Medicare Advantage Plan Exits Force 3 Million to Switch in 2026

From monitoring CMS plan finder data and enrollment snapshots across three consecutive open enrollment periods, the 2026 exit wave follows a pattern visible in plan-level bid data as early as mid-2025. The bids told the story before the enrollment numbers confirmed it: carriers priced in margin compression they could no longer absorb, and where the math stopped working, they walked.

A research letter published in JAMA on February 18, 2026 by Johns Hopkins Bloomberg School of Public Health researchers Mark Meiselbach, Matthew Lavallee, Jianhui Xu, and Dan Polsky quantified the fallout. Approximately 2.9 million Medicare Advantage enrollees in non-employer HMO and PPO plans, roughly 10% of that population, were forced to find new coverage for 2026 after their plan exited their county. That rate had averaged just over 1% annually between 2018 and 2024 before climbing to 6.9% in 2025 and hitting 10% for 2026. A tenfold increase in two years.

CMS projects total MA enrollment at 34 million for 2026, down from 34.9 million in 2025. For the first time in nearly two decades, the share of Medicare beneficiaries in MA is expected to decline, dropping from 50% to approximately 48%. The era of uninterrupted MA growth appears to have ended.

The Disenrollment Spike: What the JAMA Data Shows

The Johns Hopkins research team analyzed CMS Plan Finder data to track plan-level exits across all U.S. counties. Their findings reveal a structural acceleration in market instability, not a one-year blip. Annual forced disenrollment rates held below 2% for the six years between 2018 and 2024, averaging roughly 1%. The jump to 6.9% in 2025 signaled that carriers were beginning to rationalize their county footprints. The further escalation to 10% in 2026 confirmed it as a trend.

The study identified several plan characteristics associated with higher disenrollment risk. Enrollees in PPO plans faced significantly higher forced disenrollment rates than those in HMO plans, reflecting the broader margin pressure on PPO product structures where out-of-network utilization makes cost management harder. Non-special-needs plans were more likely to be terminated than Dual-Eligible Special Needs Plans (D-SNPs) or other SNP categories. Plans offered by smaller carriers and those rated below four stars were also disproportionately likely to exit.

These patterns make actuarial sense. PPOs carry higher medical loss ratios because beneficiaries can access out-of-network providers at partially covered rates. Smaller carriers lack the scale to negotiate competitive provider rates or absorb utilization volatility. And plans below four stars lose access to quality bonus payments worth 5% of their benchmark, creating a revenue gap that compounds the margin pressure from rising medical costs.

Which Carriers Are Pulling Back

The three largest MA carriers all reduced their geographic footprints for 2026, following through on signals embedded in their 2025 investor communications about trimming underperforming geographies.

UnitedHealthcare is offering plans in one fewer state and 109 fewer counties, with exits affecting approximately 180,000 enrollees. Despite the pullback, UHC remains the largest MA carrier by enrollment and retains the broadest county footprint in the market.

Humana made the most aggressive reductions, exiting 194 counties and three entire states. Humana plans will be available in 85% of U.S. counties in 2026, down from 89%, and in 46 states versus 48 in 2025. The pullback is particularly notable given Humana's concentration in MA, where the program generates the majority of its revenue. Humana also reported that one of its largest contracts, covering roughly 45% of its MA members, dropped from 4.5 to 3.5 stars, stripping quality bonus eligibility from a massive enrollment base almost overnight.

Aetna (CVS Health) will offer plans in 43 states and 2,159 counties, down from 44 states and 2,259 counties. Aetna is discontinuing approximately 90 MA plans across 34 states for 2026, with the majority being PPO products, consistent with the JAMA finding that PPO exits are driving much of the forced disenrollment.

Beyond the large national carriers, smaller plans are exiting entirely. Clear Spring Health announced it will close its Medicare Advantage business effective June 1, 2026, discontinuing all remaining plans across Illinois, Georgia, and Colorado. These smaller-carrier exits are significant because they disproportionately affect rural counties where plan choice was already limited.

The Provider Side: Health Systems Walking Away

Carrier exits are only half the disruption. At least 21 health systems have dropped one or more MA plan contracts for 2026, up from approximately 40 systems that took similar action in 2025. The provider-side exodus is driven by frustrations with prior authorization denial rates and slow reimbursement from MA carriers.

Several of the highest-profile health systems in the country have reduced their MA network participation:

• Mayo Clinic will be out-of-network with most MA plans from UnitedHealthcare and Humana in 2026
• Lehigh Valley Health Network went out-of-network with UnitedHealthcare MA effective January 25, 2026
• St. Luke's Health System no longer accepts Humana MA
• Kettering Health dropped Humana and Devoted Health MA contracts for 2026

When a major health system drops an MA plan, enrollees in that plan who rely on that system for specialty care face a choice between switching plans (if an alternative covers their providers) or paying out-of-network rates. In practice, many of these beneficiaries switch during open enrollment, adding to the disruption metrics captured in the JAMA study. For plan actuaries, provider network disruption also increases the uncertainty around utilization assumptions, since members who lose access to established provider relationships may delay care or shift to emergency settings with higher per-episode costs.

