The American Journal of Managed Care
November 2014
Volume 20
Issue 11

Will Medicare Advantage Payment Reforms Impact Plan Rebates and Enrollment?

Medicare Advantage enrollment decreases with lower rebates for supplemental benefits. Upcoming ACA reforms are predicted to reduce MA enrollment where traditional Medicare costs are low.


To assess the relationship between Medicare Advantage (MA) plan rebates and enrollment and simulate the effects of Affordable Care Act (ACA) payment reforms.

Study Design and Methods

First difference regressions of county-level MA payment and enrollment data from CMS from 2006 to 2010.


A $10 decrease in the per member/per month rebate to MA plans was associated with a 0.20 percentage point (0.9%) decrease in MA penetration (P <.001) and a 7.1% decline in the average MA enrollee’s risk score (P <.001). These effects are small overall, but larger in counties with low levels of traditional Medicare spending; a $10 decrease in monthly rebates was associated with a 0.64 percentage point decline in MA penetration and a 10% decrease in risk score. ACA reforms are predicted to reduce the level of rebates in lower-spending counties, leading to enrollment decreases of 1.7 to 1.9 percentage points in the lowest-spending counties. The simulation predicts that the disenrollment would come from MA enrollees with higher risk scores.


MA enrollment responds to availability of supplemental benefits supported by rebates. ACA provisions designed to lower MA spending will predominantly affect Medicare beneficiaries living in counties where MA plans may be unable to offer a comparable product at a price similar to that of traditional Medicare.

Am J Manag Care. 2014;20(11):917-924

In many counties, payments to Medicare Advantage (MA) plans exceed average tra- ditional Medicare (TM) spending. Affordable Care Act (ACA) reforms designed to reduce payments may also reduce plans’ supplemental benefit offerings.

  • Lower MA plan rebates were associated with decreases in MA enrollment, especially by beneficiaries with higher risk scores (expected utilization).
  • The enrollment response was larger in counties with low TM spending.
  • ACA payment reforms are predicted to have small effects on enrollment, especially in counties where MA plan costs are belowTM costs, despite relatively large rebate reductions.


Private plans have been offered as an alternative to traditional, fee-for-service Medicare (TM) for over 30 years. Despite optimism that managed care would reduce Medicare spending, policy changes starting with the 2003 Medicare Modernization Act have ensured higher payment rates (relative to TM) to private plans in all counties through the Medicare Advantage (MA) program.In 2006, CMS introduced a bid system to determine payments to plans.



CMS sets annual, county-level payment benchmarks, the maximum monthly amount it will reimburse a private plan to provide standard Medicare benefits to an average-risk enrollee. Historically, benchmarks were not directly tied to cost of care in the county, though this will change under Afford- able Care Act (ACA) reforms.MA plans “bid” the amount it costs them to provide Medicare benefits in the counties they will enter. When bids are below the benchmarks, plans receive a rebate covering part of the difference between the benchmark and their bid (risk-adjusted by case-mix), which must be used to provide an actuarially equivalent amount of additional benefits or reduced premiums to enrollees. This approach aims to use competition between plans to lower Medicare expenditures while increasing supplemental benefits.




Rebates were initially set at 75% of the difference between the benchmark and the bid. ACA changes will gradually reduce this to 50%, 65%, or 70% of the difference, depending on a plan’s quality rating.ACA provisions will also tie county benchmarks to TM spending (from 95% of expected spending in the highest cost regions to 115% in the lowest cost regions). More generous benchmarks relative to plans’ true costs imply larger rebate amounts. Plans can use these rebates to offer benefits such as more comprehensive drug coverage that will attract sicker enrollees, or benefits that attract healthier enrollees such as fitness programs and preventive care.Prior research has found a positive relationship between overall payments to MA plans and firm and beneficiary participation.However, profound variation in average Medicare spending and the cost of care across geographic regions means that a higher payment rate in one county versus another does not necessarily mean that plans in the higher-rate county offer more generous supplemental benefits.


