees in that category. [1] Enrollment was the one input that moved with need. That made the count both a safeguard and a data obligation.
A cap that moves with enrollment
A per capita cap differs from a block grant. If enrollment rises, the cap rises with it. CBO’s description of the limit multiplied each group’s per-enrollee amount by its enrollment. [1] A recession that swelled enrollment would raise the federal limit. A block grant fixed in total would not.
The block grant option in the Senate bill shows the contrast. Its initial amount came from the per capita cap times the number of nondisabled adult enrollees in the year before the preceding fiscal year, adjusted for population growth plus 3 percentage points. It then grew with CPI-U. [1] CBO judged the option attractive mainly to states expecting population decline. [1]
Categories as classification work
The count was not one number but five. Each enrollee had to be placed in the right category: elderly, disabled adult, nondisabled child, expansion adult or other adult. [1] Each category had its own base and its own growth rate. [1]
Classification is a systems task. An enrollee’s category depends on age, disability status and eligibility group, which live in eligibility systems. A person can move between categories during a year. In our reading, a misclassified enrollee would move money between categories with different growth rates, and the error would compound.
What the House required to be reported
The House text was explicit about reporting. For each quarter beginning on or after October 1, 2018, states would include spending by categories of services and enrollees, and enrollee counts in each category, on the CMS-64 report. [2] The number of enrollees for a category was the average monthly number reported through the CMS-64 and subject to audit. [2]
The House text also paid for the work. For quarters from October 1, 2017 to before October 1, 2019, it raised federal matching for certain systems costs to 100 percent. [2] It raised matching for additional administrative costs of the new reporting to 60 percent. [2] In our reading, Congress recognized that the cap needed new data and offered money to build it.
States that expanded later
The House text addressed states that did not cover the expansion adult category in fiscal year 2016 but did later. For them, the expansion category’s target would equal the target for the other-adult category. [2] A state could not build its own base for a group it had not yet covered.
In our reading, that rule tied a future policy choice to past data from a different group. A state weighing expansion after 2016 would have to accept a per-enrollee amount based on other adults’ costs. Whether that amount fit the new group would depend on how similar the two populations were.
Monthly averages
The House text measured enrollees as the average monthly number reported through the CMS-64. [2] An average smooths the count across a year. It also means each month’s enrollment matters. A month of delayed renewals, or a backlog of pending applications, would lower the average and the limit with it.
Counting people who are not there
A count can be wrong in both directions. Enrollees who have moved away or died, and remain on the rolls, inflate it. Eligible people who lose coverage through paperwork lapses deflate it. Under a per capita cap, the first error would raise the federal limit and the second would lower it. [1]
Federal auditors would care about the first. A state’s finance office would care about the second. In our reading, the House text’s audit provision pointed at the risk of an inflated count. [2] A state that improved its enrollment hygiene would lower its count and its limit. In our reading, the cap would have rewarded accurate counts only if auditors enforced them.
What this part asked of eligibility systems
Eligibility systems would become budget systems. Their category assignments and monthly counts would feed the federal limit directly. A state would want controls on those assignments as strong as those on its financial reports. It would want reconciliations between the eligibility system, the claims system and the CMS-64. [2]
The response: who absorbs costs above the cap
The fourth part is the consequence. If a state spent more than the cap, the federal government would not pay its share of the excess. [1] Someone would absorb that cost. CBO named the options. [1]
The menu CBO described
With less federal reimbursement, CBO said, states would need to decide whether to commit more of their own resources or to reduce spending. [1] Spending reductions could come from cutting payments to providers and health plans, eliminating optional services, or restricting eligibility, including through work requirements. [1] States could also seek more efficient delivery, to the extent feasible. [1] CBO expected states to adopt a mix. [1]
Providers and plans
Payment cuts would land on hospitals, clinicians and managed care plans. CBO noted that if states reduced payment rates, fewer providers might be willing to accept Medicaid patients. [1] For a health system with a large Medicaid share, the cap would show up as rate pressure in contract negotiations with the state or its plans.
Managed care plans would sit between the state and providers. A state facing a cap would press capitation rates. Plans would press provider rates. In our reading, each layer would need its own forecast of the cap to negotiate well.
The matching rate for expansion adults
The cap was not the only Medicaid change. From 2021, the Senate bill would reduce the federal matching rate for expansion adults by 5 percentage points a year through 2023, then set it equal to the state’s rate for other enrollees. [1] For expansion states, the two changes would stack. The cap would limit federal dollars per enrollee, and the lower matching rate would shrink the federal share within that limit.
In our reading, an expansion state’s budget office would model the two together. The matching rate change would arrive first and be easier to forecast. The cap would bite later and depend on the index. A state could not plan for one without the other.
Work requirements as a lever
CBO listed restricting eligibility through work requirements among the ways states might respond. [1] It also noted an interaction between the caps and the effects of work requirements. [1] Work requirements are a data operation: a state must know which enrollees are subject to them, verify hours or exemptions, and act on the results. The same systems that fed the cap’s counts would carry those checks.