Financial and Actuarial Drivers Behind the Exit Wave

The exit wave did not happen in isolation. It is the downstream consequence of several simultaneous pressures on MA plan economics that have been building since 2023.

Medical cost acceleration outpaced federal payment growth

Medicare spending growth remained elevated from 2023 through 2025, driven by post-pandemic utilization recovery, pharmacy cost acceleration (particularly GLP-1 drugs), and labor cost pressure in provider networks. MA plan benchmarks, set as a percentage of fee-for-service Medicare costs, did not keep pace. The CMS 2026 advance notice implied effective per-enrollee payment growth below the rate of medical trend that most plan actuaries were projecting, creating a structural margin squeeze that hit hardest in counties where provider reimbursement rates were already near or above the benchmark.

Risk adjustment revenue compressed under V28

The CMS-HCC risk adjustment model transitioned fully to Version 28 for payment year 2026, completing a three-year phase-in (33% V28 in PY 2024, 67% in PY 2025, 100% in PY 2026). V28 reduced the number of valid diagnostic codes from 9,797 under V24 to 7,770, eliminating diagnoses that CMS determined were weak predictors of cost. The net effect: CMS projected average MA risk scores would decline by 3.12%, representing approximately $11 billion in reduced payments to plans. For carriers with large enrollment bases of chronically ill members whose conditions mapped to eliminated or reweighted HCCs, the revenue hit was concentrated and material.

Overpayment scrutiny intensified

MedPAC's March 2026 report to Congress estimated that MA overpayments totaled $76 billion for 2026, representing 14% more than what spending would have been if the same beneficiaries were enrolled in traditional Medicare. The overpayment stems from three sources: coding intensity (diagnostic upcoding that inflates risk scores), favorable selection (MA enrollees tend to be healthier than their risk scores predict), and quality bonus payments. While $76 billion is actually lower than prior-year estimates due to V28's partial correction, MedPAC's continued spotlight on the payment gap increases the probability of further legislative or administrative action to tighten MA reimbursement.

Star Ratings: When Quality Bonuses Disappear Overnight

The MA Star Ratings system assigns quality scores from 1 to 5 stars based on clinical outcomes, member experience, and administrative performance. Plans rated four stars or above receive a 5% quality bonus payment added to their county benchmark. In 2026, the average MA contract rating stands at 3.98 stars, just barely below the four-star threshold. Only about 40% of contracts earned four stars or above, meaning the majority of enrollment sits in contracts that do not qualify for the full bonus.

Star Ratings volatility creates acute financial planning challenges for plan actuaries. The single most dramatic example in 2026 is Humana, where one contract covering approximately 45% of its MA members dropped from 4.5 to 3.5 stars between the 2025 and 2026 rating years. That one-point decline stripped quality bonus eligibility from millions of enrollees, immediately reducing the revenue Humana can use to fund supplemental benefits or absorb medical cost overruns.

CMS finalized a major overhaul of the Star Ratings methodology in the CY 2027 final rule, removing 11 measures and scrapping the Health Equity Index. The Congressional Budget Office estimated those changes will redirect $18.56 billion to MA insurers over the next decade by making it easier for plans to achieve four-star status. But that relief arrives in 2027 at the earliest, too late to affect the 2026 plan year that drove this exit wave.

The Benefit Squeeze: Measuring the Decline in Plan Value

Milliman's Medicare Advantage Competitive Value-Added Tool (MACVAT) provides the most granular actuarial view of how benefit richness has changed year over year. Their 2026 general enrollment analysis documents a greater than 7% decline in total value added across the MA market from 2025 to 2026, the largest annual drop in recent history.

The components of that decline tell a specific story about where plan actuaries chose to cut:

• Part C benefit value decreased approximately $17 per member per month (PMPM), reflecting higher cost-sharing imposed on members for medical services
• Part D benefit value decreased approximately $4 PMPM, driven by the Part D redesign's shift of liability to plans
• Average medical maximum out-of-pocket (MOOP) rose from $5,100 to $5,440, a 7% increase
• Medical deductibles jumped from $33 to $57, a 72% increase
• Part D deductibles increased from approximately $230 to $375, with 83% of members now facing Part D deductibles compared to just 23% in 2024
• Zero-premium MA-PD plans declined by 231 plans (a 9.5% decrease), shrinking the most visible MA value proposition

KFF's 2026 spotlight on premiums and benefits confirms the supplemental benefit erosion. Over-the-counter allowances fell from 73% of plans in 2025 to 66% in 2026. Meal benefits dropped from 65% to 57%. Transportation services declined from 30% to 24%. Remote access technologies fell from 53% to 48%. Only vision, dental, and hearing coverage remained near-universal at 98% or above.

