Association for Public Policy Analysis & Management

The Trump Administration and the Medicare Payment Advisory Commission (MedPAC), among others, have proposed an overhaul of important aspects of the Medicare Part D prescription drug program. Two important components of these multi-faceted proposed reforms are a requirement that plans share rebates with beneficiaries at the point-of-sale (currently reflected in a proposed rule to remove the safe harbor for these rebates under the Federal anti-kickback statute) and reducing the federal reinsurance subsidy, which currently pays 80% of beneficiary spending in the catastrophic coverage region.

In order to evaluate the potential impact of these proposed reforms, we investigate historical data on the empirical relationship between Part D plan bids, reinsurance bids, and beneficiary spending by linking CMS Part D plan bid data, reconciliation data, and claims data from a 100% sample of beneficiaries from 2007 through 2017. We then simulate the potential effect of these changes to the Part D program by “re-processing” claims under these alternative policy scenarios and evaluate their impact on premiums, total and out-of-pocket spending, plan bids, and federal spending. Moreover, we evaluate the distributional effects and incorporate potential behavioral responses.

In our analyses of historical data, we find that plan bids for expected federal reinsurance subsidies increased more than 125% (per member per month (pmpm)) over the study period. However, in recent years, plans increasingly under-bid their expected reinsurance subsidies compared to actual reinsurance liability on average, resulting in excess federal spending through reconciliation payments. At the same time, plan liability derived from claims data became increasingly divergent from plan bids over this time period due to considerable growth in rebates, which increased from 9.6% of total part D program spending in 2007 to 19.9% in 2016 (2018 Medicare Trustees Report). As a result, plans were able to use these growing rebate dollars to depress plan bids and hold beneficiary premiums low over this time period, despite being responsible for an increasing share of total Part D spending due to ACA provisions to fill in the doughnut hole.

These findings have important implications for the potential effects of proposed reforms to the Part D program, since these reforms would not only change the incentives of the current program structure but would also reduce the ability of plans to engage in strategic bidding behavior. We estimate that the impact of these two proposed reforms to the Part D program would result in beneficiary premium increases of around 27% (or nearly $10 pmpm), with the effect roughly evenly attributable to the requirement to pass-through rebates at the point-of-sale and reforming the structure of the reinsurance program. The impact on beneficiary out-of-pocket spending varies considerably across beneficiaries.

In sum, implementing these proposed reforms would help address current distortions in the Part D market and mitigate plans’ ability to game the bidding structure, but they are likely to result in considerable premium increases. As a result, policymakers should consider mechanisms to mitigate potential unintended consequences, such as disenrollment from the Part D program due to large premium increases.