The US budget math is looking dangerous

Key Takeaways

• The combination of higher interest rates, slower growth, and large deficits is spelling danger for the US fiscal situation. Unlike past situations where an overabundance of caution led to harmful budget austerity, we worry this time is different.
• We find the interest rate is no longer lower than the growth rate. If that persists, future debt servicing becomes a much heavier lift.
• The damage to the fiscal outlook from the GOP budget bill brings into focus the strain that deficit-financed tax cuts will put on the federal debt.
• Our analysis reveals the possible future need for the U.S. to require an "ahistorical adjustment" to its spending and taxation levels. Policymakers, economists, and budget experts should be concerned about our fiscal health and begin to plan accordingly.

In the movie The Princess Bride, one of the characters frequently refers to events unfolding before them as “inconceivable” until one of his underlings points out, "You keep using that word. I do not think it means what you think it means."   

In a similar—albeit far less humorous—world, “unsustainable” is often used to refer to the level and the path of federal debt. For decades, so-called budget hawks have warned that the level of debt carried by the U.S. federal government was unsustainable.

But what do these budget hawks mean by unsustainable? Conversely, what do doves mean when they counter that the debt is sustainable? If the hawks mean that the government will not, in near-term debt auctions, be able to find investors in government bills and bonds willing to lend at the going rate of interest, then the Princess Bride quote seems applicable. The U.S. debt market is the largest, most liquid, lowest-risk sovereign debt market in history. For decades, the interest-rate record from that market has not reflected obvious sustainability concerns; to the contrary, the historical record over the last few decades directly contradicts these concerns, as rates on government debt have generally drifted down even as publicly held debt has drifted up.[1]

On the other hand, if those warnings about unsustainability simply convey the vague sense that “surely the debt level can’t rise forever”—that somewhere in the future, there must be a limit to what credit markets would support—perhaps the hawks have a point. This, however, seems an awfully abstract application of the word, certainly not one that is helpful to policymakers and budget writers who tend to heavily discount the distant future.

Whatever the motivations of various actors in this fiscal drama, the fact remains that for many years, warnings of an unsustainable debt path seemed premature. There was, as noted, little evidence of a negative relationship between levels and changes in debt, deficits, and interest rates, including expected future rates. Had these unsubstantiated warnings been benign, they would not have mattered. But there were instances in which they contributed to an austerity agenda under which people and places were needlessly harmed by unwarranted fiscal retrenchment. For instance, in the aftermath of the 2008 crisis, the strength of recoveries in numerous countries was constrained by poorly founded deficit worries.[2],[3]

More recently, however, long-standing budget doves are sounding more hawkish, warning that “this time is different.” A recent article in The Atlantic titled “The Debt Is About to Matter Again” documents this change, citing many former doves who are beginning to worry about the country’s fiscal path. As economist and former Treasury Secretary Larry Summers summarizes, “In a short amount of time, the fiscal picture has gone from comfortably in the green-light region to the red-light region.”[4]

In this policy brief, we ask what has changed in the budget outlook that is leading more analysts to worry that this time does appear to be different.

We do so by first examining the key variables behind the shift—the economy’s growth rate (g), the interest rate on the debt (r), and the primary deficit itself (s)—hewing closely to important work in this space by Olivier Blanchard [5], as summarized in his recent book Fiscal Policy under Low Interest Rates. We show, for example, that the real interest rate, r, has risen sharply since the end of his analysis period (2021) and is now roughly equal to the growth rate, g. While we are appropriately humble about where r goes next, should this equality of r and g stick or worsen, the implications for debt sustainability are considerably more dire than if r were to revert back to its earlier, negative trend.

We then turn to current events, arguing that the trade and fiscal policies pursued by the Trump administration are potentially damaging for debt sustainability. Though the future path of interest rates is impossible to predict, recent policy developments suggest a higher likelihood of the toxic fiscal combination of slower growth, higher interest rates, and larger primary deficits.

All of which leads us to conclude that this time does appear to be different, that the moment deserves concern not just from the traditional hawks but the broader community of economists beginning to sound the alarm. We join a growing chorus who worry that the probability of a debt shock has risen and offer thoughts drawn from our own results and those of others as to what should be done.

A skeptic might argue that politicians on both sides of the aisle have little to gain from tackling the problem of debt sustainability and, in the near term, we should expect inaction from Washington.

It would be naïve not to share this concern, as politicians are incentivized to focus on shorter-term risks and have other constraints that supersede concerns over debt sustainability. Still, there are many cases throughout history where market shocks and crises have triggered responses, forcing even the most recalcitrant politicians into action. And even if we cannot find political allies in the current climate, we must still prepare for a moment when our input is needed, should such a shock occur. It is correctly said that in politics and policymaking, “plan beats no plan.”

What do we mean by unsustainable?

Recent literature has offered a few definitions of fiscal unsustainability. Blanchard writes that “debt is sustainable if the probability of a debt explosion is small,” noting that one still must define “explosion” and “small” for this definition to be useful. Abecasis et al [6]. offer a definition of “explosive” debt and add a useful historical reference:

"If the debt grows large enough, the fiscal trajectory could become 'explosive' in the sense that interest expense would be so large that stabilizing the debt-to-GDP ratio would require running persistent fiscal surpluses of a size that has seldom been sustained in the past and is unlikely to be sustained in the future because it is economically costly and politically difficult.”[7]

These definitions are unavoidably imprecise. It would not be credible to argue that a specific debt ratio or debt-service level is de facto unsustainable. It has recently been argued, for example, that last year, U.S. sovereign debt service was, for the first time, larger than defense spending and that this was a sign that fiscal consolidation was urgently needed.