Enrollees
Enrollees would feel the response through services and eligibility. CBO said that if states left programs unchanged and found other ways to offset the loss of federal funds, enrollees would notice little or no change. [1] In our reading, if states cut services or eligibility, enrollees would notice. The cap did not decide which path a state chose. It set the size of the problem.
Limited flexibility under the cap
CBO made a point that cuts against a common description of caps. It said states would not have substantial additional flexibility under the per capita caps. [1] Under the block grant option, they would have more flexibility, such as altering cost sharing and, to a limited degree, benefits. [1] In our reading, the cap shifted financial risk to states without loosening many of the rules on how they ran the program.
Program integrity as funding protection
Under open-ended matching, an improper payment costs the federal and state governments their shares. Under a cap, it would also use up room under the state’s limit. [1] In our reading, every dollar recovered from an improper claim would restore room for legitimate care. That gives program integrity units a budget argument they lacked before: recoveries would protect the state’s capacity under a fixed federal ceiling.
The House’s clawback timing
The House text set out how the excess would be recovered. If a state’s spending exceeded its target for a fiscal year, payments in each quarter of the following year would be reduced by one quarter of the excess. [2] That design put a year’s lag between spending and penalty. A state would learn its exposure after the year closed, and pay through the next year’s quarters.
For a state treasury, that lag is a cash management issue. A reduction arriving after the year it relates to has to be reserved for. In our reading, a careful state would hold a contingency against an excess it could not yet measure precisely.
How the cap would have reached different public bodies and providers
The cap was a federal formula with state consequences. The cases below are our reading of how its parts would have fallen, not findings about any state.
State Medicaid agencies
The agency would own the base, the categories and the counts. [1] [2] It would need its eligibility, claims and finance systems to agree with each other and with the CMS-64. Its fiscal agent and systems vendors would carry much of the new reporting work. [2]
The agency would also face a governance question. Today’s eligibility categories serve program rules: who qualifies, for what benefits. Under the cap, the same categories would also serve federal finance. A change in how a category is coded, made for operational reasons, could move money. In our reading, the agency would need a change control process that brought finance into decisions that had belonged to eligibility staff alone.
State budget offices and legislatures
The budget office would forecast the cap and the exposure above it. [1] The legislature would decide whether to fund the gap or direct cuts. The cap’s growth rate, set in federal law, would become a fixed constraint in state budget debates.
Budget offices would also need new reports. A quarterly comparison of spending against the cap, by category, would show where pressure was building. Without it, a state could learn of an excess only after the year closed, when the House text’s recovery from the next year’s payments would begin. [2]
Counties and local governments
The House text contained a provision decreasing target expenditures for required expenditures by certain political subdivisions, for states meeting a specific condition. [2] We do not attempt to map how each state’s local governments would have been affected. In our reading, local public hospitals and county health systems would face the same rate pressure as other providers.
Systems vendors and fiscal agents
States run Medicaid systems with contractors: fiscal agents, eligibility system vendors and data warehouse providers. The House text’s enhanced federal match for systems work ran only for quarters before October 1, 2019. [2] In our reading, that window would have compressed procurement. States would have needed change orders or new contracts quickly to use the money while it lasted.
Compressed procurement carries risk. Requirements written in haste can miss what the formula needs. A state would want its contracts to specify category definitions, reconciliation reports and audit support, all tied to the CMS-64 reporting rules the Secretary would issue. [2]
Health systems
A health system would see the cap through rates, supplemental payments and uncompensated care. CBO projected 15 million fewer Medicaid enrollees by 2026. [1] Some of those people would become uninsured patients. For a system with a large Medicaid share, the combined effect would show up in payer mix and bad debt.
A health system’s own data would matter here too. Its eligibility screening at registration, its help with enrollment and renewal, and its records of uncompensated care would all bear on how many of its patients kept coverage. In our reading, a system that tracked those flows by payer and by month would see the effect of state choices before it reached the annual financial statements.
Auditors
Auditors would follow the money into the data. The House text required federal audits of each state’s reported enrollees and expenditures for fiscal year 2016, fiscal year 2019 and each later year. [2] In our reading, state auditors would have reason to examine the same figures, since they would set state budget exposure. An auditor testing a count would trace enrollees from the CMS-64 back to the eligibility system and to evidence of eligibility. A state that could not support that trail would face findings with a funding cost.
Managed care plans
Plans would see the cap through capitation rates. They would also hold much of the encounter data a state would need to prove its costs by category. In our reading, a state’s data agreements with its plans would become part of its federal funding defense.
Objections, in order
1. The bill failed, so its design is moot
The bill failed in the Senate in July 2017. [3] [4] [5] The design matters because its core idea, federal funding tied to state data about each enrollee, did not go away. The 2025 law tied eligibility, and the counting of erroneous payments, to new data checks rather than to a funding formula. [6] The systems questions are the same.
2. Letting states choose their base quarters was generous
It was a concession. CBO’s projections show where the cap’s effect came from: the growth rate, especially after the switch to CPI-U. [1] A generous base delays the cap’s bite. It does not change its slope.
