The House budget vote signaled Medicaid churn, a cost hospitals pay at the front desk

The week

Some of the most consequential decisions in Washington arrive looking like paperwork. On February 25, 2025, the House of Representatives adopted a budget resolution by 217 votes to 215, about as narrow a margin as a vote can have.

Democrats voted against it as a bloc, and one Republican joined them. The vote came four days after the Senate had adopted its own budget resolution on February 21. The Senate plan was narrower, focused on border security, defense and energy, and left the larger questions of tax and spending cuts for later. The House chose a single, larger package instead.

A budget resolution is not a law. It is a blueprint Congress adopts to set spending and revenue targets. What made this one important was a feature called reconciliation. A budget resolution can include reconciliation instructions telling specific committees to change the law enough to hit certain targets. Legislation that follows those instructions can then pass the Senate by a simple majority, rather than needing the larger majority most bills require to overcome a filibuster. So the resolution was a procedural step that opened a faster road for the legislation that would follow.

Among its reconciliation instructions, the resolution directed the Energy and Commerce Committee, which oversees Medicaid, to reduce the deficit by at least $880 billion over ten years. Medicaid is the joint federal and state program that pays for health care for tens of millions of lower-income Americans, and it is a large share of what that committee oversees. The target was widely read as requiring substantial Medicaid cuts.

The resolution set out other instructions too. It allowed the Ways and Means Committee to add up to $4.5 trillion to deficits over ten years, mainly to extend the 2017 tax cuts. It paired that with a target of $2 trillion in spending reductions. If committees found less than the target in cuts, the tax allowance would shrink by the same amount. The resolution also called for a $4 trillion increase in the debt limit and directed the Agriculture Committee, which oversees food assistance, to find $230 billion in savings.

The arithmetic behind the Medicaid reading was confirmed shortly afterward. In early March 2025, the Congressional Budget Office set out the spending under Energy and Commerce’s jurisdiction. Excluding Medicare, which leaders had said they would not cut, that spending came to about $8.8 trillion over the decade, and Medicaid accounted for about $8.2 trillion of it. An $880 billion reduction from that base could not be found without Medicaid.

Reaction was swift. Hospital associations, state officials and patient groups warned of effects on coverage and on providers serving many Medicaid patients. They pointed to safety-net hospitals in cities and to rural hospitals where Medicaid is a big part of the payer mix.

For providers, the exposure was not evenly spread. Medicaid is a large payer for children’s hospitals, for births, for behavioral health and for long-term care. It is also a dominant payer for community health centers. Forty states and the District of Columbia had expanded Medicaid under the Affordable Care Act, and their hospitals had built budgets around the expansion population most likely to be affected.

The details came later. Over the following months the policies took shape, and they were enacted in the reconciliation law signed on July 4, 2025. That law included Medicaid work requirements for many “expansion adults,” the working-age adults who became eligible when states expanded Medicaid under the Affordable Care Act. It required more frequent eligibility checks for that group, and it set limits on mechanisms states use to finance their share of Medicaid.

Between February and July the path ran through several more procedural steps. In early April the Senate adopted an amended version of the House resolution, and the House agreed to it days later, so both chambers were working from the same blueprint. Committees then wrote the legislation, the House passed it in May, and the Senate passed a revised version on July 1 with the Vice President breaking a tie. The House agreed on July 3, and the President signed it the next day.

In February that detail was still unsettled. But the direction was visible, and for hospitals and health systems the vote raised two questions. The obvious one was how much revenue might be lost. The less obvious one was how much it would cost simply to keep track of who was covered.

Most health system reactions to the vote focused on revenue: how much Medicaid funding might be lost and what that would mean for margins, especially at safety-net and rural hospitals. That was the right first question. The second was less visible. Policies that make eligibility less stable impose a cost on providers even for patients who keep their coverage, because someone has to establish, at each encounter, who is covered and by whom.

