On March 11, 2026, the Subcommittee on Fiscal Responsibility and Economic Growth of the Senate Committee on Finance convened a hearing at which Phillip L. Swagel, the Congressional Budget Office’s Director, testified about the agency’s most recent analysis of the outlook for the budget and the economy.1 After the hearing, Senators Chuck Grassley and Elizabeth Warren submitted questions for the record. This document provides CBO’s answers. It is available at www.cbo.gov/publication/62378.
Senator Grassley’s Questions About Historical and Projected Noninterest Spending
Question. During the hearing it was asserted by one of my colleagues that federal spending excluding net interest costs is about the same today (at just under 21 percent of GDP) as it was in 2013. Has non-interest spending typically exceeded 20 percent of GDP or was 2013 a historical outlier? Please explain.
Answer. Since 1940, the U.S. government’s noninterest outlays have exceeded 20 percent of gross domestic product (GDP) in 14 years: 1942 to 1946, 1983, 2009 to 2012, 2020 to 2022, and 2024.2 Those years generally correspond to periods of war, economic downturns, or the COVID-19 pandemic. In 2025, noninterest outlays equaled 19.9 percent of GDP, about one-half of one percentage point higher than they were in 2013.
Question. Over the next 10 years, CBO projects that spending excluding net interest costs to average 19.8 percent of GDP. How much lower would debt and deficits be over the next 10 years if Congress acted to reduce non-interest spending to the pre-covid 50-year average?
Answer. Over the 50 years preceding the pandemic (1970 to 2019), noninterest outlays averaged 18.3 percent of GDP. If such outlays remained at that percentage from 2027 to 2036, the cumulative primary deficit (the total budget deficit excluding net outlays for interest) over that period would be an estimated $2.7 trillion—which is $5.6 trillion less than the $8.3 trillion in CBO’s February 2026 baseline budget projections.3 In 2036, debt held by the public would equal about 106 percent of GDP, CBO estimates, instead of the 120 percent projected under current law.
Question. What are the main reasons spending is projected to remain above pre-covid levels over the next 10 years and then escalate further in future years?
Answer. In CBO’s February 2026 baseline projections, noninterest outlays average 19.7 percent of GDP over the 2027–2036 period—higher in each year than the 19.1 percent of GDP such outlays equaled in 2019. Nearly all of that increase is attributable to spending on Social Security, Medicare (net of offsetting receipts, which are mostly in the form of premiums paid by beneficiaries), and Medicaid and is driven by an increase in the average age of the population and the rapid growth of federal health care costs.
The number of people age 65 or older is now almost three times what it was 50 years ago, and that number is expected to increase by about 15 percent over the coming decade. As a result, enrollment in Social Security and Medicare will grow. In CBO’s projections for the 2026–2036 period, growth in the number of those programs’ beneficiaries increases nominal spending on Social Security benefits and Medicare by 19 percent and 20 percent, respectively. Over that same period, rising costs per beneficiary (adjusted to remove the effects of inflation) boost nominal spending on Medicare and Medicaid by 41 percent and 18 percent, respectively.
In CBO’s projections, outlays for Medicare, Medicaid, and Social Security equal 12.2 percent of GDP in 2036; such outlays amounted to 9.8 percent of GDP in 2019. Outlays for all other mandatory programs amount to 2.8 percent of GDP in 2036, which is less than the 3.0 percent they totaled in 2019. Discretionary outlays, which totaled 6.3 percent of GDP in 2019, are estimated to equal 5.9 percent of GDP in 2026 and then decrease in each year of the projection period, relative to the size of the economy.
Senator Grassley’s Question About the Difference Between CBO’s and the Medicare Boards of Trustees’ Projected Exhaustion Date of the Balance of the Hospital Insurance Trust Fund
Question. The Congressional Budget Office (CBO) recently estimated that the Medicare Hospital Insurance (HI) Trust Fund balance will be exhausted in 2040. This was 12 years earlier than the previous CBO estimate. However, the most recent Board of Trustees for Medicare report projected the HI Trust Fund balance will be exhausted in 2033, which was revised down from 2036 in the Trustees’ prior estimate. How do CBO’s assumptions differ from the Trustees in estimating the exhaustion date of the HI Trust Fund? Please be sure to specify all material differences in the assumptions used and any reasons for the differences. Also, please address all assumptions that contribute to the difference in per enrollee spending growth.
