Introduction and summary
This century, debt began to grow faster than the economy—as measured by the gross domestic product (GDP)—driven by tax cuts that disproportionately went to the richest Americans. While the ratio of debt to GDP has been on track to grow indefinitely since the George W. Bush tax cuts of 2001, the fiscal gap—the amount of tax increases and/or program cuts it would take to stop the debt-to-GDP ratio from rising—has meaningfully widened in the past decade.1
The first and second rounds of the Trump tax cuts have since significantly increased the fiscal gap and have led to historically large primary—or noninterest—deficits outside of times of war and economic recession.2 Interest rates have also partially rebounded from their historic lows in the more than a decade following the Great Recession.3 With large primary deficits and higher interest rates, the debt ratio is now projected to climb indefinitely at a significantly faster pace than was previously projected. Importantly, this is now happening even during periods of low unemployment and sustained wage growth. While fiscal stimulus and rising debt are unquestionably necessary when the economy is doing badly, during good economic times, the debt ratio should stabilize and then decline, as it did for most of the latter half of the 20th century. In other words, policymakers should pursue a small primary surplus—though only if it can be achieved responsibly, without further burdening the most vulnerable Americans.
In addition, this report looks at the effect of potentially higher productivity growth driven by artificial intelligence (AI) on the nation’s fiscal trajectory. Based on the available evidence, while such optimism may eventually prove warranted, that is not the case at present. For now, and for the foreseeable future, the most responsible path forward regarding AI’s impact on the public debt is to hope for the best and plan for the worst.
Warnings about forthcoming fiscal unsustainability have not led policymakers to act to improve the outlook. To the contrary, compelling evidence, not least of which is the deteriorating outlook itself, suggests that Congress is increasingly unresponsive to such warnings.4 This report offers thoughts and ideas about metrics and strategies to reawaken policymakers’ fiscal reaction function.
The fiscal situation is notably worse than it used to be, even a decade ago
The Bush tax cuts created the modern fiscal gap.5 Before their enactment, even relative to a baseline that first assumed a permanently indexed alternative minimum tax (AMT),6 revenue was on track to keep pace with primary spending indefinitely, despite the aging population and per-person health care costs that routinely grew faster than the economy.7
Even after the overwhelming majority of the Bush tax cuts were made permanent on a bipartisan basis during the Obama administration, many lawmakers and experts comforted themselves about the debt ratio for two major reasons: 1) primary deficits were small enough that course-correcting—reducing the fiscal gap—would be relatively painless, and 2) interest rates following the Great Recession were low both in historical terms and relative to the GDP growth rate, a pattern that mitigates the growth of debt.8
These once relatively fiscally benign factors have changed considerably. With one exception, structural primary deficits are now at historically high levels, surpassed only by prior periods of deep recession or war—President Reagan’s structural primary deficits, the largest in history, which were caused by the largest tax cuts in history, prompted 15 years of sustained bipartisan fiscal consolidation.9 And interest rates have meaningfully rebounded from their post-Great Recession lows.10 While interest rates are not currently higher than economic growth, the historically large gap the United States experienced for more than a decade has significantly narrowed, and rates are expected to continue rising before eventually overtaking growth within three years.11
Even when debt was at 76 percent of GDP in 2016, interest payments net of U.S. Federal Reserve remittances as a percentage of GDP were the lowest since 1918.12 That is because rates were historically low. However, even with interest rates currently no higher than they were in 2008, the historically large stock of debt, combined with the partial rebound in interest rates, means there are historically large interest payments, with payments set to continue to rise indefinitely. Whether one looks at interest payments or debt as a percentage of GDP, the United States is on track to reach historic levels. Over the long run, examining interest costs and examining debt are largely identical ways to see the same budgetary issue: While interest rates might bounce around along the way, if the debt ratio grows indefinitely, so will interest payments, and if debt is shrinking indefinitely, so will interest payments.
