Medicare Advantage by Default

The Trump administration is considering a change to Medicare that could help make healthcare more efficient and cheaper for millions of American seniors and disabled people, but it comes with important caveats.

The proposal would make Medicare Advantage (MA) plans the default coverage option for Medicare enrollees, rather than traditional fee-for-service (FFS) Medicare. This move could affect a wide swath of Americans and the flow of hundreds of billions of dollars. Fifty-five percent of Medicare beneficiaries (35 million) have chosen MA, which includes 5,500 plan options offered by 164 organizations at a cost of $540 billion to the federal government. On average, a Medicare beneficiary can choose from 39 plans offered by eight organizations, although some local markets are more concentrated. MA enrollment rates are currently higher among those dually eligible for Medicaid (68 percent), black and Hispanic beneficiaries, those who self-assess as being in poor health, and those with low incomes.

One study estimated that if this default policy change were made, under certain assumptions, another seven million beneficiaries would enroll in MA, meaning roughly two-thirds of Medicare beneficiaries would be enrolled. Other research suggests the number could be even higher. This would tip the balance decidedly toward private plans and call into question the current policy structure of a program design based on FFS, suggesting instead that it should be restructured as a competitive premium-support system. This move will take time and require political capital to enact, but it should be the ultimate end-goal. Thinking of the interim, the Trump administration should ask itself, “Would changing MA to be the default coverage option increase federal costs, and would it cause harm to beneficiaries?”

The Anatomy of MA

The MA program allows those enrolled in both Parts A (hospital insurance) and B (physician insurance) to receive benefits from private plans—such as health maintenance organizations, preferred provider organizations, medical savings accounts, and other types—rather than Medicare FFS. MA plans use alternative payment models, negotiate with individual providers, employ care-management techniques like prior authorization, and provide incentives for beneficiaries to seek care from more efficient providers. These features should reduce costs, and in a competitive insurance market, those reductions will be passed on to beneficiaries. This can occur through lower premiums and cost-sharing liabilities, supplemental benefits not covered by Parts A and B (such as dental, hearing, and vision insurance), and integrated Part D (drug insurance) benefits.

In addition, MA plans are required by law to offer an out-of-pocket spending limit, a protection not included in FFS. Most FFS beneficiaries who are not covered by Medicaid (for the poor) or a prior-employer retiree health plan (which is common for unionized and government workers) will pay for a Medigap policy to fill in the FFS deductibles and co-pays. These financial disadvantages of FFS must be weighed against the freedom it offers from the constraints and limited provider networks of MA plans.

Because Medicare pays MA plans a fixed rate for each enrolled beneficiary, plans are incentivized to avoid high-cost beneficiaries or attract those with low expected costs. To counter this incentive, an elaborate payment mechanism combines a base rate and a risk score to adjust for differences in expected beneficiary medical costs. The base rate is determined in part by the FFS benchmark costs for the county where the beneficiary resides and the plan’s quality rating. If the plan bids below the benchmark, as nearly all do (at an average of 95 percent), the base rate paid is the bid plus a majority share of the difference between the benchmark and the bid. This “rebate” share increases with the quality rating.

Creating these benchmarks is its own complex calculation. Each county’s benchmark equals the projected average per capita FFS spending, with a bonus for low-spending counties and a small reduction for high-spending counties. The risk score is a beneficiary-level index value relative to the national average FFS beneficiary, calculated using a risk-adjustment model from the Centers for Medicare & Medicaid Services (CMS) that incorporates demographic information such as age, gender, Medicaid eligibility, institutional and disability status, and certain diagnoses grouped by type and severity according to similar treatment costs. CMS tracks demographic information while MA plans submit diagnostic information, which is incorporated into the risk score with a one-year lag. Diagnosis codes must be supported by evidence in the patient’s medical record and are subject to CMS audits, though such audits are infrequent.