Answer. The main reason that CBO’s estimated exhaustion date of the balance of the Hospital Insurance (HI) Trust Fund is later than the date estimated by Medicare’s Boards of Trustees is that expenditures from the fund from 2026 to 2034 are smaller in CBO’s projections than in the trustees’. In CBO’s February 2026 baseline budget projections, such expenditures over that period are $363 billion less than projected in the Medicare trustees’ June 2025 report.4 That difference is partly offset by CBO’s lower projection of the fund’s income over that same period, which is $12 billion less than the trustees projected.
The timing of the release of CBO’s and the Medicare trustees’ projections also contributes to the difference in their estimates. The trustees’ report was released on June 18, 2025—about eight months before CBO published its most recent estimate.5 During the intervening period, two things occurred that affected CBO’s estimate: CBO revised its projections to reflect the effects of Public Law 119-21, more commonly known as the 2025 reconciliation act, which was enacted in July 2025; and the agency received updated data about spending on Medicare Part A. (Such spending is funded through the HI trust fund.) CBO’s latest projections of the fund’s finances reflect that additional information, which was not available when the trustees’ report was released. CBO will continue to update its projections of expenditures from the HI trust fund as new data become available.
The difference in CBO’s and the trustees’ projections of expenditures from the HI trust fund is driven by different estimates of growth in per capita spending on services provided through Medicare Part A. CBO’s projection of annual growth in such spending over the 2026–2034 period is 1.3 percentage points lower, on average, than the Medicare trustees’ projection for that same period. CBO’s projection of the growth rate of enrollment in Part A is within 0.1 percentage point of the Medicare trustees’ projection.
In recent decades, annual growth in costs per beneficiary for Part A has varied considerably, making future growth difficult to project. Over the calendar year 2000–2010 period, costs per beneficiary grew by an average of 4.5 percent a year.6 Such growth slowed markedly over the calendar year 2010–2020 period, averaging 0.6 percent a year, but then increased to an average of 3.6 percent a year over the calendar year 2020–2024 period.
In CBO’s February 2026 baseline projections, Part A spending per beneficiary grows by an average of 3.5 percent a year from 2024 to 2034—which is similar to the rate observed over the calendar year 2020–2024 period but lower than the rate over the calendar year 2000–2010 period. In the Medicare trustees’ projections, growth in costs per beneficiary averages 5.1 percent a year from 2024 to 2034, which is closer to the rate observed over the calendar year 2000–2010 period (before spending growth slowed). CBO’s estimates of growth in Medicare Part A costs per beneficiary reflect current law, the agency’s latest demographic projections, and the latest available data about spending on Medicare Part A claims.
Differences between CBO’s and the Medicare trustees’ estimates of expenditures from the HI trust fund are largely driven by different estimates of spending on services provided in skilled nursing facilities (SNF) and inpatient hospital services. From 2025 to 2034, differences in projected spending on SNF services account for roughly 40 percent of the total difference in the projections of Part A fee-for-service (FFS) expenditures; differences in projected spending on inpatient hospital services account for about 30 percent of the total difference. In 2024 and 2025, inpatient hospital services accounted for 71 percent of Part A FFS spending, whereas SNF services accounted for 13 percent of such spending. Given the distribution of Part A FFS spending between those service lines, the difference in projected spending on SNF services is notable.
Unlike their projections of expenditures, CBO’s and the trustees’ projections of the HI trust fund’s total income from 2026 to 2034 are relatively similar. However, CBO’s projections of income from payroll taxes and the taxation of Social Security benefits are lower than the trustees’ June 2025 projections—by $47 billion in the case of payroll taxes and by $101 billion in the case of the taxation of benefits. Because the trustees’ report was released in June 2025, it did not account for the effects of the 2025 reconciliation act on individual income taxes, which are projected to reduce income to the trust fund. The difference in the projections of income from taxes is largely offset by CBO’s estimates of interest income and income from other sources, which are higher than the trustees’ estimates by $33 billion and $103 billion, respectively.
Senator Grassley’s Questions About the Household Saving Rate
Question. There is overwhelming agreement among budget experts that the nation’s fiscal trajectory is unsustainable. While most experts refuse to specify a tipping point, economists at the Penn Wharton Budget Model have estimated that “U.S. debt held by the public cannot exceed about 200 percent of GDP.”7 In their view, while Japan has been able to exceed such levels the U.S. would be unable to given the U.S has a far lower household saving rate. How does a nation’s household saving rate impact its ability to finance larger government debt?