This study used newly available administrative data detailing the average rebates paid to MA plans between 2006 and 2010 to examine the relationship between rebates, a direct measure of supplemental benefit generosity, and 2 measures of MA enrollment (penetration rates and average MA risk scores), and to assess the extent that these results varied in counties with low versus high TM spending. Results are used to simulate the enrollment response to ACA policy changes. To date, little is known about whether these changes would be passed on to MA enrollees through supplemental benefit reductions, or if they would simply reduce plan profits.Understanding the impacts of benchmark reductions and changes to the rebate formula has important implications for ACA implementation and future Medicare spending projections.



I analyzed publicly available CMS data detailing county-level MA benchmarks from 2006 to 2010 linked to information about payments, enrollment, and patient risk scores reported at the county by plan-type level. These files were combined with information about average TM spending adjusted for labor costs from the Dartmouth Atlas of Health Care. I included MA enrollment in health maintenance organizations, local and regional preferred provider organizations, and private fee-for-service plans (excluding private plans available to certain subgroups of beneficiaries, such as employer-sponsored plans). Because CMS reports information aggregated at the county by plan-type level, risk scores and rebates are enrollment-weighted averages across these plan types.


Enrollment was measured as the proportion of Medicare beneficiaries enrolled in an MA plan (penetration) and the average MA risk score. Risk scores are a summary measure of expected utilization used to adjust payments for enrollee health relative to the “average” Medicare beneficiary (risk score of 1). Higher scores indicate higher levels of expected utilization.





Plan rebates, adjusted to 2010 dollars using the Consumer Price Index (results were unchanged whether we used the overall or medical Consumer Price Index for adjustment), represent the dollar value of supplemental benefits offered. While detailed benefit data are not available for the study years, rebates are positively correlated with availability of 2 generous plan offerings; availability of at least 1 zero-premium plan with prescription drug coverage (= .40); and availability of a zero-premium plan with expanded prescription drug coverage (= .38). CMS calculates each plan’s actual rebate using benchmarks and risk scores that account for the plan’s enrollee profile across all markets served, rather than for each plan-county combination. Thus the county-level average rebates analyzed capture variation in plans’ cost of providing care in each local market and each plan’s risk profile across markets. Under the ACA, rebates are further adjusted based on plans’ quality “star” ratings. The ACA simulation assumes a 65% rebate, corresponding to a plan with a quality rating between 3.5 and 4.5 out of 5 stars. As of 2012, the majority of plans and enrollment were between 3 and 5 stars.

Statistical Methods

I estimated first-difference regressions of the change in county-level MA enrollment measures from the previous year on changes in average plan rebates and control variables to test the hypothesis that Medicare beneficiaries change their MA enrollment decisions in response to plan generosity. These models capture the immediate enrollment effects of rebate changes relative to the prior year and control for time-invariant characteristics of each county that may be related to MA payments and enrollment.






ΔEnroll= β ΔRebate+ a ΔMarket+ g ΔCost+ Time + ε



Regressions controlled for the number of firms offering at least 1 MA plan, the MA market Herfindahl—Hirschman index (HHI), and the cost of providing care in the county. The HHI ranges from 0 to 1 and measures market competition. Higher scores indicate lower levels of competition, with 1 indicating that market enrollment is concentrated in a single plan. Enrollment and patient acuity may increase in more competitive markets, as beneficiaries have access to a more diverse choice set and may be more likely to find a plan appropriate for their needs. Local costs were measured by the ratio of price-adjusted average TM spending to unadjusted TM spending from the Dartmouth Atlas of Health Care. Price-adjusting accounts for factors like local wage rates that make some markets more costly.Regressions included a constant term and indicator variables for the years 2008 to 2010 to account for other factors influencing MA enrollment and plan rebates over time.

Regressions were weighted by county Medicare enrollment (aged 65 years and up). Since ACA benchmarks are tied to TM spending, regressions were estimated first by pooling all counties, then separately for counties in the lower versus higher halves of per capita TM spending.

