Medicare Advantage (MA) plans have proven extremely successful in the market, increasing enrollment from 11.3 million in 2010 to 24.4 million in 2020.1 Part of their success is that MA plans often provide benefits in addition to what traditional Medicare offers (for example, hearing, dental, and prescription drug coverage) and, unlike Medicare Fee-For-Service (FFS), have maximum out-of-pocket limitations.2 Often, those additional benefits come at little to no cost to the beneficiary.
MA plans are designed to increase care management and control spending compared to traditional Medicare FFS. MA plans have incentives to improve care management because they receive capitated payments for each enrollee that depends on a patient’s risk score. That is, MA plans do not receive additional payments from the government for beneficiaries that require spending that is higher than anticipated (for example, because of higher or more costly healthcare utilization). However, MA plans have been criticized for so called upcoding, which inflates patients’ risk scores and makes them appear less healthy than they are, causing overpayments and an increase in revenues to MA providers.3
As a result, MA plans may reduce costs relative to Medicare FFS if better care management and spending controls dominate, or MA plans may increase costs because of higher administrative spending related to its more expensive care management and the upcoding. The Medicare Payment Advisory Commission (MedPAC) reports that in 2021, Medicare paid roughly 4 percent more for a patient in a MA plan than for a comparable patient in Medicare FFS, down from almost 18 percent in 2009.4
Medicare Advantage plans are private health plans that contract with Medicare and receive capitated payments to provide all Medicare-covered services to plan enrollees. The payment is determined through the benchmark-and-bidding system. The benchmark is a set percentage of projected average spending for Traditional Medicare enrollees in a county (percentages range from 95 to 115 percent depending on the FFS spending level of the county). Benchmarks are higher (by 5 or 10 percent) for plans that have star rating higher than 4. The bids submitted by each contract plan will be compared with the benchmarks, plans with bids higher than the benchmarks will receive the benchmark amount, plans with bids lower than the benchmarks will still receive a portion of the difference between the bid and benchmark as a “rebate” (the portion is 65 percent for plans with 4 stars, and 70 percent for plans with 5 stars). The per-enrollee payment (not including rebate) from Medicare to the private plans is also adjusted based on the risk score of the enrollee that reflects the enrollee’s predicted health cost.
Because the MA program is administered at the county level, not all 3,142 counties in the United States have MA plans available; In 2021, 56 counties, or about 2 percent of all counties with a population of about 0.8 million people, did not have any MA plan available. Among the 3,086 counties with at least one MA plan, only 1,270 had a MA plan with at least a 5-star quality rating, and 2,981 counties had a MA plan with at least a 4-star rating.
Under our illustrative policy, which would start in 2023, those aging into Medicare (i.e., those who turn 65 years old and qualify for Medicare FFS) and fail to enroll in a Medicare plan (either Medicare Plan B or D, or a MA plan), will be automatically enrolled in a high-quality, $0 beneficiary premium, MA plan. Individuals do not qualify for auto enrollment if they:
are dual eligible for both Medicare and Medicaid benefits;
receive long-term care services or have End-Stage Renal Disease (ESRD);
are covered by employer or group health plans; or
live in a service area where no applicable MA plan is offered.
Individuals who were automatically enrolled in a MA plan can disenroll at any time during the first year of auto enrollment. Eligible plans for default auto enrollment must:
have a $0 monthly beneficiary premium; and
have at least a 4-star quality rating by the Centers for Medicare & Medicaid Services (CMS).
If there is more than one qualified MA plan available in a county, individuals would be randomly assigned to the qualifying plans.
| 2022 | 2023 | 2024 | 2025 | 2026 | 2027 | 2028 | 2029 | 2030 | 2031 | 2032 | 2022-2032 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Scenario 1 | $0 | $6 | $8 | $11 | $14 | $17 | $21 | $24 | $27 | $30 | $32 | $189 |
| Scenario 2 | $0 | $7 | $11 | $15 | $19 | $24 | $29 | $34 | $39 | $44 | $46 | $269 |
Notes: Scenario 1 includes only the population that would not have enrolled in Medicare otherwise. Scenario 2 includes only the population that would not have enrolled in MA plans.
| 2022 | 2023 | 2024 | 2025 | 2026 | 2027 | 2028 | 2029 | 2030 | 2031 | 2032 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Scenario 1 | 0 | 0.4 | 0.5 | 0.7 | 0.8 | 1 | 1.1 | 1.2 | 1.3 | 1.4 | 1.4 |
| Scenario 2 | 0 | 2.2 | 3 | 3.8 | 4.5 | 5.1 | 5.8 | 6.4 | 6.8 | 7.1 | 6.8 |
Notes: Scenario 1 includes only the population that would not have enrolled in Medicare otherwise. Scenario 2 includes only the population that would not have enrolled in MA plans.