3. A per capita cap protects states in a downturn
It does, compared with a block grant, because the limit moves with enrollment. [1] That protection depends on the count being accurate and timely. It does nothing for higher costs per enrollee, such as those from a new drug or an epidemic. [1]
4. States would gain flexibility in exchange
CBO said states would not have substantial additional flexibility under the caps. [1] Flexibility came mainly with the block grant option, which CBO expected few states to take. [1]
5. Data quality is a technical detail
Under the cap, data quality would be money. A misclassified enrollee or a missing quarter of claims would change the federal limit. [1] [2] The House bill’s audit requirement and its temporary 100 percent federal match for systems work show that Congress saw the data as part of the policy. [2]
6. CBO’s projections were too uncertain to plan on
CBO said so itself. It noted that its estimates relied on its March 2016 baseline, and that enrollment and subsidies under current law would probably differ from that baseline. [1] It also said the direction of certain effects was clear, including lower Medicaid spending. [1] For our argument, the size of the gap matters less than its mechanism. The data obligations would exist at any projected gap.
7. States already lived with per capita limits under waivers
The House text recognized that waivers could carry per capita limits, and addressed them directly. It said the new limits would supersede any other payment limitations under a waiver, including limitations based on a per capita limitation. [2] In our reading, a waiver limit is negotiated and time-limited. The statutory cap would have been permanent and the same for every state. The data demands would have been similar. The room to renegotiate would not.
What a state Medicaid or budget office could check
The four parts translate into checks a state can run on its data, whatever federal financing rules apply.
- Can the agency produce consistent per-enrollee costs by eligibility category for every quarter of the last several years?
- Do enrollee category assignments in the eligibility system reconcile to claims and to the CMS-64?
- Does the budget office model federal funding under alternative growth formulas, by category?
- How quickly does the agency remove enrollees who have died or moved, and how is that measured?
- Do data agreements with managed care plans guarantee complete encounter data on a fixed schedule?
The first two checks address the base and the count. The third addresses growth. The fourth and fifth address the data a state would need to defend its funding and its eligibility decisions. [1] [2] [7]
The five checks do not depend on a per capita cap returning. Each improves a state’s position under the rules that did arrive: more frequent redeterminations, address checks and verification of community engagement from data the state already holds. [6] [7] The same data that would have defended a funding base now defends individual eligibility decisions, case by case and month by month.
With hindsight
The Senate did not pass the bill. On July 25, 2017, in floor debate that touched on the most recent version of the Better Care Reconciliation Act, amendment No. 270, offered to amendment No. 267 to H.R. 1628, was called up. [5] A motion to waive budget rules for that amendment, which required three-fifths, was rejected 43 to 57. [3] On July 28, 2017, the Senate rejected amendment No. 667 to the same bill, 49 to 51. [4] The per capita cap was not enacted that year.
Medicaid financing returned to reconciliation in 2025. Public Law 119-21 was approved July 4, 2025. [6] It did not adopt a per capita cap. Its Medicaid changes worked through eligibility and payment integrity, and the ones we describe here depend on state data.
Section 71103 requires states to have a process to regularly obtain enrollees’ address information, beginning not later than January 1, 2027. [6] Beginning not later than October 1, 2029, states must submit enrollees’ Social Security numbers and other information to a federal system at least monthly, under a section headed reducing duplicate enrollment. [6] Section 71107 requires eligibility redeterminations every six months for adults in the expansion group, for redeterminations scheduled from the first quarter beginning after December 31, 2026. [6]
Section 71106 changes how erroneous payments are counted. Payments made where insufficient information is available to confirm eligibility will count as erroneous excess payments, and the Secretary’s ability to waive the resulting reductions is limited. It applies from fiscal year 2030. [6] In our reading, that ties federal dollars to the completeness of state eligibility records, a cousin of the cap’s logic.
Section 71119 requires states to make community engagement a condition of eligibility for certain adults. They must start no later than the first quarter beginning after December 31, 2026, unless the Secretary grants a good faith effort exemption. [6] An individual can meet it by working at least 80 hours a month, among other options. [6] CMS published an interim final rule implementing the requirement on June 3, 2026, effective July 31, 2026. [7] It said states must implement no later than January 1, 2027. [7]
The rule leans on state data. It requires states to verify compliance first on an ex parte basis, using reliable information available to the state without asking the applicant for more. [7] It said states should consider their access to timely data when setting how many months a beneficiary must show compliance at renewal. [7] It said CMS intends states to use data sources and programming logic readily available to them, rather than build new methods and systems for the requirement. [7]
What held up from June 2017? CBO’s account of the cap described a policy in which state data set federal dollars. [1] The House text added audits and paid for the reporting systems. [2] The 2025 law chose a different lever. It made state data decide which individuals keep coverage, and it counts unconfirmed eligibility against states in its payment rules. [6] [7] Either way, the state’s eligibility and claims systems carry the policy. A state that built clean category data for a cap that was not enacted would find the same data useful for the checks that did.
Spotted an error? Tell us and we will check it and correct the article where needed.