ACA Predictions


Figure 1

Regression coefficients and 2010 data were combined with aggregate information about 2013 plan bids from the Medicare Payment Advisory Commission to estimate the change in rebates and corresponding enrollment response that would be expected under an immediate transition to full implementation of ACA benchmarks and rebates.The ACA benchmarks are 115% of expected TM spending in the quartile of counties with lowest TM spending, 107.5% in the next quartile, and 100% and 95% in the third- and fourth-highest quartiles, respectively. shows the 25th, 50th, and 75th percentile of plan bids and ACA benchmarks (relative to TM spending) by spending quartile. I assume that each county’s average bid follows the reported percentages of local TM spending and create an average measure that is the mean of the 3 reported bids.


Sample Characteristics

Table 1

Figure 2

The sample includes 3180 continental United States counties. ACA quartiles are county-rather than population-weighted; nearly 45% of Medicare beneficiaries lived in the highest spending quartile (). MA penetration averaged 22.9%, increasing from 19% in 2006 to 26% in 2010. MA benchmarks, payments, and rebates were highest in high TM counties. Monthly per enrollee rebates averaged $87 in the highest TM counties, compared with $52 to $57 in the first through third TM quartiles. MA enrollment was more common and had higher risk scores in high TM counties (mean risk score 1.01 in the highest quartile, 0.87 in the lowest). Rebates and risk scores fell over time, although MA penetration and county benchmarks (in constant dollars) increased ().

Regression Results


Table 2

eAppendix A,





After adjusting for market characteristics, a $10 decrease ($1 more than the average, within-county change during the study period) in monthly rebates was associated with a 0.20 percentage point decrease in MA penetration (<.001; , available at ). The magnitude of this relationship was lower in high-cost (—0.16 percentage point [pp]) versus low-cost (–0.64 pp, both <.001) counties. These effects are modest, translating to approximately 83,000 fewer MA enrollees in low-cost counties and 40,000 fewer in high-cost regions. Decreases in rebates were also associated with declines in the average MA risk score (—0.07 [7.1%], <.001), again modestly larger in magnitude in low-cost counties (—0.09 [10%)], <.001 vs —.07 [7.1%], <.001).

To address concerns that the enrollment response was driven by market factors and not the supplemental benefits funded through rebates per se, I verified that findings were robust to the exclusion of both market structure measures (the number of firms in the market and HHI), and HHI alone (eAppendix A) and inclusion of commercial market factors. If market structure influences the relationship between rebates and enrollment measures, we would expect to see the coefficients change due to omitted variables bias when these measures are excluded. However, the rebate coefficient for penetration and risk score is essentially unchanged across specifications.

Predicting Effects of ACA Benchmark Reductions

Table 3

Bids in 2013 from more than 25% of plans exceed the benchmarks plans would face under the ACA (Figure 1). These plans would have to charge premiums to MA enrollees, deliver standard benefits at lower cost, or leave the market in response to ACA reforms. Under ACA payment changes, rebates are predicted to fall in the lowest 3 TM quartiles and increase in the highest TM quartile (). Holding other market characteristics constant, rebates are predicted to fall by $27.54 (63%), $37.44 (118%), and $15.52 (37%) per member per month in the lowest 3 quartiles of TM spending, and increase by $36.69 (45%) in the highest quartile. Since most MA enrollment is in high-spending regions, the average change is a decline of only $1.84 (3.2%). Predicted re- bates would increase if enrollment was concentrated in the plans bidding at the 25th percentile (+$35.42) and exhibit a larger decline if plans in the 75th percentile dominate (—$48.51).

The enrollment response to changes in rebates is predicted to be small on average; an overall 0.43 percentage point (1.6%) decline relative to 2010 enrollment rates. Most of the decline would come from the lowest (—1.7 pp; –6.3%) and second-lowest (–1.9 pp; –8.2%) TM quartiles tempered by gains in the highest-spending counties (+0.75 pp; 2.7%). Enrollment would increase by 0.56 percentage points if plans bid at the 25th percentile of the current bid distribution, and would fall by 1.59 percentage points under bids at the 75th percentile.


