Answers to Questions for the Record Following a Hearing on Alternative Payment Models and the Slowdown in Federal Health Care Spending

United States Capitol Building in Washington, DC

On October 18, 2023, the Senate Committee on the Budget convened a hearing at which Chapin White, the Congressional Budget Office’s Director of Health Analysis, testified about alternative payment models and the slowdown in federal health care spending. After the hearing, Ranking Member Grassley submitted questions for the record. This document provides CBO’s answers. It is available at

Ranking Member Grassley’s Questions About CBO’s Projections of Federal Health Care Spending

Question. During the hearing, you stated that the Congressional Budget Office currently estimates the federal government from 2010–2033 will spend $6.3 trillion less on major health care programs compared to an August 2010 CBO estimate. You stated $1.1 trillion was attributed to the 2010–2020 period and $5.2 trillion was attributed to the 2021–2033 period. What makes up the $6.3 trillion difference? Are there different reasons for changes in estimates for the 2010–2020 period and the 2021–2033 period? How much does CBO attribute the lower estimates to Medicare Part D and why?

Answer. CBO’s 2010 projections for the 2010–2020 period overestimated spending on function 550 (Health, mostly for the Medicaid program) and function 570 (Medicare, net of premiums and other offsetting receipts) by $1.1 trillion. The difference between projected and actual mandatory outlays was $635 billion for function 550 and $431 billion for function 570. Legislative changes increased CBO’s estimate by $15 billion for function 550 and by $106 billion for function 570, but those increases were more than offset by technical changes that decreased the agency’s estimate. Most of CBO’s overestimate of spending for Medicare and Medicaid stemmed from an overestimate of spending per beneficiary and not an overestimate of the number of beneficiaries.

There were two significant sources of error in CBO’s 2010 projections. The first significant source of error was less-than-anticipated spending on prescription drugs in Medicare Part D (the program that covers the cost of beneficiaries’ outpatient prescription drugs). Actual Medicare Part D net outlays were about $333 billion lower over the 2010–2020 period than CBO projected in its August 2010 baseline. That amount represents about 30 percent of the $1.1 trillion difference. The agency identified two reasons for the slower-than-expected growth in prescription drug spending, both nationally and in Part D. First, as existing brand-name drugs lost their patent protection, they faced new competition from generic drugs, and a significant share of prescriptions shifted to less expensive generic formulations. Second, fewer new brand-name drugs, which would have been more expensive, were introduced than CBO had anticipated.

The second significant source of error in CBO’s 2010 projections was slower-than-anticipated growth in spending on long-term services and supports (LTSS) in Medicaid. The slower growth in spending was driven by two factors. First, the number of users of non­institutional LTSS grew more slowly than it did from 2000 to 2010. Second, states have increasingly shifted patients from institutional to noninstitutional settings (where care is provided at a lower cost), and more institutional services have been delivered by managed care plans (which actively seek to control costs). Both of those alternative-care delivery mechanisms are generally less costly on a per user basis.

CBO’s February 2023 projections of spending on the major federal health care programs for the 2021–2033 period were $5.2 trillion lower than what the agency had projected in June 2010 for that same period. Medicare accounted for about one-third of the $5.2 trillion difference, and Medicaid, the Children’s Health Insurance Program, and the Affordable Care Act marketplace subsidies accounted for about two-thirds of that difference. CBO’s 2010 projections for the 2021–2033 period did not have a separately identifiable estimate of net spending in Part D.

One reason for the lower projection for the 2021–2033 period is that CBO previously overestimated spending on major health care programs over the 2010–2020 period. Another reason is that in 2010, the agency’s long-term projection methods had not yet been updated to reflect the significantly slower rate of growth in federal spending on major health care programs in recent decades.

Question. Major health care program spending takes up 32 percent of federal revenue today, and it will be 45 percent of revenue by 2053. CBO projects spending on Medicare to account for more than four-fifths of the increase in spending on federal health care programs over the next 30 years. CBO has also said that Medicare’s increased spending largely stems from rising health care costs per person and demographic trends. How much of this increased spending is due to rising costs per person and how much is related to demographic trends? Does adding new procedure codes play a role?

Answer. Over the 2023–2053 period, growth in health care costs per person accounts for over two-thirds of the increase in spending, measured as a percentage of gross domestic product (GDP), on the major health care programs.1 Over that same period, about one-third of the projected increase in total spending on the major health care programs, measured as a percentage of GDP, is attributable to the aging of the population.

CBO’s long-term projections of federal health care spending are driven, in part, by the continued emergence of new technologies and by rising incomes, which allow for the adoption and diffusion of such technologies. The agency does not project the number of procedure codes or the emergence of specific procedure codes.

1. The analysis of the causes of the growth in spending on the major health care programs encompasses gross spending on Medicare and does not reflect receipts credited to the program from premiums and other sources.