Moreover, primary deficits are now projected to be roughly 3 percent of GDP indefinitely, higher than at any point in U.S. history outside of a war or recession. (see Figure 1)
Primary deficits and higher interest rates matter because only three factors determine debt ratio stability: 1) the size of the primary deficit, 2) the difference between the growth rate of the economy (which economists abbreviate as g) and the compound interest rate on all federal debt r, and 3) the level of debt as a percentage of GDP. As discussed in greater detail below, both the higher primary deficits and higher interest rates the nation is now experiencing and projected to bear in the future make debt ratio stability substantially more difficult than it was a decade ago, while the level of the debt ratio amplifies the impact of g minus r.
What is the problem with ever-rising debt?
Economists believe that government debt crowds out private investment over the long run because more capital will be used to finance government borrowing rather than new businesses.13 But when less capital goes to private investment, there is lower productivity growth than there otherwise would have been and therefore lower wage growth for American households than there otherwise would have been.
For example, the Congressional Budget Office (CBO) projected last year that, between 2025 and 2055, real gross national product (GNP) per capita will grow by 47 percent.14 But the CBO also projected that, if the debt ratio were to stabilize instead of growing forever, real GNP per capita would instead grow by 52 percent by 2055. In dollar terms, this means the CBO believes the upward trajectory of the debt ratio will leave real GNP per capita roughly $4,600 (in 2025 dollars) lower in 2055 than it otherwise would have been. The authors estimate that this will leave average wages for American workers roughly $4,200 lower than they would have been absent the increase in the debt ratio—noting that this is an average result and does not account for how average wage growth is distributed. This is significantly larger than the average tax cut in the Big Beautiful Bill.15
It is worth stating explicitly that, even under these projections, real wages are estimated to continue growing substantially. While projected debt increases are estimated to lower Americans’ wages by 3.4 percent, or $4,200, on average in 2055 relative to where they would otherwise have been, real wages are, on average, still projected to grow by 45 percent or $37,350 (instead of by 50 percent, or $41,550, if the debt ratio were instead stabilized).16 In other words, the current debt trajectory is estimated to make it so wages grow by 45 percent by 2055, instead of by 50 percent. Future generations are better able to handle income losses than current generations are because they are, on average, richer. Of course, “good debt” that fights recessions or that funds such things as investments in productive infrastructure or research and development is worthwhile and generates returns that help boost growth and offset its fiscal cost. However, relative to a stable debt ratio scenario, the percentage loss to wages grows forever. In other words, the wedge between where wages would have been were the debt ratio stable and where they would be when the debt ratio rises forever widens. In this regard, Americans will be better off in the future if Congress can figure out a responsible way to stabilize the debt ratio, and the sooner it is done, the larger the benefit. Being responsible means not hurting vulnerable Americans by cutting the services upon which they rely. It also means not compromising long-term investments in the future such as education funding, biomedical research, and other cutting-edge research and development.
It is crucial that the measures Congress takes do not do more harm than good. A direct benefit of stabilizing the debt ratio is higher wages than workers would otherwise have. But if, in achieving that, Congress significantly reduces aid for vulnerable Americans, it has done more harm than good. For instance, if Congress were to kick people off Medicaid to reduce the debt, those people’s lives would become a lot more precarious than they would otherwise be. Even if their incomes are 3.5 percent higher than they otherwise would have been—and Medicaid recipients, almost by definition, are those least likely to see rising real wages redound to them—people are not better off if they lose crucial assistance. People are not better off if they have earned 3.5 percent more but lost their health care, given that the median Medicaid spending per enrollee was $7,909 in 2023.17
In particular, progressives should care about stabilizing the debt ratio because of the very real risk of overreaction by either markets or Congress. While there is no risk that the United States will be unable to pay its creditors—short of deliberate sabotage by Congress’ refusal to raise the debt limit—there is a risk that the current unsustainable fiscal outlook leads investors who buy U.S. debt to demand extra compensation to hold that debt. This has happened in other countries, most notably the United Kingdom.18 While the dominant position of the U.S. dollar in global markets and the size, depth, and liquidity of the market for U.S. debt provides the United States with some insulation, it does not render the nation immune to negative market judgments.