Under current law, 2,890 counties (92 percent of all counties) have an auto-enrollment eligible MA plan available.5 In total, there are 25,760 such plans available in the country. The analyzed policies would increase the population that is enrolled in MA plans, and, because of the increase in the potential market size, those policies also increase the number of counties with eligible MA plans.
Scenario 1: Taking current MA plan availability as fixed and assuming that only people who do not enroll in any Medicare plan are auto enrolled, PWBM projects that outlays increase by $189 billion over the 2022-2032 period (see Table 1). In that scenario, an additional 1.4 million people would be covered by a MA plan in 2032 (see Table 2). PWBM also projects that ten more counties will have at least one eligible MA plan if the proposed policy is enacted, thereby increasing MA plan availability for roughly 0.03 million people by 2032.
Scenario 2: Of course, if people are automatically enrolled in a high quality, $0 beneficiary premium, MA plan, the incentive to enroll in Medicare part B or D decreases. If we assume that those who age into Medicare and do not enroll into another MA plan are automatically enrolled in a MA plan, an additional 6.8 million people would be enrolled in a qualifying MA plan by 2032 (see Table 2). Outlays would increase by $269 billion over the 2022-2032 period (see Table 1).
PWBM also projects 50 more counties will have at least one eligible MA plan if the proposed policy is enacted, thereby increasing MA plan availability for as many as 0.13 million people in 2032. Including those additional MA plans, outlays under the first scenario would be $193 billion over 2022-2032. Under the second scenario, outlays would be $276 billion over the same period.
The average cost per person to the federal government of providing MA coverage would be $14,500 in 2023 under the first illustrative policy, while the average cost is $3,300 in 2023 under the second policy. The cost in the first scenario is higher, because those people are currently not enrolled in any Medicare plan and each additional enrollee would require the full per person cost of providing MA coverage. In the second scenario, the group of people who would be newly enrolled in MA plans is a mix between those who would not have been enrolled in any Medicare plan and cost $14,500 per person in 2023, and those who would have otherwise been enrolled in an FFS plan and would cost an additional $900 per person. The second group is less expensive to enroll in MA plans, because they already incur the cost of being enrolled in an FFS plan.
Two potential options that might lower total outlays are 1) auto enrolling people into the lowest cost MA plan instead of randomly enrolling them into an eligible plan; and 2) auto enrolling people into lower quality rated MA plans. Enrolling people into the lowest cost MA plan could decrease outlays over the 2023-2032 period by between 7 percent in our first scenario and by about 26 percent in our second scenario. However, enrolling people into lower quality MA plans does not decrease spending. The reason is that the average cost of MA plans with star ratings greater than three is almost identical to the average cost of plans with a three star-rating and lower. In addition, over 2,300 counties (or 74 percent of all counties) do not have any MA plans with star ratings of three or less.
PWBM’s budget estimate is based on county-level enrollee and cost data. Because the proposed policy may dis-incentivize people from signing up for traditional Medicare (FFS), and it may encourage insurance companies to provide MA plans in counties where they are not providing services now, we performed a three-step analysis: First, we assume that those who age into Medicare do not change their enrollment behavior; i.e. the same fraction of the elderly population will enroll in MA and Medicare FFS as we project under the baseline. Second, we assumed that under the new policy those turning 65 have little to no incentive to enroll in traditional Medicare anymore and will all be auto enrolled in an MA plan. And third, we assume that health insurance companies will enter additional counties in response to the increase in potential MA customers.
Under our first assumption, we estimate the eligible population based on data from the PWBM Microsimulation model and the county population distribution from the American Community Survey (ACS). In each year, the eligible population include people who turn 65. Based on enrollment choices made in 2019, we take 7 percent of the eligible population and auto enroll them into an eligible MA plan; the remaining 93 percent enroll in an MA or Medicare FFS plan.6 Next we exclude people who are having ESRD based on the state level percentage of people with ESRD and those who have employer plans.7 At the end of the first year, we assume that 30 percent of those who were auto enrolled choose to switch to a Medicare FFS plan. We then add those enrollees to the group of people who were automatically enrolled and remained in an MA plan in the prior year. Of that group, 2 percent disenroll every year and switch back to Medicare FFS.8
Under our second assumption we add to the 7 percent of the eligible population that would not have enrolled in any MA or FFS Medicare plan all those who would have enrolled in a FFS plan. Based on the MA penetration rate projected by 2020 Medicare Trustees Report, along with the county level population, we estimate the population size that would sign up for FFS.9 Among the additional eligible population, we exclude the beneficiaries who are dually eligible based on the state level percentage of dual eligible population and those have employer plans as supplemental coverage.10
Under our third assumption, we estimate how many additional counties will have MA plans available. We use county-level MA plan data from CMS to calculate the number of counties that currently have eligible MA plans.11 The MA plan data contains plan premiums, star ratings, risk scores, plan types, and the cost to the federal government (including rebates) for each contract plan. We then estimated MA plan availability using a logit regression of the number of eligible MA plans on the Medicare eligible population and Medicare enrolled population in a county. Based on that estimate we calculate the additional MA plans available under the policy using an estimate of the newly eligible population in each county under the policy.12
Next, we calculate the per capita cost for those enrolled in MA plans as the average plan cost among eligible MA plans for each county.13 For counties that currently do not have an eligible MA plan, we assume that the per capita cost is the state average. We adjust the per capita cost for future fiscal years based on projections of excess cost growth from the Congressional Budget Office.14 We calculate Medicare FFS spending by county based on data from CMS and adjust is so that the 2021 cost difference between MA and FFS is 4 percent as report by MedPAC.15
This analysis was written by Junlei Chen and Yan He under the direction of Richard Prisinzano and Felix Reichling. Prepared for the website by Mariko Paulson.