The 2023 to 2024 Medicaid unwinding, when states resumed eligibility reviews after the pandemic and many people lost coverage for procedural reasons, had already shown how large that cost could be. “Procedural” is the key word. Many of those people were still eligible; they lost coverage because a form went to an old address or a renewal deadline was missed. For the hospitals that treated them, the result was the same as if they had become ineligible: uncertainty at the front desk, unpaid claims and difficult conversations about bills.

The cost lands in three places, defined by when in the patient’s journey it is incurred. The first is before and at registration, when coverage is checked. The second is between visits, when patients need help keeping or regaining coverage. The third is after care, when claims meet coverage that has changed. Each has a different owner, access, financial counseling and the revenue cycle, and each can be planned before the policy details are final.

The weight of each place differs by type of provider. A safety-net or academic medical center sees all three at volume, with emergency departments that cannot screen for coverage before care. A rural or critical access hospital has thinner staffing, so a few extra hours of eligibility work each day is a real share of its front office. Community health centers and other ambulatory networks see patients more often but in lower-cost visits, which makes the cost of each coverage check large relative to the payment. Behavioral health providers serve many enrollees whose circumstances change often. Location matters as much as type: a hospital in a state that never expanded Medicaid faces far less exposure to rules aimed at the expansion population than one in an expansion state.

Churn, renewals and the revenue cycle

Churn is the health policy word for people cycling on and off coverage. Someone is enrolled in Medicaid, loses coverage at a renewal, goes uninsured for a few months, then re-enrolls. Every one of those transitions creates work and uncertainty for the people who provide their care.

Churn between plans

Churn also happens inside Medicaid. In states that rely on managed care, an enrollee who loses coverage and re-enrolls may be assigned to a different plan, sometimes automatically if they do not choose one. Each plan has its own network, prior authorization rules and claims address. A patient who is covered throughout, but by a new plan, can still generate a denial if the claim goes to the old one. For a hospital, a change of plan creates almost as much work as a gap in coverage.

A few other terms recur. Eligibility is whether a person qualifies for coverage on a given day. A redetermination, or renewal, is the periodic review in which a state checks that a person still qualifies. Medicaid managed care plans are private insurers that states pay to run Medicaid benefits for many enrollees. A patient’s coverage may therefore sit with one of several plans rather than with the state alone. And the revenue cycle is the whole chain of work by which a hospital gets paid: registering the patient, confirming coverage, recording the care, sending the claim, chasing denials and collecting any balance.

How does a hospital actually check coverage? In most cases, its registration system sends an electronic query to the state Medicaid system or to the patient’s plan, which replies with whether the person is covered and under which plan. It is a little like a shop running a card through a payment terminal: the answer comes back in seconds. But a card can be declined after the meal has been eaten. In the same way, coverage confirmed on Monday can turn out to have lapsed, or changed plans, by the time the claim is processed weeks later.

When eligibility becomes less stable, every one of those steps gets harder and more expensive, even for patients who stay covered.

What the electronic check actually returns

The query and reply follow a federal standard. Under HIPAA’s transaction rules, eligibility inquiries and responses use standard electronic formats known as the 270 and 271 transactions. The reply can say whether a person is active on a given date, which plan is responsible and, depending on the payer, what benefits apply.

The answer is only as current as the data behind it. State eligibility systems update on their own schedules, and assignment to a managed care plan can lag enrollment. A renewal that is pending, or a termination that will take effect at month end, may not show at all. So an “active” reply on the day of a visit is strong evidence, not a guarantee. A hospital that checks once has a single point-in-time answer for a patient whose coverage may change before the claim is paid.

What the unwinding showed

During the pandemic, Congress paid states extra Medicaid funding on condition that they keep enrollees covered continuously. That protection ended on March 31, 2023, and states then worked through renewals for everyone over roughly a year. KFF, the health policy research organization, tracked more than 25 million disenrollments across the unwinding and found that most, where the reason was known, were procedural rather than findings that people no longer qualified.

System design mattered too. In 2023 federal officials found that many states were running automated renewals at the household level rather than for each individual. Children who still qualified lost coverage because a parent did not. The states involved had to pause procedural terminations and reinstate coverage for about half a million people, many of them children. For hospitals, the episode showed that coverage loss could come from how a state system was configured, as well as from rules or from patient behavior.