Should such a market reaction occur, it would in turn raise the likelihood that Congress pursues irresponsible fiscal consolidation, further cutting programs on which the poorest Americans rely and important investments for the future. Finding ways to address the country’s fiscal gap responsibly somewhat decreases the likelihood of deep cuts to vital programs that support the most vulnerable Americans.
Yet another reason to care about debt is that higher interest rates lower housing construction—another factor, in addition to higher mortgage rates, making housing less affordable. Residential construction—both single and multifamily—is one of the forms of investment that is most sensitive to interest rates since more of it is financed with debt than business investment. The CBO says that a “1 percent increase in mortgage rates would have a proportionately larger impact on residential investment than a 1 percent increase in corporate bond rates would have on businesses’ purchases of computers.”19 The CBO’s models assume that a 2 percentage point increase in mortgage rates would reduce housing starts by around 400,000 over the next three years—housing starts would gradually drift back toward baseline levels. Still, the number of houses would be permanently smaller.20 The smaller supply of houses increases rents and the cost of homeownership relative to a scenario of higher supply.
Finally, the nation’s rising debt may create a debt spiral that would be increasingly difficult to break out of. A rising debt ratio pushes up r and pushes down g, and, as g minus r becomes smaller and eventually becomes negative, debt stabilization will become more difficult; the larger g minus r is, the larger primary deficit can be run while still stabilizing the debt ratio. As g minus r shrinks, that fiscal benefit shrinks, and when it becomes negative, the nation will need to start running primary surpluses—of an increasingly large size as r becomes larger and larger than g—making debt stabilization more difficult. This means that, as debt rises, all else being equal, even if primary deficits remained constant as a percentage of GDP, debt stabilization would grow increasingly difficult.
However, the problem is exacerbated by the growing stock of debt because the level of the debt ratio amplifies the impact of g minus r. When debt is high and g is larger than r, debt becomes even easier to stabilize for that given level of g minus r. But when debt is high and g is smaller than r, debt becomes harder to stabilize for that given level of g minus r. So, because the growing U.S. debt ratio is pushing the country toward r being larger than g and this high level of debt-to-GDP itself amplifies the impact of r being larger than g, an ever-growing debt ratio feeds back into itself in two separate ways. The current path of an ever-rising debt ratio will eventually create a debt spiral that is very difficult to break out of. At the point where a high debt ratio has pushed r above g, the high debt ratio will itself make course correction meaningfully more difficult than it would have been. (see Table 1)
The long history of declining r reinforced the view among analysts that g minus r was necessarily positive and would stay positive in the long run, so growing debt was less of a problem than it might seem. But the idea that r is going to drop back down to where it was for the 14 years following the Great Recession seems less persuasive these days. For one, as Jared Bernstein and former Federal Reserve Chair and Treasury Secretary Janet Yellen recently argued, extensive capital investment around AI is putting upward pressure on the cost of capital.21 The risk that the medium-to-long-term g minus r is zero rather than positive has clearly increased. This, therefore, increases the concern relative to what a reasonable person would have felt in the past decade. At a minimum, it suggests that ever-rising debt entails more downside risk than economists and budgeteers previously thought.
The fiscal gap—the average amount of primary deficit reduction needed to keep the debt ratio from rising—is now sufficiently large that course correction is quite difficult.22 Assuming the temporary tax cuts and spending increases in the Big Beautiful Bill are made permanent and that only one-third of the revenues from the new tariffs imposed by President Trump remain after his administration ends, primary deficits would need to be reduced by 2.6 percent of GDP on average to stabilize the debt ratio. In 2026 terms, this would be roughly $840 billion. Over 10 years, legislation would need to reduce primary deficits by $10.5 trillion. To put that figure in perspective, the sum total of all taxes House Democrats proposed in Build Back Better totals only $2 trillion, after adjusting for changes in the size of the economy, and the historic tax cuts congressional Republicans and President Trump enacted in the Big Beautiful Bill totaled $5.8 trillion, after adjusting for changes in the size of the economy, or $6.6 trillion also assuming that the temporary provisions are made permanent.