Understanding Overpayment

But the risk-adjustment model is far from perfect. It is calibrated using FFS claims data, which—according to the Medicare Payment Advisory Commission (MedPAC)—produce two types of overpayments to MA plans.

First, MedPAC estimates that risk scores overpredict spending for MA enrollees by 11 percent, owing to what it calls “favorable selection.” This means that, even after controlling for demographics and diagnoses, MA beneficiaries inherently spend less than FFS beneficiaries.

Second, MedPAC estimates that MA plans have higher coding intensity than FFS does, owing to the obvious incentive for MA plans to code diagnoses rigorously, and even liberally, to increase their payments from Medicare. After an automatic statutory adjustment of 5.9 percent, the overpayment rate was 4 percent in 2026. After standardizing for differences in risk scores and the county of residence between MA and FFS beneficiaries, MedPAC estimated that payments to MA plans, which are composed of MA plans’ bids and rebates, were about 99 percent of FFS spending.

On net, MedPAC estimates the overpayment to MA plans is 14 percent, or $76 billion. This figure accounts for most of the advantage MA plans hold in terms of lower premiums and extra benefits rather than reflecting genuine healthcare efficiency, net of administrative costs and insurer profits. (The CMS Office of the Actuary believes the amount of overpayment is about 8 percent currently, as it noted in the 2026 Medicare Trustees Report.)

Taken at face value, MedPAC’s 14 percent overpayment estimate would imply that default enrollment into MA plans would significantly increase the cost of Medicare to taxpayers—which is not a desirable policy result. While others push back by claiming the coding intensity of MA plans gives providers better information about their patients, there are also several reasons to consider this estimate to be a significant overstatement.

First, the coding intensity estimate has already been reduced (from 10 percent to 4 percent) in prior years, owing to improvements in the CMS risk-adjustment model. In 2027, another change will further lower payments by excluding diagnosis information from unlinked chart review records not associated with specific beneficiary encounters from risk scores and by incorporating updated FFS diagnosis and spending data. In short, the coding intensity overpayment will likely shrink to zero in the near future, although CMS will need to continue its efforts in this area to stay current with evolving industry practices.

Second, there is strong reason to think that MedPAC’s favorable-selection estimate is too high. Although the calculation is quite involved, in essence it measures the health spending of FFS beneficiaries in the year before they switch to MA and compares that spending to similar non-switching FFS beneficiaries. MedPAC implicitly assumes that the expenditures of FFS beneficiaries before they switch are indicative of their underlying costs in MA, while the expenditures of those who remain in FFS serve as the appropriate counterfactual.

This approach has several problems. First, as noted above, FFS has lower coding intensity, meaning those with higher FFS spending have risk scores that are too low, biasing the comparison. This is a notable difference, accounting for about 2 percentage points of the favorable-selection estimate. There is also evidence that any initial favorable selection recedes over time—that is, there is a regression to the mean. Hospice care is provided only by FFS, so terminally ill patients with high costs are overrepresented in FFS spending data.

It is also possible that beneficiaries contemplating a switch to MA will postpone discretionary health spending until after they enroll in MA, when coverage is cheaper. MedPAC excludes beneficiaries who move to different counties from its calculation, which may introduce upward bias if geographically stable FFS beneficiaries are more costly, all else being equal. Moreover, it excludes those who enroll in MA and stay, and as the trend toward MA has grown over time, these enrollees may differ considerably from switchers. MedPAC further excludes from its estimate FFS beneficiaries who are enrolled in Part A only, though their costs are lower than Part A and B enrollees and they are included in the benchmark.

Finally, individual healthcare spending in any given year is somewhat random, even after controlling for demographics and health status. If those with temporarily low spending are more likely to switch to MA—perhaps because they feel they are not getting enough value from their costly Medigap policies or are missing supplemental benefits—and those with temporarily high spending are less likely to switch, then MedPAC’s estimate is biased upward.

