AT A GLANCE

• Medicaid managed care accounts for roughly half of all Medicaid spending nationally.
• Capitation payments fund not only care delivery, but also plan administration, reserves, and margin.
• Evidence that managed care improves outcomes in long-term services and supports is mixed and inconclusive.
• Nursing facility care is rate-driven, capacity-constrained, and structurally unlike retail health care markets.
• States are entering a period of rising demand and tightening federal support.
• Plan overhead—not care—should be the first target for structural reform.

I. The Question States Are Not Yet Asking Clearly

Medicaid is approaching a constraint point. Demand for long-term services is increasing. Federal financial support is tightening. State budgets are under pressure. None of those trends are temporary.

In that environment, states face a choice. They can reduce services—or they can reduce the cost of delivering those services. The first path is politically and morally difficult. The second requires an honest examination of how Medicaid dollars move before they reach providers.

In most states, a substantial portion of those dollars flows through managed care organizations before reaching the nursing facility. That structure is rarely questioned. It should be.

When budgets tighten, states should cut the middleman before cutting care.

Long-stay nursing facility care is not a discretionary service. For most residents, it is the only setting in which they can safely receive the level of support they require. Before any state considers reducing benefits, reducing provider rates, or deferring quality investments, it should first examine whether the architecture of its delivery system is more expensive than it needs to be.

This brief argues that in the nursing facility market specifically, managed care has not clearly demonstrated the value required to justify its cost as a delivery structure—and that states now have both the policy tools and the technological capacity to reconsider it.

II. Managed Care Was Built for a Different System

Managed care organizations are paid through capitated rates structured to cover medical services, administrative operations, care coordination, reserves, and margin. Federal rules require plans to maintain a minimum 85% medical loss ratio (MLR)—the share of premium revenue that must go toward actual care delivery—meaning that by design, a portion of every Medicaid dollar paid to an MCO is allocated outside direct care. In practice, reported loss ratios in Medicaid managed care are typically in the range of 90 to 91 percent—leaving the balance to fund administrative functions, utilization management, plan infrastructure, and profit.

The original rationale for this structure was straightforward. When states first moved Medicaid populations into managed care at scale, they lacked the administrative capacity to process large volumes of claims efficiently, aggregate and analyze data across a fragmented provider landscape, and coordinate care for complex populations. Plans performed functions that state agencies genuinely could not.

That premise has materially weakened. Modern state systems and vendors can now automate claims adjudication, verify eligibility in real time, integrate data across provider types, deploy targeted care management tools for high-risk populations, and monitor utilization and quality directly. The question is no longer whether a state can perform these functions. It is whether paying a full-risk insurer to perform them—at plan overhead rates—remains the most efficient use of constrained Medicaid dollars.

States no longer need to buy 1990s-era insurer administration to perform 2026-era data, claims, and coordination functions.

III. The Nursing Facility Market Is Not Retail Health Care

The case for managed care is strongest in markets where plans can meaningfully influence consumer choice, steer utilization, and negotiate prices. Those conditions are largely absent in the nursing facility market.

Nursing facility care is characterized by state-set reimbursement methodologies, limited bed supply in many jurisdictions, provider-specific placement decisions driven by clinical need and geography, and long-stay populations with stable and predictable care requirements. Once a resident is admitted, care is not shopped. Prices are not negotiated in real time. Utilization is not meaningfully redirectable by plan design.

The state already defines the core economics of the nursing facility market through licensure, bed policy, case-mix classification, quality add-ons, supplemental payment programs, and rate methodology. In that context, the MCO is rarely functioning as a true risk manager. It is functioning as an administrative layer on top of a system the state has already largely structured and priced.

That distinction matters. Where managed care's value lies in its ability to manage risk and steer behavior, its value is limited in a setting where risk is relatively stable and behavior is highly constrained by regulation and clinical reality. The administrative functions that remain—claims routing, prior authorization, encounter data collection, care coordination—are exactly the functions that modern infrastructure can perform without routing dollars through a full-risk plan.

IV. The Evidence Does Not Compel Deference

The policy case for managed care in long-term services and supports is not settled. Federal advisory bodies have found no definitive conclusion that managed care improves access or quality in this population. Multi-state evaluations of managed LTSS programs show inconsistent results that vary by program design, population, and state context—not a consistent pattern of improvement. Studies focused specifically on the nursing facility setting show limited measurable impact on quality indicators or utilization patterns.

This does not establish that managed care is ineffective. It does establish that its value is not sufficiently demonstrated to justify ongoing immunity from scrutiny—particularly under fiscal pressure.

If a state is considering reductions to care, benefits, or provider payments, it is obligated to first examine every other component of system cost. Overhead spending that has not proven its value relative to alternatives does not get a pass simply because it has become the default.

V. The Middle Layer Represents Identifiable Cost

The financial structure of managed care in Medicaid is not opaque. Capitation rates are actuarially derived to fund care delivery plus plan overhead. The MLR framework makes clear that a defined category of spending sits between state appropriations and provider payments. That category is not hypothetical. It is structural.

Not every dollar in that layer is recoverable. Some administrative and coordination functions would need to be performed by the state or contracted vendors regardless of whether a risk-bearing MCO is in the picture. Eliminating plan overhead does not eliminate the need for claims processing, eligibility systems, or care transitions support. It changes who performs those functions and at what cost.