Policymakers’ goal should not just be to close the fiscal gap but instead to slightly overshoot it. While the government should deliberately run large and growing deficits during bad economic times to boost aggregate demand, structural deficits—what the deficit would be when economic growth is good—should be sufficiently small so that the debt ratio declines somewhat during good economic times. This would achieve a similar outcome to lowering total deficits, including interest payments, to 3 percent of GDP, as some debt analysts have recommended.
It should also be emphasized that the current fiscal projections assume historically modest interest rates. In the CBO’s 2026 long-term budget outlook, the simple rate of interest the government pays on its outstanding debt never gets above 4.1 percent, last seen in 2008. In other words, this is the outlook even with relatively low interest rates. (see Figure 2) If interest rates were 1 percentage point higher than the CBO projects, the country’s fiscal outlook would be significantly worse. Of course, the outlook may instead be better than the CBO projects. But the effect of higher or lower interest rates on the fiscal outlook is larger when debt itself is high. That itself is a reason to prefer a lower debt ratio.
Due to data limitations, the figures in this report reflect projections made by the CBO in its 2025 long-term budget outlook, as adjusted by the authors, rather than projections made in 2026 with similar adjustments made. If the authors had used the projections made in 2026, the fiscal outlook would look even worse, due in large part to interest rate projections being more than 50 basis points higher in the long run. (see Figure 2)
Policymakers should not rely on AI-driven productivity growth to solve the problem
Can AI boost productivity enough to stabilize the debt, obviating the need for the fiscal discipline advocated above? The economy’s growth rate is the product of productivity growth and the growth rate of aggregate hours worked—that is to say, labor supply times how efficiently the workforce turns inputs into outputs. Should AI significantly increase productivity growth, the result would be faster GDP growth and, all else equal, lower debt ratios going forward.23
This hope raises several questions, including: What is a realistic expectation for the extent to which AI might raise productivity? Will that be enough to put the nation on a more sustainable path? And will AI’s productivity-growth impact persist such that the country can stay on that path? With the strong caveat that this is highly speculative—though much has been said and/or postulated about AI’s impact, very little is known—this section tackles those questions.
Summarizing the findings, it is possible that AI will boost productivity, and thereby GDP growth, by enough to stabilize and perhaps even reduce the debt ratio within the budget window. However, to do so, productivity would have to grow at rates well above recent historical averages. Moreover, along with the risk of betting on an outsize forecast, there are two other risks. First, while AI will certainly raise the level of productivity, it is less certain that AI will have an impact on the long-term growth rate. If that concern is borne out, then the debt trajectory will, after temporarily stabilizing, eventually return to the unsustainable path documented above. Second, should AI boost productivity growth enough to improve the fiscal trajectory significantly, it could be quite disruptive to jobs and incomes of many Americans, a scenario that would require new, potentially extensive government spending to offset.
What is the potential impact of artificial intelligence on productivity?
There is an active debate among economists and those in the technology sector over whether AI will augment labor—improving the productivity of incumbent workers—or displace labor. According to a 2026 report from Goldman Sachs Research, workplace adoption, defined as firms using AI in “their regular business functions,” is just under 20 percent and rising at a good clip.24 At this point, however, there is not much evidence on either side of the debate. The key insight for fiscal analysis is that either way—whether AI makes workers more productive or replaces them—output per hour goes up.
While the expectation that output per hour will increase is widely shared among economists, there is a wide range of predictions about its magnitude. Tom Cunningham of Model Evaluation and Threat Research provides a review of recent research on forecasts of AI’s growth effects, noting that economists’ predictions tend to be considerably lower than those of AI industry experts.25 Reviewing estimates of how much faster productivity might grow over the next decade or beyond, he reports that “most economists’ forecasts are 0.1–1.5%/year,” while the predictions of most “AI insiders” are much higher than that. The difference, according to Cunningham, stems from the insiders’ greater confidence about AI’s future capabilities.
