The Shifting Budgetary Outlook for Medicare Part D  | American Enterprise Institute

Medicare part D, which covers outpatient prescription drugs, is undergoing a planned transformation from multiple policy interventions, most of which were advertised as relieving financial pressure by lowering the prices charged by the manufacturers. What was not signaled as a possible outcome was a significant net increase in federal part D costs, but that is what is now occurring. 

Figure 1 compares projected part D spending from the 2023 Medicare trustees report, which was the first forecast published after approval of the Inflation Reduction Act’s part D revisions, with the estimates contained in the 2026 release. The latter forecast placed cumulative part D spending over the period 2023 to 2032 at $2.3 trillion, which is more than $0.6 trillion above what was projected in the 2023 report. The current projection for part D spending in 2032 is now $99 billion, or 48 percent, above what was estimated three years ago.  

Figure 1. Projected Medicare Part D Spending, 2023 to 2032 

The Congressional Budget Office (CBO) also underestimated part D spending in the aftermath of the IRA’s passage. In its February 2026 baseline, the agency acknowledged raising projected part D spending over the period 2026 to 2035 by $0.6 trillion compared to its forecast from January 2025. Across all of Medicare, CBO raised its baseline estimates by $1.0 trillion over this same period. In a May 2026 letter, the three Republican chairmen of key House committees (Budget, Energy and Commerce, and Ways and Means) pointed to CBO’s February 2026 part D spending revision and asked the agency to explain how its estimates at the time the IRA was approved could have been so far off the mark.  

The long-term consequences of higher Medicare prescription drug coverage spending, if it persists, will be substantial. In the 2026 trustees report, the Medicare actuaries estimated the present value of part D costs in excess of premiums and other receipts over the next 75 years at $11.5 trillion, or 0.6 percent of future GDP. In 2023, the estimated shortfall was $7.6 trillion, or 0.4 percent of GDP.  

Under the original part D law, the share paid by beneficiaries (25.5 percent) versus the government (74.5 percent) was modeled on part B for ambulatory services. To guard against large spikes from the law’s many complex program adjustments, the IRA capped the annual growth of the base beneficiary premium at 6.0 percent which, if operative in any year, would be expected to shift more of the cost burden onto taxpayers. At the same time, to prevent a permanent downward slide in the share paid by beneficiaries, the IRA set a new floor for the base premium at 20.0 percent of total costs starting in 2030, which the 2026 trustees report projects will trigger a 48 percent increase. As 2030 approaches, pressure is likely to build on Congress to find a way to prevent this large year-over-year increase in beneficiary out-of-pocket costs.  

Both the Biden and Trump administrations have added to the budgetary pressure with Medicare demonstrations that liberalized benefits with little prospect of offsetting savings.   

In 2024, to further smooth the transition to the IRA’s new part D design for standalone prescription drug plans (PDPs), Biden officials approved a temporary relief program to offset potential premium shocks over the period 2025 to 2027. CBO estimated the cost of the demonstration in 2025 alone at $5.0 billion. (The Trump administration later modified the demonstration to provide less support to PDPs in 2026.)  

For its part, the Trump administration recently approved time-limited Medicare coverage for GLP-1 products for weight control for a $50 monthly beneficiary copayment (part D plans will not be charged as part of the test). This new “Bridge” demonstration, which is scheduled to run from July 2026 through 2027, may create political pressure to continue broad and permanent Medicare access to GLP-1 therapies even though Congress has never approved that policy.  

Both the Biden and Trump demonstrations rely on section 402 of the Social Security Act Amendments of 1967 which does not require budget neutrality. Although section 402 tests are supposed to be research experiments, there is little to suggest these interventions were developed for that purpose.  

Part D’s recent cost saga fits a pattern. Major entitlement liberalizations are often promoted as budget neutral (at worst) based on the assumption that other changes will offset the costs of more generous program benefits. Then, after enactment, the calculation shifts, and the promised benefits turn out to be more costly than expected, while the provisions relied upon to cut future spending underdeliver for any number of reasons. It is happening again with the IRA. Unfortunately, the benefit cliffs built into the Trump demonstration in 2027 and the IRA in 2029 could offer new opportunities to repeat the story again.