The average risk score is predicted to fall by only 0.03 points (—3.2%), reflecting relatively large declines (–0.25 and –0.36) in the lowest TM counties and gains in the highest quartile (+0.26). In most cases, bids at the 25th percentile would be associated with changes in risk score of about 0.10 points (decreasing in the low TM quartiles and increasing in high TM counties), roughly 1 standard deviation. Only 25% of counties in the lower TM counties had average risk scores below 0.84 in 2010, so changes of this magnitude may be unlikely, particularly if beneficiaries accept higher premiums to keep their MA plans. As of 2013, 90% of Medicare beneficiaries have access to at least 1 plan that currently bids below their county’s ACA benchmark. Under ACA reforms, enrollees may face fewer plan choices but will retain the MA option.Since the available aggregate data precludes estimates of how MA enrollees might switch plans within a county rather than leave MA in response to ACA reforms, these results likely represent an upper bound on the risk score changes.


eAppendix B

The ACA also authorized Pay-for-Performance (P4P) bonuses to high-quality plans. The current demonstration version of this program has been criticized for providing bonuses to nearly all plans.I replicated the simulation assuming the continuation of the current P4P program, which increases benchmarks by 3.5% for 3.5-star plans. Maintaining the P4P bonuses would blunt the effects of the other ACA provisions; re- bates would experience smaller declines in the low TM counties (—$11.19 vs –$27.54 and –27.28 vs –44.18) and increase in high TM counties ().

Median plan bids would be associated with a $16.81 increase in monthly rebates, 0.14 pp increase in MA penetration, and 0.11 pp increase in MA risk scores.


Using administrative data from 2006 to 2010, I found that decreases in rebates to MA plans used to offer supplemental benefits are associated with decreases in both the number of MA enrollees and their average risk score. Efforts to reduce payments to MA plans under the Affordable Care Act will have small, negative effects on MA enrollment on average. Disenrollment effects will be larger in regions with low levels of TM spending, even though benchmarks in these regions will range from 107.5% to 115% of local TM spending. Findings suggest that implementation of ACA provisions designed to lower MA spending will predominantly affect Medicare beneficiaries living in counties where MA plans have been relatively unsuccessful in competing with TM and may be unable to offer a comparable product at similar price.


The ACA simulation assumed a 65% rebate, the level that would be offered to the majority of plans based on current quality metrics. However, either P4P bonus payments or true plan quality improvements could undo much of the cost-cutting from the other ACA provisions. If ACA payment changes restricting P4P bonuses to high-scoring plans are implemented, MA quality of care may improve as plans strive to reach the 4-star designation, conducting quality improvement efforts or exiting the market. Similarly, relatively inefficient or high-cost plans may be driven out of the market. There is a long history of plan withdrawal from unprofitable markets in response to a variety of payment changes to the Medicare Advantage program.

Higher rebates through quality improvement have mixed implications for beneficiaries. Exits by lower-quality plans may leave some Medicare beneficiaries with limited plan choice. Improving performance on elements of the star-rating quality scores may increase plan administrative costs through compliance with data collection and patient monitoring to achieve process and intermediate outcome end points. Plan investments to improve quality ratings may have the unintended consequences of crowding out spending on supplemental benefits that beneficiaries may value more than those targeted by the CMS rankings.




While most specifications found that effects of ACA changes would be modest on average, there was considerable regional heterogeneity. These policy changes would largely adversely affect MA beneficiaries in counties with low TM spending and benefit those in the highest-cost counties. These findings highlight a long-time challenge in the MA program that has important implications not only for short-term changes to the program itself, but for more comprehensive reforms such as premium support, which would provide Medicare beneficiaries with a defined contribution to purchase health insurance.A flat amount nationally would produce enormous inequity in the level of coverage Medicare beneficiaries are able to purchase.Even setting rates as a fixed proportion of local TM spending would also lead to substantial geographic differences in benefit availability.While plans in low-cost regions will struggle to provide any supplemental benefits at payment rates of up to 115% of TM spending, those in relatively high spending counties will be able to offer generous benefit packages when paid rates below local average TM spending.

