Sections

Commentary

The forces affecting America’s fiscal future: Introduction and overview

Shutterstock / JOAT

Introduction

The U.S. government faces an unsustainable fiscal trajectory under current policies. The federal debt, measured as a share of gross domestic product (GDP), was just under 100% at the end of 2025. That is just shy of the historic peak of 106% at the end of World War II. Back then, fiscal restraint, rapid economic growth, interest-rate caps, and other factors pushed that ratio down sharply over the following three decades. Today, however, debt is projected to rise significantly and inexorably. The Congressional Budget Office (2026) projects that the debt-to-GDP ratio will rise to 175% by 2056 if no policies change. Auerbach and Gale (2026) show that under scenarios that maintain spending and extend temporary tax provisions, the ratio could rise as high as 211% over the next 30 years. In both sets of forecasts, there is no natural turning point, and the ratio would be expected to continue to rise thereafter.

The sources of the fiscal problem are many. Economists have known for decades that population aging, including the aging of the baby boom generation, would put enormous pressure on Social Security and Medicare spending. Gradually rising costs of healthcare have added to that spending burden. On the revenue side, several major tax cuts over the past 25 years, made permanent in recent legislation, have substantially reduced revenues. The Global Financial Crisis of 2007 to 2009 and the COVID-19 pandemic, along with the associated federal policy responses, led to substantially higher public debt. Recent increases in interest rates, combined with higher debt, mean that federal interest payments are projected to rise much more rapidly than GDP.

Rising government debt will crowd out private investment, reducing the capital that workers can use. This in turn reduces their productivity, wages, and standard of living (Friedman 1988, Auerbach 1994). Rising debt also raises the risk of a fiscal crisis (Edelberg, Harris, and Sheiner 2025), potentially leading to a sharp increase in interest rates and a disruptive fiscal adjustment. In the international political arena, high debt can erode U.S. power and economic position relative to other nations.

Addressing the nation’s fiscal imbalances is essential. To this end, the Urban-Brookings Tax Policy Center, Brookings’ Hutchins Center on Fiscal and Monetary Policy, and the University of Utah’s Marriner S. Eccles Institute for Economics and Quantitative Analysis convened a conference on “The Factors that Affect America’s Fiscal Future,” at the University of Utah on June 24 to 26, 2026. The conference brought together more than 50 researchers, including key experts from the academic community, government, and think tanks. Authors were asked to examine the factors driving the rise in debt and ways to improve the outlook.

This volume includes papers prepared for that conference. They are organized in three groups. Several papers address external factors that affect the budget outlook, including the rise and diffusion of artificial intelligence, and demographic changes. Other papers address major spending priorities—healthcare, energy and the environment, and defense. Finally, two papers examine revenue issues—one proposing business tax reform and the other exploring how earmarking taxes could increase support for revenues.

Two major themes emerged from the papers and conference discussions. First, although the fiscal situation is daunting, policy changes of a magnitude well within historical experience could put the debt on a sustainable path. The second theme was the importance of how revenues are raised and spending is contained. Budget solutions should not focus just on bean counting; they should take the opportunity to incorporate broader policy objectives. It would be a Pyrrhic victory to stabilize the debt at the cost of unnecessarily damaging living standards or weakening the U.S. position in the world.

Desirable reforms would improve the budget outlook as well as reach other policy goals. For example, the structure of healthcare reimbursement can reduce spending while encouraging research into the most promising medical technologies. Energy policy can affect revenues and spending, as well as direct resources to uses that produce minimal pollution. Business tax reform can raise revenues and make taxes more efficient and equitable. Earmarking tax revenues can build political support for tax increases by linking them to spending on popular programs.

External factors

The rapidly expanding—and still largely unknown—artificial intelligence frontier could profoundly affect the economy and the fiscal outlook.

In Artificial intelligence and the economy, Benjamin F. Jones presents a framework for understanding AI’s economic role. He highlights three aspects of AI’s role in producing output—the share of tasks AI performs, its productivity at those tasks, and the strength of bottlenecks (the parts of production that remain difficult to automate). A key result, given what we know about substitution between capital and labor, is that “bottlenecks are stronger than infinities.” Because different tasks are not perfectly substitutable in production, even vast improvements in productivity on some tasks have limited benefits for overall productivity if other tasks remain unchanged. For example, over the past 50 years, a 17-order-of-magnitude increase in computing power has accompanied only a tripling of overall productivity. A corollary is that AI may affect the economy most rapidly by expanding the tasks it can perform rather than taking a selected few to infinite productivity.

