Medicare Spending was 27% More for People who Disenrolled from Medicare Advantage than for Similar People in Traditional Medicare
This analysis uses data from the 20% Medicare Beneficiary Summary File (MBSF) and Medicare claims files to compare Medicare spending (across Parts A and B) for beneficiaries who were previously enrolled in Medicare Advantage (Disenrolled from Medicare Advantage) to beneficiaries who have been continuously covered by traditional Medicare (Always Traditional Medicare). The Disenrolled from Medicare Advantage cohort includes people who were enrolled in Medicare Advantage in 2021 and in traditional Medicare in 2022. The Always Traditional Medicare cohort of beneficiaries were enrolled in traditional Medicare in both 2021 and 2022.
To be included in either cohort, beneficiaries had to have 12 months of coverage under Medicare Part A and Part B in 2021, as well as coverage in Parts A and B in all months in 2022 they were enrolled in Medicare (i.e., not deceased). The analysis includes all 50 states and DC. People with ESRD are excluded because 2021 was the first year people with ESRD were broadly eligible to enroll in a Medicare Advantage plan. For the Disenrolled from Medicare Advantage cohort, the analysis excludes people who were enrolled in a cost plan, Medicare Medicaid plan, or PACE plan in 2021.
Compared to the Always Traditional Medicare cohort, a somewhat larger share of the Disenrolled from Medicare Advantage cohort were ages 65 to 69 (24% vs. 17%), Black (10% vs. 7%) or Hispanic (7% vs. 5%), or dually eligible for Medicare and full Medicaid benefits (16% vs. 12%). There were also differences in the share of beneficiaries with certain health conditions (Appendix Table 1). To account for differences in age, dual eligibility, and health status, the risk score for each person was calculated using the CMS-HCC Risk Adjustment model based on diagnoses present in 2022. Note that the risk model is used to adjust payments to Medicare Advantage plans and does not account for race. Thus our approach also does not control for race and ethnicity.
Compared to all people enrolled in Medicare Advantage in 2021, those who disenrolled were more likely to have been covered by traditional Medicare in 2020 than those who stayed enrolled in Medicare Advantage. Specifically, among people who disenrolled from Medicare Advantage, 31% were covered by traditional Medicare in 2020; among people who stayed in Medicare Advantage, 11% were in traditional Medicare in 2020. These differences may be driven by the ability for people to return to the same Medigap policy within 12-months after switching to Medicare Advantage for the first time (note, Medicare beneficiaries are limited to a single “trial period”), relieving concerns about Medigap underwriting or cost sharing without a supplemental policy for some subset of people who disenrolled. We are unable to determine what share of the people who disenrolled from Medicare Advantage were in their “trial period” because it was the first time they had tried Medicare Advantage.
Our approach to analyzing spending follows the approach used by MedPAC in calculating average spending for people who switch from traditional Medicare to Medicare Advantage compared to those who remain in traditional Medicare. First, all Medicare payments, beneficiary liability, and other primary payer payments for all Part A and Part B services were summed for each person in the sample for Calendar Year 2022. Next, each person’s total spending is divided by their risk score (calculated using the CMS-HCC Risk Adjustment model and 2022 diagnoses) to derive risk-standardized spending, which allows for comparisons on an apple-to-apple basis for people with differences in health risk (such as differences in diagnosed health conditions). Within each county, the risk-standardized spending for each cohort is then averaged. To estimate spending given the health risk of the population of people who disenrolled from Medicare Advantage, the risk-standardized spending is then multiplied by the average risk score of people who disenrolled from Medicare Advantage. The national average spending for each cohort is then calculated as a weighted average of each county spending, using the number of people who disenrolled from Medicare Advantage as the weight. Counties in which no beneficiaries disenrolled from Medicare Advantage are excluded from the analysis. For each subgroup analysis a similar approach is followed, limiting the sample to people with the specific characteristic.
The Medicare Chronic Conditions Warehouse definitions for chronic conditions were used as a basis for identifying people with specific chronic conditions. To maintain consistency across the cohorts, only the 2022 claims were used. At least 1,000 members of the Disenrolled from Medicare Advantage cohort had to have a chronic condition for it to be included in the list.
Additionally, to test the sensitivity of the analysis, spending was modeled using a regression that controlled for risk score and county. The regression was specified using both a general linear model with a log link and a probit model. The average difference in each case was similar to those presented in this brief.
Overview of Medicare Advantage Payments
Medicare pays Medicare Advantage plans, in part, based on county-level spending for similar people in traditional Medicare, though payments may be higher or lower depending on a range of factors, such as a plan’s quality star rating, the county in which it operates, and the amount requested annually by the plan (referred to as a plan’s “bid”).
Plans that bid above traditional Medicare spending must charge enrollees an additional premium to make up the difference in cost. A growing number of plans choose to bid below traditional Medicare spending, which allows them to recoup a portion of the savings as a “rebate.” Rebates must be put towards certain plan features, such as lower cost-sharing, reductions to enrollees’ Part B or Part D premiums, or extra benefits (such as dental, vision, and hearing) not included in traditional Medicare. Payments are further adjusted to reflect the health status of enrollees, as well as certain other characteristics, such as age, sex, and Medicaid enrollment. Plans whose enrollees are expected to incur larger health costs receive larger payments (a process referred to as “risk adjustment”).
The Medicare Payment Advisory Commission estimates that payment to Medicare Advantage insurers is 122% of Medicare spending for similar beneficiaries covered by traditional Medicare. The higher payments are driven by favorable selection into Medicare Advantage and more intense coding of diagnoses among these plans. Those higher payments allow Medicare Advantage insurers to offer additional benefits and lower out-of-pocket spending, but place pressure on the hospital insurance trust fund and raise Medicare Part B premiums for all beneficiaries, including those in traditional Medicare.