Before and at registration: checking coverage more often

Every encounter starts with an eligibility question, answered through electronic checks against state systems and Medicaid managed care plans. If enrollees must re-establish eligibility more often, or lose coverage for failing to document work or an exemption, more patients will arrive with coverage that has lapsed, changed plans or is in dispute.

Front-desk and scheduling staff feel it first. The practical response is automation: eligibility checks run automatically several days before each scheduled visit and again on the day, with changes flagged to staff before the patient arrives. Most modern registration systems can do this; the question is whether it has been switched on, tuned and staffed so that someone acts on the flags.

Health systems should also track a small set of measures, such as the share of visits with a coverage change since the last encounter and the share where coverage is unresolved at check-in. Those numbers show whether churn is rising before denials do.

Sizing the front-desk workload

A health system can estimate the extra workload before the rules take effect. The inputs are the number of Medicaid visits per month, the share of those patients whose coverage is likely to change at each renewal point and the staff minutes needed to resolve a flag. Moving from annual to six-month renewals roughly doubles the renewal points for expansion adults. The estimate will be rough. It still turns an abstract policy change into a staffing request with a basis, which is more persuasive in a budget meeting than a general warning.

Accurate demographic data also matters more. A patient’s current address and phone number, confirmed at every visit, are what the state and the plan need to reach them at renewal. The July 2025 law directs states to use reliable data sources to keep enrollee addresses current, but the hospital often learns of a move first. Front-desk scripts that confirm contact details, and a route to encourage patients to update the state, cost seconds per visit.

Scheduling is also the cheapest moment to warn a patient. A call or message that says coverage could not be confirmed, with an offer of a counselor appointment before the visit, resolves many problems early. Otherwise they surface at the front desk, or weeks later as a bill.

For unscheduled care, especially in emergency departments, hospital presumptive eligibility, for hospitals that have qualified under their state’s rules, gives staff a route to temporary coverage at the time of service. A qualified hospital can, under the state’s rules, make a quick provisional decision that a patient likely qualifies, so that care is covered while a full application is processed.

Presumptive eligibility in more detail

Since 2014, federal law has let qualified hospitals in every state make presumptive eligibility determinations for Medicaid, under rules each state sets. Hospitals must elect to take part, train the staff who make the decisions and meet the state’s accuracy standards. Temporary coverage generally runs to the end of the following month, or until the state decides a full application if one is filed. The program is underused partly because it needs trained staff on every shift, including nights and weekends, which is exactly when unscheduled patients arrive.

Federal law also shapes the emergency case. Under EMTALA, hospitals with emergency departments must screen and stabilize patients regardless of ability to pay. Coverage questions therefore cannot delay emergency care, and registration in the emergency department has to work around care rather than ahead of it. That is why the preparation for unscheduled patients has to be in place before they arrive.

What the new rules change at the front door

The law that followed the February vote changed the front-door problem in specific ways. Six-month redeterminations for expansion adults double the number of points each year at which coverage can lapse. Work and community engagement requirements add a new reason for coverage to end even for someone whose income still qualifies.

Two further provisions reach the front desk directly. The law shortened retroactive coverage, the period before an application that Medicaid will pay for once a person is approved. For applications from 2027, it falls from three months to one month for expansion adults and to two months for other enrollees. A lapse discovered late now costs more, because less of the gap can be covered after the fact. And from October 2028 states must impose cost sharing of up to $35 per service on expansion adults with incomes above the poverty line. The exceptions include primary care and some behavioral health services. Front desks that rarely collected from Medicaid patients will need to start.

Different settings, different checks

A clinic that sees scheduled patients can run checks days ahead and on the day. An emergency department can check only after the patient arrives and must not let the check delay care. An ambulatory network with many small sites may run checks centrally, with a shared team working the flags for all locations. Telehealth visits need the same check, though there is no front desk to catch a problem. The design question for each setting is who sees the flag and how quickly they can act on it.