A common concern with technological change is its impact on income distribution.

Jones finds that expanding the tasks AI performs raises the capital share. But improvements in automation can raise the labor share by raising the productivity of workers performing other tasks. With past innovations, he argues, these factors have roughly balanced, keeping labor and capital shares relatively constant as technology improved.

Past technological improvements have tended to replace lower-skill work. AI, by contrast, may be particularly effective at performing tasks traditionally done by higher-skilled workers, such as coding. AI’s effects on the shares of income going to capital and labor, as well as on the distribution of earnings across workers, have important implications for the fiscal outlook and the policy choices discussed in the remaining papers.

In How might fiscal policy respond to the rise of artificial intelligence?, Karen Dynan, Doug Elmendorf, and Louise Sheiner construct four scenarios for AI’s long-term impact on the federal budget and economic outcomes, using CBO projections and outside sources. The scenarios illustrate a range of possible outcomes rather than predict the future. For each, the authors estimate the effects on federal debt in 2056 and consider possible policy responses.

In the first three scenarios, AI raises annual total factor productivity (TFP) growth by 0.5 percentage points above CBO’s baseline of 1.1% over the next three decades. In the first scenario, unemployment and labor force participation are unchanged, and AI-driven income gains are proportional to workers’ and capital owners’ current earnings. Federal debt reaches 136% of GDP in 2056, compared with CBO’s baseline projection of 175%.

In the second scenario, all additional income accrues to households in the top fifth of the income distribution. Debt in 2056 is 126% of GDP, lower than in the first scenario, because the wealthy pay higher tax rates. But inequality would rise sharply, as incomes for the bottom 80% would remain at the CBO baseline.

The third scenario adds AI-driven job losses, with unemployment rising and labor force participation falling by 1 percentage point each. Debt is the same as in the second scenario.

The fourth scenario doubles the productivity boost to 1 percentage point, while unemployment rises by 2 percentage points and labor force participation falls by 4 percentage points. Yet debt still falls only to 126% of GDP in 2056, the same as in the second and third scenarios, because the additional income goes to capital owners, whose income is taxed relatively lightly.

The scenarios suggest different policy responses. The first scenario requires little policy adjustment since the gains from AI are broadly shared. The second scenario improves the fiscal outlook but increases inequality, potentially strengthening the case for redistribution. A natural response to the third scenario, in which workers are displaced, would be to expand unemployment insurance, wage insurance, and job training. In the fourth, policymakers may need to foster employment opportunities, consider a universal basic income, and increase taxes on capital. Dynan, Elmendorf, and Sheiner conclude that fiscal policies “robust to [all four] different scenarios would be especially valuable,” particularly those that expand social insurance, foster employment, and tax capital at higher rates.

Louise Sheiner and David N. Weil, in Demographics and the fiscal challenge, evaluate the role of aging, fertility, and immigration in shaping the long-term fiscal outlook. They calculate that population aging—both over the past 20 years and in the decades ahead—contributes substantially to the projected rise in the debt-to-GDP ratio by increasing spending on retiree programs. They then ask how changes in fertility and immigration could affect fiscal trends. Even large increases in fertility, they find, will not help the fiscal outlook for a very long time. They document a “crossing effect”: higher fertility initially worsens the fiscal outlook because of increased spending on children but eventually improves it as the larger population enters the workforce and pays taxes. For the federal government, the net effect turns positive after about 35 years. For all levels of government combined, it takes roughly 60 years, reflecting the fact that state and local governments bear much greater costs of children, particularly for education, than the federal government does.

The fiscal benefits of immigration arise much more quickly. Working-age immigrants pay taxes that exceed benefits received as soon as they arrive, and higher immigration rates improve the federal balance sheet across a wide range of scenarios. This effect is especially large for high-skilled immigration, since higher earnings generate more tax revenue without increasing the cost of public services. Although higher fertility and immigration can improve the fiscal outlook, neither can fully substitute for tax increases or spending cuts. Even with 3 million new immigrants a year (high relative to recent experience), debt in 2056 would still be well above today’s levels. The effect is equivalent to a permanent 6% increase in federal revenues—a substantial improvement, but not enough to stabilize the debt.