The Budget Lab at Yale recently explored this very question about the impact of AI-powered faster productivity growth on the U.S. fiscal outlook.26 Figure 5 below shows four debt-ratio paths from that study, including the latest CBO baseline along with three scenarios based on different assumptions about the extent to which AI increases productivity growth.27 A 2026 study by Ezra Karger et al. collected and analyzed a broad survey of opinions and forecasts to estimate that productivity growth will increase 2.5 percent annually, as opposed to a bit below 2 percent, the average of the trend over the past decade.28 The Yale Budget Lab looked at the effect on debt of these scenarios as well as faster (3.2 percent) and slower (2 percent) growth scenarios.
Figure 5 below reveals that under the Yale Budget Lab’s model, Karger et al.’s medium productivity growth assumption leads to debt stabilization at about 105 percent of GDP, while the 3.2 percent productivity growth assumption leads to a downward debt trajectory by the end of the window.
Using the CBO’s debt simulation model produces similar results.29
The debt paths in Figure 5 above appear plausible, given the models’ assumptions. But are the assumptions themselves credible? The Yale Budget Lab study offers numerous thoughtful tweaks to their model that partially offset the debt-reduction impacts of AI-induced faster productivity growth, including diminished labor supply due to job displacement and higher government spending made necessary by the loss of labor income from those displacements. Under these scenarios, the debt ratio tends to grow more slowly than the CBO baseline, but it does not stabilize within the budget window, except in the high-productivity scenario.
However, the biggest challenge to the “AI-to-the-rescue” scenario pertains to the productivity growth rate itself. While forecasters’ optimism that AI will boost labor productivity is warranted, the key question for sustainability dynamics is whether that elevated growth rate will persist. Because debt sustainability relies in part on growth rates—and not level shifts—it is not enough for AI to improve workers’ productivity as it is adopted. To boost the growth rate over long enough periods to change the debt trajectory, it must continually improve workers’ productivity.
To make this somewhat confusing distinction concrete, consider a factory wherein 5 people make 10 widgets. Output per person is 2 widgets—the total of 10 is divided by 5 workers. Productivity is typically measured as output per hour versus per person, but that does not matter here. Now imagine AI enables the workers to increase their output by half, to 15 widgets, boosting their productivity level to 3 (15 divided by 5). Productivity growth is 50 percent, from 2 to 3. Now, assume the 5 workers keep using AI to produce 15 widgets. Their productivity level stays at 3, and thus their productivity growth rate does not rise further. For continued productivity growth, AI would have to make them increasingly productive.
This distinction is not theoretical. Figure 6 below shows yearly productivity growth in the nonfarm business sector, along with a smooth trend to pull the growth signal from the noisy underlying series. In the 1990s, internet adoption significantly raised the growth rate, generating considerable excitement about a new era of technology-induced faster growth. However, after peaking around the year 2000, productivity growth returned to its pre-internet rate. Of course, the internet did not go away. But it turned out to have a greater long-term impact on the level of productivity than on its growth rate.
For this technological change to stabilize or shrink the debt ratio in the long run, it would have to not just boost the level of productivity but boost its long-term growth rate as well.
When it comes to AI’s impact on the fiscal sustainability outlook, therefore, one should hope for the best and plan for the worst. For this technological change to stabilize or shrink the debt ratio in the long run, it would have to not just boost the level of productivity but boost its long-term growth rate as well, and it would have to do so without requiring significant new government spending to offset its labor market impacts.