Milliman also observed that carriers with declining value added experienced enrollment losses of 4% to 33%, while one carrier that improved its value-added position saw 22% enrollment growth. The market is punishing benefit cuts in real time.

Risk Pool Implications: The Adverse Selection Spiral

The enrollment decline creates a feedback loop that plan actuaries need to model carefully. When carriers exit counties, the beneficiaries who lose coverage must choose between enrolling in a different MA plan (if one is available), switching to traditional Medicare with a Medigap supplement, or enrolling in traditional Medicare without supplemental coverage.

The selection dynamics of these transitions are not symmetric. Healthier beneficiaries, who have lower switching costs because they are less dependent on specific provider relationships, are more likely to move to traditional Medicare if MA benefits have eroded. Sicker beneficiaries, who depend on MA plan features like care coordination, supplemental benefits, and out-of-pocket caps, are more likely to stay in whatever MA plan remains available. This differential sorting worsens the risk profile of the remaining MA population.

The risk pool concern operates on two levels simultaneously:

Within MA plans: As healthier members exit, the remaining enrolled population has higher average acuity, pushing medical loss ratios upward and creating further pressure for benefit cuts or additional market exits in subsequent years. This is the classic adverse selection spiral described in actuarial literature.

Between MA and traditional Medicare: If the members returning to fee-for-service are healthier than average, they increase the relative cost of the remaining MA population compared to the FFS benchmark used to set MA payments. Since MA benchmarks are tied to FFS spending, a compositional shift in the FFS population toward healthier enrollees could actually lower the benchmark, further squeezing MA plan revenue. Conversely, the same shift would increase per-capita FFS spending if the returning MA members are lower-cost, which is more consistent with the favorable selection patterns MedPAC has documented.

The direction of this selection effect is an open empirical question. But for plan actuaries building 2027 bids, the uncertainty itself is the problem. The distribution of risk scores in a plan's enrolled population is shifting at a pace that makes recent experience data a less reliable predictor of next-year costs.

The CMS 2027 Rate Notice: Relief or Reprieve?

The CMS 2027 MA final rule delivered a 2.48% effective growth rate, a dramatic improvement from the 0.09% proposed in the advance notice. That 239-basis-point swing from the NPRM to the final rule reflected revised fee-for-service normalization, changes to the V28 risk score trend, and the Star Ratings methodology overhaul. For plan actuaries filing June 2026 bids, the 2.48% rate provides more room to maintain benefits or slow the pace of benefit reductions.

Whether 2.48% is enough to reverse the exit trend depends on whether it covers the gap between payment growth and medical cost trend. Most industry estimates place 2026-2027 medical trend in the 6-8% range for MA populations, driven by utilization recovery, pharmacy costs, and labor inflation in provider networks. A 2.48% payment increase covers roughly one-third to one-half of that trend gap, meaning plans still need to find margin improvement through benefit design, network management, or membership growth.

The rate reversal analysis we published earlier this year walked through the component mechanics of how the 2.48% number was constructed. The key insight for plan actuaries evaluating 2027 county-level viability is that the rate increase is an average; geographic variation in FFS spending trends means some counties will see above-average effective rate increases while others will see below-average, potentially perpetuating the exit incentives in the lowest-benchmark counties.

What Plan Actuaries Should Be Modeling Now

The Long View: Is the MA Growth Model Broken?

Medicare Advantage grew from 13% of Medicare enrollment in 2005 to 50% in 2025 on the back of a simple economic proposition: CMS benchmark payments exceeded the cost of covering the enrolled population, leaving a surplus that carriers could deploy as supplemental benefits (dental, vision, hearing, gym memberships, OTC allowances) to attract more enrollees. The flywheel worked because favorable selection meant MA enrollees cost less than their risk scores predicted, and coding intensity meant risk scores were systematically higher than they should have been. MedPAC estimates these dynamics generated $76 billion in excess payments in 2026 alone.

V28 is designed to close part of that gap. The Star Ratings overhaul will redistribute some of the pressure. The 2.48% rate for 2027 provides a bridge. But the fundamental question is whether MA can sustain 50%+ market share at payment rates that are closer to the true cost of covering its enrolled population. The 2026 enrollment decline to 48% suggests the answer is not yet clear.

Patterns we have observed across three open enrollment cycles point to a market that is stratifying rather than collapsing. Well-capitalized carriers with strong Star Ratings, efficient provider networks, and scale-driven administrative cost advantages can still operate profitably under tighter economics. Smaller carriers, plans with below-four-star ratings, and PPO products in high-cost geographies face structural headwinds that the 2027 rate increase alone will not resolve.

For the actuarial profession, the MA enrollment decline resets a question that had seemed settled: how large a share of Medicare can the private market sustain, and at what cost to the federal budget? The next two enrollment cycles will determine whether 2026 was a cyclical low point or the beginning of a longer-term market correction.