Spending priorities

Healthcare is a major component of federal spending. David M. Cutler asks, Will medical care strangle or save the federal budget? He observes that healthcare spending grew persistently as a share of GDP for decades before leveling off after 2010. Is this a permanent shift or a temporary slowdown? If future healthcare spending grew at the rates prevailing between 1970 and 2010 (2.4 percentage points faster than GDP annually), it would reach 26% of GDP in 2040, compared with just 19% if it grew at rate prevailing since 2010 (0.3 percentage points faster than GDP growth annually). Even an intermediate scenario, in which healthcare spending reached 22% of GDP by 2040, would put extreme pressure on the federal budget and make private health insurance unaffordable for many families.

Cutler also examines what could drive future spending. High growth could result from expensive new “rescue therapies” and “blockbuster drugs.” Leqembi, for example, slows Alzheimer’s progression but costs more than $25,000 a year. If 5 million people took it annually, the cost would equal about 2.5% of national health expenditures. Slower growth could result from use of minimally invasive surgeries, moving care out of hospitals and nursing homes, better disease prevention, and lower administrative costs.

Cutler calls AI “the biggest unknown on the medical care horizon.” It could save money by automating paperwork, helping patients choose cheaper care, and speeding the development of new therapies. But it could also raise spending by customizing diagnoses or helping providers increase reimbursements through more aggressive billing practices. Ultimately, Cutler concludes, AI’s effect on spending will depend on human decisions, including reimbursement policies, regulation, and patient and provider trust.

Joseph Aldy, Jun Gao, and Kenneth T. Gillingham explore energy and environmental policy in Tax policy and power sector decarbonization. They argue that a carbon tax is the best way to encourage renewable energy, because it puts a price on carbon emissions that reflects the damage they cause, rather than rewarding particular technologies.

But a carbon tax would raise electricity prices and face steep political obstacles, so the United States has instead relied on tax credits and other subsidies to encourage renewable energy. In recent years, wind and solar generating capacity has expanded dramatically, while costs have fallen below the costs of coal and natural gas. While a carbon tax raises electricity prices and generates government revenue, subsidies tend to lower electricity prices at a cost to the federal budget. For example, energy tax expenditures have run about $10 billion to $20 billion annually over the past decade, and the Inflation Reduction Act’s tax credits were projected to exceed $1 trillion over a decade before being curtailed by the One Big Beautiful Bill Act in 2025. Moreover, subsidies do less to encourage conservation or reductions in emissions from existing fossil-fuel plants.

The authors propose reforms to make renewable energy subsidies more cost effective. They favor tax credits tied to electricity generation rather than investment spending, citing evidence that production tax credits delivered wind power at about 28% lower cost than investment-based grants. They also recommend targeting subsidies toward projects that reduce emissions the most, for example, by encouraging renewable energy development in areas where electricity is currently generated by particularly polluting power plants.

In addition, they identify regulatory barriers that raise the costs of renewable energy projects, including delays in connecting to the electricity grid, permitting requirements, and inefficient reliability standards. Reducing these barriers could encourage more renewable energy investment while allowing the government to provide smaller subsidies per project.

Such reforms could become increasingly important as electricity demand rises with the expansion of AI and data centers. Together, better-targeted tax credits and regulatory reforms could help reduce emissions at a lower cost to the federal budget.

Defense spending is sometimes an overlooked part of the fiscal landscape. In The defense budget: Costs and choices, Mark Cancian examines the budgetary implications of decisions driven primarily by military and diplomatic considerations. Defense spending has fallen from 51% of federal outlays in 1955 to 12% today, but it remains the largest discretionary spending program. Several factors put upward pressure on spending: changing perceptions of threats from China, Russia, Iran, and North Korea; the rising cost of successive generations of weapons systems; and the struggle to fill ranks from a small recruiting pool (only 12% of Americans between the ages of 17 and 24 were qualified and available to serve as of 2020).

Cancian does not argue for cutting defense spending. Rather, he presents a menu of specific spending reductions across areas including military strategy, compensation, weapons acquisition, personnel, and infrastructure that could contribute to fiscal restraint. Two of the largest potential savings come from adopting a strategy of restraint, which could save about $210 billion a year but would slow potential responses to a conflict over Taiwan or in Europe, and means-testing veterans’ disability compensation, which could save up to $50 billion annually.