Between visits: helping patients keep coverage

Many health systems employ financial counselors who help patients apply for and renew Medicaid. The unwinding showed that procedural losses, such as missed renewal notices, out-of-date addresses and incomplete forms, could be prevented with timely help. Work and community engagement requirements add a new category: patients may need to understand what counts, gather documentation, claim an exemption and report on time.

Health systems choose how much of this support to provide, but the cost of not providing it returns as uncompensated care, the care a hospital delivers without being paid. Planning questions include counselor capacity for the expected volume, training on the new rules once states publish them, and outreach timed to each patient’s renewal date.

Data drives that timing. Where states share renewal dates or managed care plans pass them on, a health system can contact patients before their review, which costs far less than re-enrolling them after a lapse.

Many exemptions, such as those for people with certain medical conditions, rely on documentation clinicians already hold, so workflows that let counselors request it from the care team quickly will matter. A counselor who has to chase a busy physician by email for a letter will lose days. One who can send a structured request through the electronic health record, with a set turnaround, will not.

What the work requirement asks of an enrollee

Under the July 2025 law, expansion adults aged 19 to 64 generally must show 80 hours a month of qualifying activity. That can be work, community service, a work program, at least half-time education or a combination. Monthly income equal to 80 hours at the federal minimum wage also qualifies. States must check compliance at application, for a period of one to three months chosen by the state, and again between renewals.

The exemptions are long. They include parents and caretakers of young children or of people with disabilities, and pregnant and postpartum women. People who are medically frail are exempt, as are people with disabling mental or substance use conditions, people in substance use treatment and some veterans with disabilities. People already meeting work rules for SNAP or TANF are exempt too. States must use data they already hold to confirm compliance or exemptions where they can, so that enrollees do not have to report what the state can verify itself.

For a health system, the practical burden concentrates on the medically frail and similar exemptions, because the evidence for them lives in clinical records. A counselor needs a fast way to identify patients likely to qualify, ask the care team for documentation and send it to the state with the patient’s consent. A patient who qualifies for an exemption but loses coverage for lack of a letter is the most avoidable loss in the system.

The economics of counseling

Counseling is cheap relative to an uncompensated inpatient stay and expensive relative to a clinic visit. That shapes where it should go first: patients with frequent or high-cost care, scheduled procedures and chronic conditions. Many hospitals use eligibility vendors, often paid a share of what they recover. That model scales quickly but can tilt effort toward high-balance accounts and away from the routine renewals that prevent losses. Contracts should reward coverage kept as well as applications filed, and should set standards for how vendors contact patients on the hospital’s behalf.

After care: claims that meet changed coverage

Coverage instability produces denials, retroactive eligibility changes and patients moving between Medicaid, marketplace plans and no coverage. Each transition creates billing work: rebilling, appeals, charity care decisions and difficult conversations about patient balances.

Revenue cycle leaders can model scenarios for denial rates and self-pay volume, review charity care policies for a larger uninsured population, and decide whether to accelerate automation in denials management. Vendor lead times for such tools are long, which argues for deciding early.

Managed care contracts also matter. Plans facing their own enrollment and financial pressure may tighten prior authorization, the requirement to get a plan’s approval before certain care, or slow payment. Terms on retroactive eligibility, timely filing and dispute resolution decide who absorbs the cost when coverage changes after care is delivered, so they deserve review before renewal. “Timely filing” is the deadline by which a claim must be submitted. If the hospital learns weeks later which plan a patient belonged to, a tight deadline can turn a payable claim into a write-off.

Shorter retroactive coverage

Retroactive coverage has long been a safety valve for hospitals. A patient admitted while uninsured could apply afterward, and if approved, Medicaid would pay for care up to three months before the application month. From 2027, that window shrinks to one month for expansion adults and two for other enrollees. A hospital that relied on retroactive enrollment to recover payment for uninsured admissions will recover less. The response is to move enrollment earlier: apply during the stay, not after discharge, and use presumptive eligibility where it fits.