The death of Congress’ ‘reaction function’ and what can be done to revive it
To state the obvious, policymakers have been unmotivated to act against the nation’s unsustainable fiscal path in sufficient measure to close the fiscal gap. In fact, recent research has shown that in the decades before the last budget surplus occurred in the latter 1990s and early 2000s, Congress reacted to the CBO’s forecasts of higher deficits by changing policy to at least slightly reduce the forthcoming imbalances. But since then, Congress’ “reaction function” to analysis such as that discussed above—more importantly, to the CBO’s analysis—has been nonexistent.30
One problem is that policymakers heavily and increasingly discount the future. Given the incentives they face, it is politically rational for them to provide deficit-financed “goodies” to their donor base or constituents rather than to insist they accept cuts to services or higher taxes to stave off higher interest rates and debt-service crowding out down the road.31
While it is important to tell and show the story of the country’s challenging fiscal path, a quarter century of graphs showing debt ratio projections heading north have not moved the political system to act. To the contrary, Congress has become even less responsive over time. Clearly, analysts need to think of new ways to more emphatically impress the extent of the costs of fiscal recklessness on policymakers and the public.
Recognizing the aggressive future discounting noted above, the connection between debt and interest rates strikes the authors as relevant. The CBO’s model assumes that each 1 percentage point rise in the federal debt ratio pushes up long-term interest rates by 2 basis points, or 0.02 percentage points.32 Applying that to the rise in the debt ratio since 2001, from 31.5 percent to 99.4 percent of GDP, would imply that interest rates are roughly 1.36 percentage points higher now than they would otherwise be absent that rise in the debt ratio.
This, then, must be put in more concrete terms. For example, it is well understood that yields on U.S. Treasuries are highly correlated with the 30-year fixed-rate mortgage. For a home that costs $400,000—about the current median home price—if all $400,000 is owed, an increase in the mortgage rate from, for example, 5.5 percent to 6.5 percent would increase the monthly mortgage payment by about $260, or more than $3,000 per year.
Of course, there are many factors beyond public debt that determine interest rates, and the fact that the U.S. dollar is the dominant global reserve currency means creditors from all over the globe will purchase U.S. debt. But while the ongoing affordability crisis is often cast in terms of consumer prices, the interest rate on mortgages, automobiles, small businesses, and home loans is also a highly salient price to consumers. Insofar as higher public debt is contributing to higher rates, fiscal imbalances are relevant to affordability concerns.
Next, Jared Bernstein, Adam Shaw, and Daniel Posthumus have developed a line of research around what they refer to as “ahistorical adjustments.”33 Given the future discounting problem, it is commonly suggested that policymakers will not act on sustainability until they are forced to do so by a prompting event, such as a lasting spike in interest rates driven by creditors insisting on higher term premia. Should that spike occur, fiscal authorities could face the need to very quickly close the fiscal gap, which, as the authors point out, amounts to 2.6 percent of GDP.
In their 2026 analysis, Bernstein, Shaw, and Posthumus noted, “Outside of bounce backs from recessions, there is only one notable move from deficit to surplus of such a large magnitude in the time series: the approximately 6 percentage point increase from 1992 to 2000.”34 But this was a gradual change, one which was facilitated by both policy and luck, as those years saw a capital gains boom; a tax increase on high-income earners in the 1993 budget, on top of the gains from a previous tax increase from the George H. W. Bush administration; the productivity bump shown above; and spending cuts during the period from 1985-2000, including defense cuts facilitated by the end of the Cold War.
If the country had to quickly reduce primary deficits by 2.6 percent of GDP to close the fiscal gap, it would mean deeply painful spending cuts or tax increases with negative growth impacts—occurring by definition, as it is important to remember, at a time of higher interest rates. For example, this would mean raising taxes by 15 percent (more than $5,100 on the average employed individual if borne entirely by them) or cutting program spending by 13 percent—or 57 percent if Social Security, Medicare, Medicaid, the Supplemental Nutrition Assistance Program (SNAP), veterans spending, and defense spending are left untouched, or 23 percent if just Social Security and Medicare are left untouched.