Cancian emphasizes that defense cuts require difficult political choices. Cuts to military benefits and weapons programs rarely succeed because of resistance from service members and military institutions. He suggests that commissions modeled on the successful military base closure process could help overcome this resistance by allowing Congress to approve packages of reforms rather than voting on individual cuts. Such an approach might make it easier for defense hawks, deficit hawks, and advocates of domestic spending to reach agreement.

The revenue side

Fiscal consolidation will require both spending restraint and higher revenue. In Improving the destination-based cash-flow tax: Combining a value-added tax with a worker credit, William Gale, Adam Looney, and Elena Patel argue that the leading alternative to the corporate income tax—a destination-based cash-flow tax (DBCFT)—has much to offer. The DBCFT would tax business cash flow—domestic sales less expenditures on wages, materials, and investment—rather than profits, with border adjustments taxing imports and exempting exports. It would remove distortions related to investment, financing, organizational form, and the location of profits and production. But the resulting appreciation of the dollar would create a one-time $11 trillion transfer to foreign wealth holders.

The authors propose an alternative that retains the economic advantages of the DBCFT while reducing transfers to foreigners: a subtraction-method, border-adjusted value-added tax (VAT), combined with a refundable tax credit for workers. The key difference is the location of the subsidy for wages—the DBCFT allows firms to deduct wages, while the authors’ proposal taxes wages under the VAT but compensates workers directly through the credit. This seemingly small change alters the natural adjustment path for the economy. Under the VAT, the natural adjustment would occur through an increase in domestic prices, rather than an appreciation of the dollar. The refundable worker credit would compensate U.S. workers for the one-time price increase. The reform would reduce the transfer to foreigners by about $4 trillion. Additional transition rules could further reduce or almost eliminate the transfer.

The authors note that the Federal Reserve should accommodate the one-time price increase, as central bankers have done in other countries that adopted a VAT, and that doing so would avoid a recession. They estimate that the proposal could raise about $390 billion more annually than the corporate income tax it replaces, roughly 1.3% of GDP.

Jean Baptiste Colbert, King Louis XIV’s public finance minister, is often quoted as saying, “The act of taxation consists in so plucking the goose as to procure the largest quantity of feathers with the least possible amount of squealing.” In a democracy, however, the “goose” must first be convinced it wants to be plucked.

In Convincing the goose: Framing and earmarking taxes to help reduce the federal debt, Ian Berlin, Ellis Chen, Vanessa Williamson, and William Gale examine how earmarking—linking revenues from particular taxes to spending on particular programs—could increase public and political support for raising taxes. Dedicated revenue programs already account for one-third of federal revenues. Polls show greater support for tax cuts posed in isolation than when respondents are told the cuts would require reductions in government spending. The authors also cite evidence that earmarking revenues for popular programs increases support for tax increases.

But public support alone is not enough. Congress must be persuaded to act. Earmarking a tax, the authors argue, can create constituencies eager to defend it because they support the program it funds. These constituencies can, in turn, give legislators political support for enacting and maintaining the tax.

Two case studies illustrate these arguments. In Kansas, the legislature reversed major tax cuts following voters’ disapproval of the resulting spending cuts. In Colorado, voters approved earmarked tax increases for paid family leave and school meals while separately voting to cut income taxes. The authors caution that earmarks create some concerns of their own, but conclude that earmarking new taxes for popular existing programs could still help raise revenues while freeing up other resources for deficit reduction.

 

Authors

  • Acknowledgements and disclosures

    Gale and Patel are affiliated with Brookings and the Urban-Brookings Tax Policy Center. Looney is affiliated with the University of Utah and the Urban-Brookings Tax Policy Center. Sheiner and Wessel are affiliated with Brookings and the Hutchins Center for Fiscal and Monetary Policy. The authors thank Sona Wyse and David Skidmore for outstanding research support and editing, respectively, and the University of Utah and the California Community Foundation for generous financial support.

  • Footnotes
    1. For recent views on the fiscal situation, see Auerbach and Gale (2025), Auerbach and Yagan (2024), Congressional Budget Office (2026), Dynan and Elmendorf (2025), Edelberg, Harris, and Sheiner (2025), Furman (2024), Riedl (2025), and Summers (2023).
    2. Although Social Security is an important part of the fiscal outlook, the conference did not focus on that program because it has already attracted numerous analyses and proposed solutions.

The Brookings Institution is committed to quality, independence, and impact.
We are supported by a diverse array of funders. In line with our values and policies, each Brookings publication represents the sole views of its author(s).