Humana study finds Medicare Advantage favorable selection lower
Humana Healthcare Research reveals that favorable selection in Medicare Advantage is 42%-65% lower than MedPAC's 10%-15% estimates, placing it at 4.9%-5.8% for 2020-2022. Published in Health Affairs Scholar, the study uses comprehensive encounter data from the full MA population, challenging previous extrapolation methods. This finding suggests a smaller impact on MA payments than previously thought, urging further methodological refinement in payment policy debates.

*this image is generated using AI for illustrative purposes only.
Humana Healthcare Research (HHR) has published a new study in Health Affairs Scholar finding that favorable selection in Medicare Advantage (MA) is substantially lower than estimates previously cited by the Medicare Payment Advisory Commission (MedPAC). The research suggests that the impact of favorable selection on MA payments may be smaller than current policy debates indicate, driven by a methodological refinement that utilizes comprehensive encounter data rather than extrapolation from limited subsets.
The study analyzed data from the full MA population between 2020 and 2022. By evaluating this complete dataset directly, HHR estimated favorable selection at 4.9%-5.8%. This figure represents a significant downward revision from the 10%-15% range frequently referenced in policy discussions and earlier MedPAC analyses. The primary driver of this discrepancy was the shift from relying on external studies and limited enrollee subsets to using comprehensive encounter data for direct evaluation.
Key Findings
The core findings of the HHR study are summarized below:
| Metric | Previous Estimates (MedPAC/Others) | New HHR Estimate | Reduction |
|---|---|---|---|
| Favorable Selection | 10%-15% | 4.9%-5.8% | 42%-65% |
| Data Source | Limited subsets/Extrapolation | Full MA Population | N/A |
| Period Analyzed | N/A | 2020–2022 | N/A |
The lower estimates were achieved through a single methodological refinement: the use of comprehensive encounter data to evaluate the full MA population directly. This approach contrasts with prior methods that relied on extrapolation derived from external studies and a limited subset of enrollees. The findings imply that policy assumptions based on higher favorable selection rates may overstate the financial impact on MA payments.
Implications for Payment Policy
Dr. Sanjay Shetty, President of CenterWell, Humana’s health services organization, stated that the study contributes important new evidence to the Medicare Advantage payment conversation. He noted that the substantial reduction in estimates resulting from a single methodological change suggests there is still more to learn about measuring this complex issue. The findings highlight the value of comprehensive data and the importance of continued research to inform MA payment policy.
More broadly, the study demonstrates that full-population comparisons can produce more grounded estimates. This enables extrapolations that are more directly informed by the actual MA population rather than inferred from smaller samples. The research underscores the need for continued methodological refinement in assessing favorable selection to ensure accurate policy formulation.
How might CMS adjust Medicare Advantage risk-adjustment models in response to HHR's lower favorable selection estimates?
Will this study influence the upcoming budget negotiations regarding the overall funding levels for the Medicare Advantage program?
Could competitors like UnitedHealth Group or CVS Health challenge these findings by proposing alternative methodologies for calculating favorable selection?

