Answer. Household saving (or personal saving, which is the term used by the Bureau of Economic Analysis and includes saving by households and nonprofit institutions serving households) is an important source of funding for the purchase of the government’s debt. The higher the personal saving rate, the greater a nation’s capacity to finance larger government debt domestically. Conversely, a low personal saving rate requires the government to rely more on foreign investors to purchase its debt.
Question. As the national debt ascends into uncharted territories, what assumptions does CBO make about the nation’s household saving rate?
Answer. In CBO’s most recent baseline budget projections, the personal saving rate increases from 5.1 percent in 2026 to 6.6 percent in 2036, nearly reaching its average of 6.8 percent over the 40 years (1980 to 2019) before the COVID-19 pandemic. A rising personal saving rate would expand the nation’s capacity to absorb the growing amount of debt held by the public, which is projected to total 120 percent of GDP in 2036.
Senator Warren’s Questions About the Cost and Economic Effects of Tariffs
Question. The Congressional Budget Office’s report, The Budget and Economic Outlook: 2026 to 2036, analyzes the impact of President Trump’s erratic tariffs on the U.S. economy. The report predicts that foreign companies will absorb just 5% of tariff costs. Of the remaining 95%, American companies will absorb 30% in the short term, with the rest falling to American consumers in the form of price increases. Does this contradict President Trump’s oft-repeated claim that foreign companies will simply “eat” the cost of tariffs?
Answer. In CBO’s assessment, people and businesses in the United States have borne most of the cost of the tariffs imposed since the start of 2025. If foreign exporters had fully absorbed that cost, import prices—that is, the prices paid by U.S. importers before tariffs are applied—would have declined by an amount that offset the cost of the tariffs paid by importers. CBO’s analysis of data on import prices since the tariffs were implemented indicates that pretariff price decreases have been considerably smaller than would be consistent with foreign exporters’ fully absorbing the tariffs’ cost.
Available evidence suggests that foreign exporters have absorbed a relatively small share—about 5 percent, on average—of the cost of the tariffs. That conclusion is consistent with results of analyses by academic researchers, think tanks, and other forecasters.8 Foreign exporters might have fully absorbed the cost of tariffs on some products, but evidence indicates that such cases do not reflect overall trends.
Question. Has your office done any subsequent analysis to determine if this prediction has borne out in practice?
Answer. The assessment of tariffs published in the latest edition of CBO’s The Budget and Economic Outlook relied on analysis conducted by the agency last fall and was supported by findings reported by other researchers at that time.9 Some more recent evidence indicates that the share of the cost of tariffs absorbed by foreign exporters may have increased over time, possibly reaching a range of 10 to 15 percent toward the end of calendar year 2025.10 CBO will continue to monitor data and research on responses to tariffs and will update its estimates as additional information becomes available.
Question. Your report also predicted that, while American consumers will shoulder the costs of 70% of tariffs in the “near term,” they will ultimately end up paying for 95% of tariffs in the long term. Please define “near term” in the context of this report.
Answer. In the context of that discussion in CBO’s report The Budget and Economic Outlook: 2026 to 2036, “near term” refers to the time period until the end of calendar year 2026. CBO projected that by then, businesses would fully adjust their prices in response to the tariffs put in place as of November 20, 2025. Those price adjustments reflect higher prices from businesses whose production inputs are affected by those tariffs as well as price increases by domestic businesses whose products compete with foreign imports.
Question. Since releasing this report, has your team done any analysis to determine what percentage of tariff costs have already fallen to consumers?
Answer. Incoming data on consumer prices have been consistent with CBO’s recent forecast and the agency’s assessment of the effects of tariffs on such prices. Recent data have shown elevated core goods inflation—particularly in tariff-exposed categories such as apparel, furniture, and appliances—that is consistent with heightened costs of tariffs passing through to consumers. Those findings are subject to uncertainty because completely isolating the effects of tariffs from other factors that affect consumer prices is not always possible. CBO will continue to monitor the relevant data and will update its expectations as new information becomes available.
Question. Your report noted that President Trump’s erratic tariffs will have negative impacts on the economy, by both discouraging investment and growth while pushing up prices. Specifically, you noted that: “The higher and frequently changing tariffs put in place in 2025 are expected to affect the U.S. economy through several interrelated channels. Those policies—together with the actions of major trading partners—will, in CBO’s assessment, temporarily raise the rate of inflation, reduce real investment (that is, investment adjusted to remove the effects of inflation), lower the level of real gross domestic product (GDP), and reduce employment in relation to what would have occurred without the changes in trade policy.” Does the above indicate that CBO assesses that President Trump’s tariffs could result in both inflation and slowed growth, or stagflation?
