Coding and Favorable Selection Associated With Increased MA Rebates
Key Takeaways:
- Implied rebates below the annual average: In an analysis of Medicare Advantage (MA) monthly rebates, calculations based on spending reductions alone yielded lower rebates than the actual 2024 average.
- Coding and selection raised rebates: After including the impact of coding and favorable selection, implied rebates increased and were closer to the observed average.
- Implications for future studies and policy makers: Further research is needed to evaluate other factors that might increase MA rebates, which may help inform policymakers during program reforms.
MA plans are designed so that rebates are financed by lower medical spending compared to traditional Medicare. However, coding and selection can increase MA spending plans, which were estimated to exceed traditional Medicare spending by 19% in 2024.
Using the statutory rebate formula and 2024 Centers for Medicare & Medicaid Services (CMS) administrative data, researchers calculated implied rebates for plans based on MA spending reductions while assuming no impact of coding or favorable selection.
MA Monthly Rebates
Across all plans included in the analysis, implied rebates were lower than the observed average rebate for 2024, which was $211 per member per month (PMPM) for non-employer plans. Assuming MA spending was 20% below traditional Medicare spending with 15% administrative costs, the implied rebate was $119 PMPM. The implied rebate was even less ($74 PMPM) at a 15% reduction and 15% administrative costs.
For conventional plans, the observed average rebate was $194 PMPM in 2024. At 20% lower spending and 10% administrative costs, the implied rebate was $137 PMPM. With the same spending reduction and 15% administrative costs, the implied rebate dropped to $107 PMPM.
Rebates for Special Needs Plans (SPNs) followed a similar pattern, though they were higher than rebates for non-employer and conventional plans, likely because the enrolled beneficiaries have higher expected spending. In 2024, the observed average rebate was $263 PMPM. Assuming a 20% spending reduction with 10% administrative costs, the implied rebate was $200 PMPM. When MA spending decreased by 20% and had 15% administrative costs, the implied rebate lowered further to $155 PMPM.
The Impact of Coding and Selection
When researchers repeated their analysis for non-employer plans, they calculated the effect of coding and selection according to Medicare Payment Advisory Commission (MedPAC) estimates: 19% in aggregate, with 8% coding and 11% selection.
Accounting for coding and selection significantly increased rebates for non-employer plans. Assuming 20% lower spending and 10% administrative costs, the implied rebate was $269 PMPM. At a 15% spending reduction and 15% administrative costs, the implied rebate was $202 PMPM, which was closer to the 2024 observed rebate.
Implications for Managed Care
The study suggests that additional sources of funding contribute to MA rebates, as implied rebates based on spending reductions were significantly lower than observed rebates. These rebates rose when coding and selection were included in the calculations but remained slightly below the observed rebate. Future analyses should consider other explanations for observed rebates, taking into account factors such as cross-subsidization and supplemental benefits.
These findings carry various considerations for policymakers, including the level and value of benefits, factors that generate rebates, and financing between MA and traditional Medicare.
The authors said, “Policy makers may opt for a gradual approach, focusing on technical updates to address program integrity concerns instead of becoming mired in disputes over the precise magnitude of coding and selection effects.”
Reference
Chernew ME, Sun J, McWilliams JM. Medicare Advantage rebates are too large to be financed by reductions in plans’ medical spending alone. Health Aff. 2026;45(9):978-985. doi:10.1377/hlthaff.2026.00332