Findings should be interpreted in light of several limitations. This analysis is based on aggregate data, precluding study of variation in plan costs and offerings within counties. It is unknown whether the increased enrollment within a county actually moves into the plans with the lowest bids and highest rebates. Plans that submit the lowest bid within a county may have low costs because they are good at attracting enrollees who will use less care than their risk score suggests (positive selection), or have restrictive provider networks that would discourage sicker (higher-risk) beneficiaries from enrolling. As sensitivity analyses showed, the expected effect of changes in the benchmark will vary considerably for an area served by relatively low-cost versus high-cost plans.


While the first-difference regression models used in this analysis allow me to control for time-invariant, county-level characteristics that may confound the relationship between MA rebates and enrollment, I cannot rule out bias from other, time-varying unmeasured characteristics, and results should not be interpreted causally. I am unable to model all aspects of plan entry and exit decisions and Medicare beneficiaries’ responses to these choices. Some MA enrollees already choose plans that charge additional monthly premiums, presumably because they value certain plan features. The enrollment effects estimated in this paper may overstate the consequences of rebate reductions if beneficiaries are willing to pay additional premiums to maintain plan choices or if plans are able to strategically bid in ways that enable them to continue to provide constant benefit levels. For example, Song and colleagues found that MA plans increase their bids in response to increases in county-level benchmarks.



Despite these limitations, this paper broadens our understanding of the ways that changes to the MA payment formula may impact program participation, especially in counties where TM spending is already low. Policy makers should assess whether additional spending necessary to sustain MA enrollment in these regions is desirable. Higher MA penetration appears to influence physician practice styles to lower TM spending and hospitalization rates,though it is unclear whether the value of supplemental benefits to MA enrollees meets or exceeds their cost of provision.Future CMS demonstrations could explore whether other benefit options such as expansion of prescription drug benefit subsidies through Part D produce greater health and satisfaction gains compared with increased MA payments.


Dr Nicholas had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. This study was supported by a grant from the Commonwealth Fund. Dr Nicholas is also supported by career development award K01AG04173 from the National Institute on Aging. The views expressed in this article do not necessarily represent the views of the Commonwealth Fund or the US government.

Earlier versions of this work were presented at the 2013 annual meetings of the International Health Economics Association and AcademyHealth.

Author Affiliation:

The Johns Hopkins Bloomberg School of Public Health, Baltimore, MD.

Source of Funding:

This study was supported by a grant from the Commonwealth Fund.

Author Disclosures:

Dr Nicholas received institutional grants from Commonwealth Fund, National Institute on Aging, and Russell Sage Foundation. She also presented research at the 2013 annual meetings of AcademyHealth and International Health Economics Association.

Address Correspondence to:

Lauren Hersch Nicholas, PhD, MPP, Johns Hopkins Bloomberg School of Public Health, 624 N Broadway, Baltimore, MD 21205. Email:

The Cost of Privatization: Extra Payments to Medicare Advantage Plans: 2005 Update

1. Biles B, Hersch Nicholas L, Cooper BS, Adrion E, Guterman S. . NewYork, NY: Commonwealth Fund; 2004.

Am J Manag Care

2. Song Z, Landrum MB, Chernew ME. Competitive bidding in Medicare: who benefits from competition? . 2012;18(9): 546-552.

Payment Basics: Medicare Advantage Program Payment System

3. MedPAC. . Washington, DC: Medicare Payment Advisory Commission; 2012.

Health Serv Res.

4. Atherly A, Dowd BE, Feldman R. The effect of benefits, premiums, and health risk on health plan choice in the Medicare program. 2004;39 (4 pt 1):847-864.