While reductions of this magnitude will seem conspicuously unambitious to those worried about the outlook, Yagan and Auerbach’s approach recognizes a simple truth: When in a deep hole, the first thing one should do is stop digging, even if that requires a substantial redefinition of the hole!
Finally, in a new paper, economists Danny Yagan and Alan Auerbach, the researchers behind the measurement of Congress’ historical reactions to deficit projections, offer a more gradual solution to the problem.35 First, they redefine the sustainability problem from stability in the debt ratio over the 10-year budget window to the minimum deficit reduction needed to create a 95 percent likelihood that the debt ratio stays below 250 percent over the next 100 years. In other words, they are “grading on a curve”—and a steep one at that. But the point is that the budget goal requires only 0.3 percent of GDP in the near term, climbing to 1.7 percent by 2036. While reductions of this magnitude will seem conspicuously unambitious to those worried about the outlook, their approach recognizes a simple truth: When in a deep hole, the first thing one should do is stop digging, even if that requires a substantial redefinition of the hole!
To be clear, this question of how to more effectively elevate the extent of the debt problem is a highly challenging one, and it is not fully answered here. This is an important area for future work, even more so than continuing to document the nation’s fiscal deterioration, as unfortunately, such presentations have had little impact on addressing the fiscal outlook.
Conclusion
In the near term, Congress must rediscover its reaction function and seek responsible ways to lower primary deficits. But in the medium term, Congress should aim to eliminate primary deficits fully and pursue a small primary surplus. Crucially, however, Congress should ensure that any deficit reduction does not harm the poorest or vital public services on which Americans rely. Irresponsible deficit reduction is significantly worse than doing nothing, but doing nothing is worse than responsible deficit reduction. Congress should look first and foremost to undo some of the many tax cuts significantly tilted toward the wealthy, which are responsible for the fiscal gap in the first place.
Acknowledgments
The authors would like to thank Emily Gee, Jasmine Amoako, Jiun Park, and Kennedy Andara for help with research; Anh Nguyen and Bill Rapp for graphic design and data visualization; Carl Chancellor, Bianca Serbin, and Steve Bonitatibus for helpful edits for clarity and readability; and Audrey Juarez for legal review. This report was made possible in part by a grant from the Peter G. Peterson Foundation. The statements made and views expressed are solely the responsibility of the authors.
Methodology
The authors have adjusted the CBO’s 2025 long-term budget outlook to 1) incorporate the impact of the Big Beautiful Bill and, from there, assume its temporary tax cuts and spending increases are extended; 2) incorporate the new tariffs imposed in President Trump’s second term but assume only one-third of that new tariff revenue remains after Trump leaves office; 3) assume appropriations followed the budget deal; 4) assume disaster spending matches historical levels in the outyears; and 5) assume any other emergency spending does not continue, with the exception of emergency-designated money that is intended to fund ongoing base operations.36
To estimate the impact of the Bush tax cuts, the authors used various CBO and Joint Committee on Taxation (JCT) estimates.37 These estimates include both the revenue and outlay effects of these laws but use only the tax code changes. The authors first assumed the AMT was permanently indexed for inflation at Clinton rates and therefore only attributed to the Bush tax cuts their portion of the cost of the 2013 permanent AMT patch.38
To estimate the impact of the first round of the Tax Cuts and Jobs Act, the authors used the cost estimate provided in the CBO’s April 2018 baseline and, from there, extrapolated the long-term impact.39 As with the Bush tax cuts, these estimates include both the revenue and outlay effects of the law as well as the increased interest costs from the measures.
For the Big Beautiful Bill, the authors used estimates provided by the Committee for a Responsible Federal Budget.40 This includes both the tax and spending changes to the law, encompassing both the spending increases and spending cuts.
To estimate the alternative scenarios for debt, interest, and primary deficits without the Bush or Trump tax cuts, the authors incorporated the macrodynamic changes to growth and interest rates. This includes both incorporating the effect that never having done the tax legislation in the first place would have on growth and interest rates, as well as the impacts from the lower resulting debt.