Answer. CBO estimated that the tariffs in place as of November 20, 2025, would temporarily increase inflation and reduce economic growth in relation to what they would otherwise have been. Stagflation generally refers to periods of persistently high inflation and low economic growth. That term is most often used to describe the economic difficulties in the United States during the 1973–1982 period, when inflation regularly exceeded 8 percent annually and real (inflation-adjusted) GDP growth was often weak or negative.11 CBO has not assessed the likelihood that recent changes to tariff policies would result in stagflation. The agency’s estimates of the effects of recent tariffs on prices and economic output are small in magnitude compared with historical periods of stagflation. Current tariffs alone are unlikely to result in the persistent inflationary pressure and weak economic growth that characterize stagflation.
Question. Since this report was released, the President has launched an illegal war in the Middle East, which has rattled markets—decreasing investments—and choked off supply chains—increasing prices. Do you anticipate that President Trump’s war could exacerbate the stagflationary impacts of President Trump’s trade agenda?
Answer. CBO has not completed an analysis of the likely economic effects of the conflict with Iran. Many outside experts expect that a prolonged conflict, particularly one that disrupted shipments of oil and related products through the Strait of Hormuz, would put upward pressure on prices and slow global economic growth.12
Senator Warren’s Questions About the Golden Dome for America
Question. On May 5, 2025, CBO released an analysis that concluded that SBI constellations could cost between $161 billion and $831 billion. On March 17, 2026, the Department of Defense announced that the cost estimate for “Golden Dome for America” increased by $10 billion in recent months. Will the CBO update its own cost estimates with any new information it receives that changes its analysis for the total cost of the “Golden Dome for America”?
Answer. Yes. In response to a Congressional request, CBO is currently estimating the cost of implementing the executive order titled “The Iron Dome for America” (which refers to the defense initiative now commonly known as the Golden Dome for America, or Golden Dome).13
Question. What does CBO estimate to be most likely to drive further cost risks or increases for Golden Dome?
Answer. CBO has not determined which aspects of Golden Dome are most susceptible to cost growth. To meet the goals of the executive order, the Department of Defense (DoD) would need to deploy a range of missile defense sensors and interceptor systems and a comprehensive command and control system to connect all the components of Golden Dome. Some of those components would probably be based on existing systems, particularly the ground-based interceptor and radar systems that have been deployed in the United States, the Middle East, and Asia. Other components, such as space-based, boost-phase interceptors and space-based sensors have never been deployed and must be developed. New types of systems are more susceptible to cost growth than existing systems.
Senator Warren’s Questions About Insourcing DoD’s Work
Question. DoD’s Office of Cost Assessment and Program Evaluation has assessed there could be significant cost savings from insourcing Department of Defense (DoD) work. Is CBO capable of completing an analysis on the cost of insourcing versus outsourcing the Department of Defense workforce?
Answer. CBO has not analyzed potential savings from insourcing DoD’s work, and the agency’s ability to quantify such savings would be limited by access to data and information about specific contracts. In general, insourcing can reduce costs in some situations (for example, when specialized expertise is an enduring need), as can outsourcing (for example, when specialized expertise is needed for a short time). The savings or costs would depend on the specific circumstances of the work and the details of any contracts involved. Other considerations can also affect the choice of insourcing or outsourcing work. For example, outsourcing could provide flexibility to quickly expand DoD’s workforce in response to new tasks, and insourcing could help DoD develop and retain human capital.
Producing a detailed quantitative analysis of all DoD’s contracts would require far more resources than CBO has. However, CBO may be able conduct a qualitative analysis to describe the types of circumstances in which insourcing or outsourcing, respectively, could reduce costs and to identify other noncost factors that could affect decisions about which approach to take. In that case, data from DoD would be useful in providing examples of such circumstances.
Some of CBO’s past analyses are related, but do not directly apply, to the question of potential savings from insourcing DoD’s work. In 2015, CBO analyzed the savings from replacing military personnel in support positions with federal civilians, but that study did not examine how the cost of contractors compared with costs of military personnel or federal civilians, because there were not sufficient data to do so.14 And in 2023, for testimony before the Subcommittee on Personnel of the Senate Armed Services Committee, CBO analyzed the size and cost of DoD’s overall workforce of service contractors, but that analysis did not examine potential savings from insourcing.15
Question. If CBO is capable of conducting this analysis, would CBO release any findings in the next calendar year?