Charity care obligations

Nonprofit hospitals face their own rules. Section 501(r) of the tax code requires them to have a written financial assistance policy and to limit charges to patients who qualify for it. They must also avoid extraordinary collection actions until they have made reasonable efforts to determine eligibility. More patients losing Medicaid means more patients who may qualify for financial assistance, and more applications to process before any collection can begin.

Some hospitals use presumptive charity scoring, which estimates financial need from external data, to grant assistance without a full application. That can reduce administrative cost and patient burden, but it should not replace an attempt to restore Medicaid for those who still qualify. Coverage pays for future care. Charity care covers only the bill in hand.

Denials data should flow back to the front of the process. If a growing share of denials trace to coverage that changed between scheduling and service, the fix is earlier and repeated checks, not more appeals staff. A monthly review that sorts eligibility denials by root cause and assigns each cause to registration, counseling or billing keeps the three owners working on the same problem.

Reading the denial codes

Payers explain denials using standard claim adjustment reason codes in the electronic remittance advice. Several codes point directly to coverage. Code 27, for example, means expenses incurred after coverage terminated, and code 31 means the patient cannot be identified as the payer’s insured. Grouping denials by these codes, and by the plan and location involved, separates eligibility problems from coding or authorization problems. It also shows which front-end fix would have prevented each one.

The same data supports conversations with plans. A pattern of coverage-related denials for patients the plan’s own eligibility response showed as active is a contract issue, not a registration failure, and the hospital should raise it as one.

One missed renewal notice, two outcomes

Take a patient at a community hospital: a working-age adult covered by Medicaid through a managed care plan, with a chronic condition that brings her in every few months.

In the spring, her renewal notice goes to an apartment she moved out of. She does not see it, misses the deadline and loses coverage, though she still qualifies. Two weeks later she has a scheduled follow-up. If the hospital runs an automated eligibility check five days before the visit, the lapse is flagged, a scheduler calls, and a financial counselor helps her reapply before she arrives. If not, the front desk discovers the problem at check-in, the visit goes ahead with coverage unresolved, and a claim is later denied.

In the first version, the cost to the hospital is a phone call and some counselor time. In the second, it is a denied claim, staff time on rebilling, a possible charity care application and a confused patient with a bill she cannot pay. There is also a strong chance she delays her next visit, making her condition, and her eventual care, more expensive.

Under the 2027 rules, the late discovery costs more still. If her coverage lapsed two months before the hospital noticed, and she is an expansion adult, retroactive coverage after reapplying reaches back only one month. One month of her care has no payer.

Now suppose a second patient arrives unscheduled, in the emergency department, late on a Saturday. There is no scheduling call and no chance of a check five days ahead. His coverage comes back as inactive. Here the hospital’s options depend on preparation done long before he walked in. If the hospital has qualified for presumptive eligibility and has trained staff who can use it at any hour, he may leave with temporary coverage and an appointment with a counselor. If not, his care is provided anyway, as it should be, and the question of payment is left to the billing office weeks later, when he is much harder to reach.

The two cases point to the same conclusion. The cheapest moment to resolve coverage is the earliest one, and the earliest moment only works if the systems, staff and procedures are in place before they are needed.

Multiply that difference across thousands of patients and the case for investing early in automated checks, counseling capacity and denial analysis becomes a financial argument as well as a mission one.

Building the shared view: who owns what

One reason the cost of churn stays hidden is that it is spread across departments that rarely meet. Patient access, which runs scheduling and registration, sees unresolved coverage at check-in. Financial counseling sees the patients who need help renewing. The revenue cycle team sees the denials weeks later. Each sees a slice; no one sees the whole.

A practical fix is to give the problem a single executive owner, often the chief financial officer or the head of revenue cycle. That owner needs a short monthly report that follows patients across all three stages. The report does not need to be elaborate. It needs a handful of numbers that move together: coverage changes found before visits, coverage unresolved at check-in, patients helped by counselors and how many kept coverage, and eligibility-related denials by root cause. Broken down by location and service line, those figures show where churn is concentrated, whether early checks are working and where the money is being lost.