The relevant question is not whether states can capture every dollar currently allocated to plan overhead. The relevant question is whether states are paying a full-risk insurer premium for functions that could be performed more efficiently through direct administration, targeted automation, or specialized vendor contracts—and whether that premium is justified by outcomes the literature cannot consistently demonstrate.

Before any state reduces services or provider payments, it should identify every dollar in the system that is not flowing directly to care—and justify each one.

In the nursing facility market, where the state already controls the regulatory and payment structure, that examination is long overdue. The middle layer should not be the last thing reviewed when budgets tighten. It should be the first.

VI. Indiana: A State Draws a Line

Indiana's recent experience in this space is instructive. In 2024, the state launched PathWays for Aging, a managed LTSS program that expressly included nursing facility services, home health, and hospice. The program reflected the conventional approach: route the population through MCOs and leverage plan infrastructure to manage care and cost.

What happened next is more interesting than the launch. Governor Braun signed legislation in March 2026 providing that individuals who have received nursing facility services for more than 100 consecutive days will transition out of managed care and into fee-for-service Medicaid, with transitions beginning July 1, 2027.

Indiana did not declare managed care a failure. It drew a line. That line reflects a substantive policy judgment: whatever value managed care may provide in transitions, short-stay coordination, or early intervention, its case is weaker for long-stay institutional residents whose care needs are established, whose providers are fixed, and whose trajectories are not meaningfully redirectable by plan activity.

Indiana's retreat is not an isolated event. It is an early signal that states with direct experience managing this population through MCOs are beginning to question whether the structure is well matched to the setting. Other states should examine their own programs through the same lens before fiscal pressure forces the question.

VII. A More Direct Model Is Now Feasible

States considering reform are not choosing between managed care and chaos. A third model is available, and it is increasingly viable at scale.

Under a direct administration model, the state pays providers through a regulated fee schedule or value-based payment structure, retains control over quality incentives and program design, and procures specific administrative functions—claims processing, eligibility verification, care coordination, data analytics—through targeted vendor contracts rather than full-risk plan overhead. The functions are unbundled. Each is sourced and priced on its own merits.

This approach is not a return to paper-based state administration. It is the application of modern infrastructure to a population whose characteristics make direct administration considerably more tractable than in acute care. Long-stay nursing facility residents are a defined, identifiable population. Their providers are licensed, regulated, and largely known to the state. Their service needs are documented and relatively stable. The information environment for managing this population directly is more favorable than managed care's original architects assumed.

Care coordination for this population does not require a full-risk insurer. It requires competent discharge planning, reliable transition support from hospital to nursing facility and from nursing facility to community, and direct quality incentives tied to outcomes the state actually wants to drive. All of those functions can be designed and procured without paying plan overhead on top of them.

VIII. Policy Options for States

Reform does not require a single, all-or-nothing decision. States can begin with targeted changes and build from there.

Option 1: Long-Stay Carve-Out Transition residents who have been in a nursing facility for a defined period—typically 90 to 120 days—out of managed care and into fee-for-service or a state-administered payment model. Indiana's 100-day threshold is a reasonable reference point. This approach captures the population for whom managed care's value proposition is weakest while preserving MCO involvement in short-stay and transitional care.

Option 2: Mandatory Value Demonstration Require MCOs serving nursing facility populations to demonstrate measurable improvement in cost, quality, or access specific to that population—not to their Medicaid book of business broadly. Plans that cannot meet the demonstration requirement within a defined period are subject to carve-out. This shifts the default: the burden falls on the plan to prove value, not on the state to prove the absence of it.

Option 3: Administrative Cost Transparency Mandate detailed public reporting of plan-level administrative expenses, utilization management costs, and care coordination spending attributable to nursing facility populations. Transparency does not by itself produce reform, but it creates the informational foundation for it. States cannot evaluate whether those costs are justified without knowing what they are.

Option 4: Replace Plan Overhead with State-Administered Automation States do not need to redesign their entire Medicaid program to begin reducing plan overhead. Claims processing, prior authorization, and encounter data collection are high-volume, rule-driven functions that modern state systems and vendors can perform directly—without routing dollars through a full-risk plan. Automating these functions within the Medicaid agency recovers overhead cost incrementally, builds internal capacity, and positions the state for deeper reform over time.

Option 5: Full Carve-Out Remove nursing facility services entirely from managed care and return to state-administered payment with direct quality and performance incentives. This is the most structurally significant option and the most appropriate for states where the evidence of managed care value in this population is weakest and state administrative capacity is sufficient to support direct operation.

IX. Conclusion

The structure of Medicaid matters. When dollars flow through multiple administrative layers, each layer must justify its existence—not by default, not by inertia, but by evidence of value that no alternative arrangement can match.

In the long-stay nursing facility market, that justification has not been made. The evidence base is inconclusive. The market characteristics that make managed care most valuable are largely absent. The administrative rationale has weakened as state technological capacity has grown. And the fiscal environment no longer allows states the luxury of treating system complexity as a fixed cost.

Indiana has already drawn a line. Other states should draw theirs.

Managed care may remain appropriate elsewhere in Medicaid. But in the long-stay nursing facility market, the burden of proof has shifted. States should not reduce benefits, cut provider rates, or defer quality investments while paying a full-risk insurer premium they cannot demonstrate is earning its cost.

Preserve dollars for care. Reduce system overhead. Remove costs that do not produce measurable value.

The tools to do this exist. The policy models are available. The fiscal pressure is real and growing. States that examine this question now will be better positioned than those that wait until the question is forced.