Answer. Depending on Congressional interest, the availability of data, and when work began, CBO could, in 2027, complete a qualitative analysis of the effects of insourcing specified portions of DoD’s workforce.
1. Testimony of Phillip L. Swagel, Director, Congressional Budget Office, before the Subcommittee on Fiscal Responsibility and Economic Growth of the Senate Committee on Finance, The Budget and Economic Outlook: 2026 to 2036 (March 11, 2026), www.cbo.gov/publication/62154.
2. Unless this document indicates otherwise, all years referred to in describing budget projections are federal fiscal years, which run from October 1 to September 30 and are designated by the calendar year in which they end.
3. Congressional Budget Office, The Budget and Economic Outlook: 2026 to 2036 (February 2026), www.cbo.gov/publication/61882.
4. Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, 2025 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds (June 2025), www.cms.gov/oact/tr/2025.
5. For that most recent estimate, see Congressional Budget Office, “CBO’s Updated Projections of the Hospital Insurance Trust Fund’s Finances,” CBO Blog (February 23, 2026), www.cbo.gov/publication/62165.
6. The Medicare trustees report “average incurred per beneficiary costs” in calendar years.
7. Jagadeesh Gokhale and Kent Smetters, “When Does Federal Debt Reach Unsustainable Levels?” (Penn Wharton Budget Model, October 6, 2023), https://tinyurl.com/yeyjbmnj.
8. See, for example, Gita Gopinath and Brent Neiman, The Incidence of Tariffs: Rates and Reality, NBER Working Paper 34620 (National Bureau of Economic Research, January 2026), www.nber.org/papers/w34620; and Julian Hinz and others, America’s Own Goal: Who Pays the Tariffs? Kiel Policy Brief 201 (Kiel Institute for the World Economy, January 2026), https://tinyurl.com/yztf3a56.
9. Congressional Budget Office, The Budget and Economic Outlook: 2026 to 2036 (February 2026), www.cbo.gov/publication/61882.
10. Pablo Fajgelbaum and Amit Khandelwal, “Tariffs in 2025: Short-Run Impacts on the U.S. Economy” (paper presented at the Spring 2007 Brookings Papers on Economic Activity Conference, March 26–27, 2026), https://tinyurl.com/586ay8dd; and Mary Amiti and others, “Who Is Paying for the 2025 U.S. Tariffs?” Liberty Street Economics (blog entry, Federal Reserve Bank of New York, February 12, 2026), https://doi.org/10.59576/lse.20260212. Other recent work studying changes in tariffs in 2018 and 2019 indicates that the share of the cost absorbed by foreign exporters (40 percent) may have been higher than previously thought, when accounting for quantity discounts and other related factors. That work does not analyze changes in tariffs in 2025, however. See Sharat Ganapati and Colin J. Hottman, Did Foreigners Pay America’s Tariffs? Quantity Discounts, Scale Economies and Incomplete Pass-Through, NBER Working Paper 34901 (National Bureau of Economic Research, February 2026), www.nber.org/papers/w34901.
11. Robert B. Barsky and Lutz Kilian, “Do We Really Know That Oil Caused the Great Stagflation? A Monetary Alternative,” NBER Macroeconomics Annual, vol. 16 (2001), pp. 137–183, https://doi.org/10.1086/654439.
12. See, for example, Samantha Gross, “The Iran Conflict’s Energy Shocks Are Not Yet Fully Realized” (Brookings Institution, April 1, 2026), https://tinyurl.com/ycyjumm6; United Nations Conference on Trade and Development, Strait of Hormuz Disruptions: Growth and Financial Implications (April 1, 2026), https://tinyurl.com/58yw9pbh; and Lutz Kilian, Michael Plante, and Alexander W. Richter, “What the Closure of the Strait of Hormuz Means for the Global Economy,” Dallas Fed Economics (blog entry, Federal Reserve Bank of Dallas, March 20, 2026), www.dallasfed.org/research/economics/2026/0320.
13. Executive Order 14186, “The Iron Dome for America,” 90 Fed. Reg. 8767 (January 27, 2025), https://tinyurl.com/mryx7t37.
14. Congressional Budget Office, Replacing Military Personnel in Support Positions With Civilian Employees (December 2015), www.cbo.gov/publication/51012.
15. Testimony of David E. Mosher, Director of National Security Analysis, Congressional Budget Office, before the Subcommittee on Personnel of the Senate Committee on Armed Services, Approaches to Reducing the Department of Defense’s Compensation Costs (July 26, 2023), www.cbo.gov/publication/59375.