Technology helps, but it is not the starting point. Most of the data already exists in the registration system, the counseling team’s case notes and the billing system. The work is in connecting them, agreeing on definitions and making sure someone with authority reads the result every month. A common trap is to commission a large analytics project and wait for it. A simpler report built from existing systems, produced by hand at first if necessary, delivers most of the value sooner.

Definitions deserve care. “Coverage change” should mean the same thing in the registration system and in the billing report, and a change of plan should be counted separately from a loss of coverage. Data access needs the same discipline. The report combines eligibility, clinical and financial information, so it should be built from the minimum data each measure requires, with access limited to the people who act on it. A report that compliance has approved at the outset is one that survives the first audit.

The same view also supports conversations outside the organization. Managed care plans, state Medicaid agencies and community partners respond better to specific evidence than to general concern. Some hospitals can show which communities are losing coverage, at what point in the renewal process and with what effect on care. They are in a stronger position to ask for better data sharing, simpler renewal processes or help with outreach.

Data the hospital does not hold

Renewal dates are the most useful data point and the hardest to get. States hold them. Managed care plans often know them for their members. Some states share renewal information with plans and providers, and HIPAA generally permits disclosures needed for payment and treatment, though state rules and contracts govern what is shared. A health system should ask each major plan and its state Medicaid agency what renewal data they will provide and in what format. A monthly file of upcoming renewals turns outreach from guesswork into a schedule.

Finally, the view helps the board. Directors are used to seeing payer mix and margin. Adding a small set of churn measures alongside them turns an abstract policy risk into something the board can track quarter by quarter. It also connects the hospital’s mission to its finances in a way that is easy to explain.

A sensible order of work

The work does not all have to happen at once. In the first quarter, the essentials are switching on automated checks and building the simplest monthly report. The second quarter adds renewal outreach and a documentation route for exemptions. Over the following year, the work moves to managed care contract renewals, denials automation and staff training on each state’s rules as they are published. Sequenced this way, each step produces data that informs the next.

Thin margins and unglamorous preparation

All three places draw on the same data: eligibility status, renewal dates and payer mix by population and location. A shared view of that data, reported monthly to leadership, shows where coverage loss is concentrated and what it costs by service line. In February 2025 the details of the law were unknown, but the direction was clear enough to start building that view.

The stakes are concrete. Margins at hospitals that serve many Medicaid patients are thin, so a rise in denials or uncompensated care can be the difference between surplus and deficit. Staff at the front desk and in counseling roles absorb the stress of harder conversations, which affects retention. Patients who lose coverage delay care. And communities notice which hospitals helped them keep coverage and which simply sent bills.

Good preparation is unglamorous. Automated checks are running, with someone acting on the flags. Churn measures are tracked monthly. Counselor numbers are matched to expected need and trained on each state’s rules. Managed care contracts are reviewed with coverage changes in mind. And denial analysis loops back to registration and counseling rather than living only in the billing office.

The failures are equally familiar. One is treating eligibility as a one-time check at registration. Another is measuring counseling by applications filed rather than coverage kept. Others are buying denials software late, after volumes have already climbed, and letting each department solve its own piece without a shared view of the data.

The case for doing less

A reasonable chief financial officer could argue against much of this. If policy reduces coverage, revenue falls whatever the hospital does, and spending on counselors and automation adds cost to a shrinking base. Some of the people who lose coverage will not qualify again, so outreach effort will often fail. Vendors selling eligibility tools have an interest in making the problem sound larger than it is.

Each point has some force. Not every hospital needs new software, and counseling capacity should follow evidence rather than projections. Still, the costs described here fall on patients who remain eligible or could become eligible again, and those patients are numerous in any period of churn. Early automated checks cost little once configured. The arithmetic of a single avoided inpatient write-off pays for a great deal of counselor time. The case for preparation rests on that difference, and on the fact that the work takes months to build.