Med Care.

5. Ng JH, Kasper JD, Forrest CB, Bierman AS. Predictors of voluntary disenrollment from Medicare managed care. 2007;45(6): 513-520.

N Engl J Med.

6. Cooper AL,Trivedi AN. Fitness memberships and favorable selection in Medicare Advantage plans. 2012;366(2):150-157.

Int J Health Care Finance Econ.

7. Pizer SD, Frakt AB, Feldman R. Nothing for something? estimating cost and value for beneficiaries from recent Medicare spending increases on HMO payments and drug benefits. 2009;9(1):59-81.

Journal of Economics & Management Strategy.

8. Cawley J, Chernew M, McLaughlin C. HMO participation in Medicare+ Choice. 2005;14(3): 543-574.

J Health Econ.

9. Chernew M, DeCicca P,Town R. Managed care and medical expenditures of Medicare beneficiaries. 2008;27(6):1451-1461.


10. Frakt AB, Pizer SD, Feldman R.The effects of market structure and payment rate on the entry of private health plans into the Medicare market. 2012;49(1):15-36.

Report to the Congress: Medicare Payment Policy.

11. Medicare Payment Advisory Commission. Washington, DC: MedPAC; 2013.

12. Medical Spending, County-level, 2006-2010. 2013. Accessed 15 March 2012.

Introductory Econometrics: A Modern Approach

13. Wooldridge JM. . Mason, Ohio: South-Western Cengage Learning; 2003.

14. Cosgrove J, Emmanuelli-Perez E.Testimony Before the Committee on Oversight and Government Reform, House of Representatives: Medicare Advantage—Quality Bonus Payment Demonstration Has Design Flaws and Raises Legal Concerns. Washington, DC: US Government Accountability Office; 2012.

Health Aff (Millwood).

15. Biles B, Dallek G, Nicholas LH. Medicare Advantage: déjà vu all over again? 2004; Suppl Web Exclusives: W4-586-597.

Health Aff(Millwood).

16. Berenson RA. Medicare disadvantaged and the search for the elusive “level playing field.” 2004;Suppl Web Exclusives:W4-572-585.

Health Aff (Millwood).

17. Berenson RA, Dowd BE. Medicare advantage plans at a crossroads--yet again. 2009;28(1):w29-w40.

N Engl J Med.

18. Antos JR. The Wyden—Ryan proposal--a foundation for realistic Medicare reform. 2012;366(10):879-881.

J Health Polit Policy Law.

19. RiceT, Desmond KA.The distributional consequences of a Medicare premium support proposal. 2004;29(6): 1187-1226.

Health Aff (Millwood).

20. McBrideTD. Disparities in access to Medicare managed care plans and their benefits. 1998;17(6):170-180.

Health Aff (Millwood).

21. Wennberg JE, Fisher ES, Skinner JS. Geography and the debate over Medicare reform. 2002;Suppl Web Exclusives:W96-W114.

J Health Econ

22. Song Z, Landrum MB, Chernew ME. Competitive bidding in Medicare Advantage: effect of benchmark changes on plan bids. . 2013;32(6):1301-1312.

J Health Econ.

23. Baicker K, Chernew ME, Robbins JA.The spillover effects of Medicare managed care: Medicare Advantage and hospital utilization. 2013;32(6):1289-1300.

Related Videos
Ali Khawar
Neil Goldfarb, CEO, Greater Philadelphia Business Coalition on Health
Screenshot of Mary Dunn, MSN, NP-C, OCN, RN, during a video interview
Inma Hernandez, PharmD, PhD, professor at the University of California, San Diego Skaggs School of Pharmacy and Pharmaceutical Sciences
Carrie Kozlowski
Carrie Kozlowski, OT, MBA
Carrie Kozlowski, OT, MBA
kimberly westrich
Bruce Sherman, MD
Related Content
CH LogoCenter for Biosimilars Logo