In summary, MedPAC’s implicit static estimate of the cost of a default MA policy is overstated, likely by at least half. Moreover, if more beneficiaries are enrolled in MA through inattention—that is, by default—any favorable-selection effect will diminish. The remaining favorable-selection cost is open to different policy interpretations, particularly given that it is more likely to reflect beneficiary preferences for the level and style of healthcare. This is to say that high FFS comparison costs reflect that higher-income and wealthier FFS beneficiaries prefer more care and choice, rather than that MA enrollees have better health status.

Finding the Right Balance

Some who study these programs have expressed concern that default enrollment in MA plans would harm beneficiaries by disrupting continuity of care between patients and their doctors (owing to beneficiaries potentially having to change provider networks), access to prescribed drugs (due to different plan formularies), or timely access to services (as a result of prior authorization). These concerns are quite paternalistic, however, because beneficiaries who place high importance on these issues can select FFS or whichever MA plan best fits their needs. Default enrollment only offers patients a new starting point, but their overall options remain the same.

A paternalistic argument could also cut in the opposite direction: The care-coordination features of MA plans, especially for those with special needs, may produce superior health outcomes compared with the degree of self-direction that FFS requires. Critics also have more nuanced concerns, like whether those who wish to switch out of MA after the default enrollment would be granted a special enrollment period for Medigap without preexisting condition restrictions, particularly if they choose to switch beyond the current 12-month trial period.

This default MA policy could be implemented in various ways. The government could randomly assign beneficiaries to one of the plans in their area, one of the zero-dollar premium plans, the area’s lowest-premium plan, plans with quality ratings of four (out of five) or greater, or an area’s lowest-cost plan. Beneficiaries with retiree coverage from past employers could be excluded if their retiree health plan does not include MA plans in the benefit menu. The administration could also get this ball rolling without a fight in Congress, as default MA could be implemented without legislation through a testing model under the Center for Medicare and Medicaid Innovation or through CMS demonstration authority.

This change can also lay the groundwork for a new paradigm of competition with FFS, which could bring costs down to benefit the consumer. Rather than explicitly lowering rebates to MA plans on account of favorable selection, the implicit higher cost of FFS could be addressed through policies targeting FFS spending directly and eventually by introducing more competition between the two segments through a premium-support structure. FFS costs could be reduced through measures relating to prior authorization already being tested in a Center for Medicare and Medicaid Innovation model or by limiting the extent of Medigap and retiree plan coverage, both of which have been shown to significantly increase FFS costs.

More fundamentally, as MA becomes the predominant coverage model in Medicare, the current benchmarking and risk-adjusting rules make little sense. Instead, a premium-support approach would be a better policy. In premium support, the federal government gives each beneficiary a fixed payment or subsidy, which beneficiaries use to purchase coverage from competing private plans or traditional FFS Medicare. If someone chooses a plan that costs more than the government contribution, they pay the difference themselves; if they choose a cheaper plan, they may keep some savings or pay lower premiums.

This approach is considerably simpler than the current hybrid model and would yield several advantages. It would intensify competition among private insurers for beneficiaries’ business, putting downward pressure on costs and improving services. It would also control federal spending by allowing the government contribution to grow at a slower, predetermined rate rather than rising automatically with healthcare spending. A premium-support approach would encourage efficiency by giving plans stronger incentives to negotiate prices with providers, manage care, and reduce unnecessary spending. Lastly, it would expand consumer choice by allowing beneficiaries to select plans tailored to their needs and preferences.

Americans desire high quality care, choice, and lower costs in their healthcare. The Trump administration’s potential moves to make MA plans the default option are a good step toward those goals. Even with a small increase in the federal budget, it is a plan that brings no harm to beneficiaries and serves as an ideal precursor for more fundamental changes in Medicare and Medicaid and eventually healthcare more broadly.

Mark J. Warshawsky is a senior fellow and Wilson H. Taylor Chair in Health Care and Retirement Policy at the American Enterprise Institute, where he focuses on Social Security and retirement issues, pensions, long-term care, disability insurance, and the federal budget.