There is a policy counter-argument too. Supporters of work requirements argued that they encourage employment and focus the program on those most in need. Evidence from Arkansas, which applied a work requirement in 2018, cut against that claim. More than 18,000 people lost coverage before a federal court halted the program, and a peer-reviewed study found coverage fell without a measurable gain in employment. Whatever one’s view of the policy, hospitals must plan for its administrative effects.

Where churn will show up first

1. Do we check coverage automatically before every scheduled visit and again on the day, and who acts on the changes it finds?
2. What share of visits last month involved a coverage change since the patient’s previous encounter, and what share had coverage unresolved at check-in?
3. How many financial counselors do we have, how many patients could need renewal or work-requirement help, and how will we train counselors when states publish their rules?
4. What do our managed care contracts say about retroactive eligibility, timely filing and disputes, and when do they renew?
5. What share of our denials trace to coverage changes, and who owns fixing each root cause?

With hindsight

The reconciliation law signed in July 2025 turned the February blueprint into specific obligations. It required states to apply work and community engagement requirements to many adults in the Medicaid expansion population, generally from the start of 2027. States showing good-faith effort could seek temporary exemptions. It required eligibility redeterminations every six months for expansion adults, rather than the annual reviews most enrollees had been used to.

Some ideas debated in the spring did not make it into the final law, including per-capita caps on federal Medicaid funding and a lower federal share for the expansion population. The savings came instead through eligibility rules and financing limits, the route that most directly raised administrative costs for providers.

The details reached further than the headline provisions. Work requirements apply to expansion adults aged 19 to 64, with a long list of exemptions, and states must verify compliance using data they already hold where possible. The law required the Department of Health and Human Services to issue an interim final rule on the requirements by June 2026. Six-month renewals apply to renewals scheduled after the end of 2026. Retroactive coverage shrinks for applications from 2027, and cost sharing for expansion adults above the poverty line begins in October 2028.

The law also postponed, until 2034, parts of two recent federal rules that had been meant to simplify Medicaid enrollment and renewal. Some of the streamlining that would have reduced procedural losses will therefore not arrive during the period when the new requirements take effect.

The law also phased in limits on provider taxes and state-directed payments. Those are two of the mechanisms states use to finance their share of Medicaid and to channel extra funding to hospitals. The limits therefore mattered directly to provider finances as well as to state budgets. The law also created a rural health transformation fund.

That fund, the Rural Health Transformation Program, provides $50 billion over five years starting in fiscal 2026, distributed to states under plans approved by the Centers for Medicare & Medicaid Services. It offsets some of the pressure on rural providers, though it is structured as transformation funding to states, not as a direct replacement for lost Medicaid payments to particular hospitals. Provider tax limits begin to phase in from fiscal 2028, so their full effect on hospital payments arrives later than the eligibility changes.

The Congressional Budget Office, the nonpartisan agency that estimates the effects of legislation for Congress, projected that the law would increase the number of uninsured people by about 10 million by 2034.

Through 2026 states worked on the systems and rules needed to verify work requirements, many through vendor-operated eligibility platforms. That work is itself a technology project of real scale: states need to collect, check and track new kinds of information about enrollees, on a deadline.

Earlier work-requirement programs were a guide to what the systems must handle. Georgia’s Pathways program, which began in July 2023, enrolled far fewer people than the state had expected, and its experience was widely cited as evidence that reporting systems themselves drive coverage outcomes. Arkansas’s 2018 program had shown the same thing earlier. States designing for 2027 had to decide how much they could verify automatically from wage, benefit and other data, and how much enrollees would have to report.

For health systems, the front-desk, counseling and revenue cycle costs described here moved from scenario to schedule. Six-month redeterminations mean twice as many moments each year when a patient’s coverage can lapse. Work requirements add a new reason for coverage to end even when a person would otherwise qualify. Both multiply the number of encounters where staff must establish who is covered.

Policy changes to eligibility carry operational costs that do not show up in headline revenue estimates. Health systems that began building automated checks, counseling capacity and a shared view of coverage data in early 2025 had time on their side. Those that waited for the final rules faced the same work with less runway.