-
https://tinyurl.com/yv49j6cu, Figure 12-2. ↩
-
Kelsey Waddill, “Medicare Advantage Quality of Care Surpasses Traditional Medicare”, HealthPayer Intelligence, December 10, 2020, https://tinyurl.com/by785jzp; Centers for Medicare & Medicaid Services, “Understanding Medicare Advantage Plans”, CMS Product No. 12026, November 2020, https://tinyurl.com/chw364xy; Better Medicare Alliance, “Medicare Advantage Is a High-Value Option for Consumers”, May 2018, https://tinyurl.com/3zj497m2. ↩
-
https://www.cbpp.org/blog/medicare-advantage-upcoding-overpayments-require-attention ↩
-
https://tinyurl.com/yv49j6cu, Figure 12-4. ↩
-
There are 91 counties where 4-star plans are available, but their beneficiary premiums are greater than $0. ↩
-
Based on the population and number of total Medicare enrollees of age 65 to 74 in 2019 (either with FFS or MA), the Medicare enrollment rate among people ages 65-74 is 93 percent. See United States Census Bureau, “2019 Population Estimates by Age, Sex, Race and Hispanic Origin“, June 2020, https://tinyurl.com/ypsdrjvz; Centers for Medicare & Medicaid Services, “Medicare enrollment,” 2019, https://tinyurl.com/2p9bk6un; Jenny Yang, “Distribution of Medicare enrollees in 2019, by age group”, September 2021, https://tinyurl.com/mr4c6zja. Note that this estimate automatically excludes Medicare enrollees who are dually eligible and who have employer plans as supplemental coverage. ↩
-
Kaiser Family Foundation, “Medicare Beneficiaries With End-Stage-Renal Disease (ESRD)”, 2019, https://tinyurl.com/2p85r423; Kaiser Family Foundation, “A Snapshot of Sources of Coverage Among Medicare Beneficiaries in 2018,” March 2021, https://tinyurl.com/2p8asv2j. We assume that the ESRD rate and Employer Plan rate among elderly without Medicare coverage are the same as those for Medicare beneficiaries. ↩
-
Steven C. Martino et al., “Rates of Disenrollment From Medicare Advantage Plans Are Higher for Racial/Ethnic Minority Beneficiaries,” Medical Care 59, no. 9 (September 2021): 778-784, doi: 10.1097/MLR.0000000000001574. ↩
-
Medicare Trustees, “2020 Annual Report”, April 2020, Table IV.C1, https://tinyurl.com/ykdwmyjf. ↩
-
Kaiser Family Foundation, “Dual Eligibles as a Percent of Total Medicare Beneficiaries,” December 2021, https://tinyurl.com/5n7hu4ms. ↩
-
Centers for Medicare & Medicaid Services, “Part C and D Performance Data,” https://tinyurl.com/5n767mk5; Centers for Medicare & Medicaid Services, “Plan Payment Data,” https://tinyurl.com/4pd6h87x. ↩
-
We predict the market change based on county level potential enrollment. ↩
-
Using the average across eligible MA plans is equivalent to randomly selecting among eligible MA plans as stipulated under the policy. ↩
-
Congressional Budget Office, “The 2018 Long-Term Budget Outlook,” 2018, https://tinyurl.com/2p8km93b (accessed December 17, 2021). ↩
-
See Centers for Medicare & Medicaid Services, “Public Use File,” https://tinyurl.com/yzn3hbnp; and MedPAC, “For the record: MedPAC’s response to AHIP’s recent “Correcting the Record” blog post,” March 2021, https://tinyurl.com/7vp4cuxv. ↩