But it is not clear why that’s an unsustainable barrier,[8] especially if the imbalance occurred in a period when growth was relatively strong and interest rates were relatively low. And, of course, some countries, most notably Japan, have maintained very high debt ratios without exploding debt.[9]

We will therefore hew more closely to the Blanchard and Abecasis et al definitions, which rest on degrees of fiscal space: How heavy a fiscal lift would be needed in terms of revenue increases and spending cuts to stabilize the debt ratio?

We look at history and at political constraints to ask how realistic such adjustments would be without major economic fallout. Because there are plausible scenarios of our current fiscal trajectory that we believe would require historically large fiscal consolidations—very “heavy lifts” in the above phrasing, or what we label “ahistorical adjustments”—and because we believe these consolidations could be both economically painful and politically very challenging, we are thus more worried about budget sustainability than at any time in the recent past.

The holy trinity of debt stability: growth, interest, and primary balances

We begin with a discussion of the simple formula for the stability of the debt ratio (the ratio of government debt to GDP), as written in Equation 2 below. Note that our definition of debt ratio stability says nothing about the actual level of the debt ratio, only its change over time. That is, debt can be stable at a ratio of 50 percent, 100 percent, 200 percent, or higher.

This point has at least two implications. First, as discussed above, we know of no credible arguments that any specific debt level is inherently unsustainable. As we noted, many countries, such as Japan, maintain high debt ratios for long periods of time.

A scatterplot of countries’ debt ratios (x-axis) against their long-term interest rates (y-axis) does not show a clear pattern (Figure 1).[10] There are, for example, a cluster of countries with (nominal) yields between around 2.5 percent to 3.5 percent, yet with net debt ratios that range from close to zero to 100 percent.

Figure 1: Net Government Debt vs. 10-Year Bond Yield, Advanced Economics

The level of the debt ratio, however, does matter in some respects. At a given interest rate, the higher the debt ratio, the higher the total debt service costs. And high debt service costs can lead potential creditors to assume higher risks when considering investing in sovereign debt, a feedback effect that pushes up the rate of interest. This implies an endogeneity between rising debt levels and the interest rate on that debt.

A common rule of thumb is that an added point on the debt ratio raises the interest rate on government debt by one to three basis points.[11] But, at least in terms of first-order effects, the actual level of the debt is not a major factor in the simple mechanics of debt trajectory.[12]

Equation 1 shows that the debt ratio, b, is equal to the ratio of the rate of interest on publicly held debt, r, (plus one) to the growth rate, g, (plus one) times last period’s debt ratio b(-1). The last term, s, is written here as the primary surplus, the surplus in receipts over spending in the current period, net of interest payments on the debt. A primary surplus subtracts from the debt and a primary deficit adds to it. See Figure A-2 in the appendix for a set of simulations designed to show the impacts of different values for r, g, and s on the debt ratio’s trajectory.

Equation 1

Equation 2 rearranges 1 to derive the sustainability equation. The left-hand side changes from the level of debt to the change in the debt from one period to the next, which is a function of the difference between the interest rate and the growth rate, the last period’s debt ratio, and, again, the primary surplus. 

Equation 2

We introduce these equations to underscore the centrality of r, g, and s. The (r – g) term has been highly scrutinized in research on debt sustainability. In this brief, we also focus on s, the primary surplus or deficit, as we expect the latter to become increasingly negative in coming years, posing a credible threat to future sustainability. We will also look at s to show the magnitude of adjustment needed to correct the fiscal path risks being outside the historical range.

For now, however, assume s equals 0 (a primary balance). With this assumption, the debt ratio rises or falls depending on whether r<g or vice versa.[13] As Blanchard summarizes, “debt increases at rate r, output increases at rate g, so the debt ratio increases at (r – g).” Intuitively, if the economy is growing faster than the interest rate on its debt and there is no primary deficit, the government is generating enough income to both service and roll over its debt without worrying about that debt unsustainably spiraling upward.

The formulas also suggest that if r>g in the long run, meaning borrow rates outpace growth rates, the government must run primary surpluses to keep the debt sustainable. The larger the debt, and the larger (r – g) is, the larger the surpluses must be. Again, this is intuitive: If growth is insufficient to service the debt because g<r, then government must save elsewhere. Conversely, this implies that if r<g, the government can run primary deficits without fearing a debt spiral. However, that outcome is less comforting than it sounds. While the debt ratio won’t explode with r<g, it can still relentlessly climb.[14]

Where are r and g today?

Kogan et al. find that over the long historical record, r is generally less than g.[15] More recently, real and nominal interest rates have trended downward. The reasons for this are debated, but the trend, which is common across advanced economies, is not (proposed explanations include aging demographics, global savings gluts, increased risk aversion leading to greater demand for safe assets). Meanwhile, while real growth rates have drifted slightly down, r<g dynamics, have been driven by the decline in the safe real rate, r.  We see this in the figure below from Blanchard with data through 2021 (Figure 2).[16]

Figure 2: Blanchard's r and g paths, 1992-2021.

However, in Figure 3, we update Blanchard’s figure using data through 2025. Over the last few years, r has climbed sharply such that r now equals g, leading to very different and much less favorable debt dynamics.

Figure 3: Real Interest Rate (r) and Real Growth Rate (g) Through 2024

Of course, the update begs the question whether r’s sharp rise is temporary or more lasting. As noted, the factors for r’s long decline are widely debated, not to mention the fact that interest rate movements are very hard to accurately forecast. Our update shows r rose quickly, and it could similarly fall quickly. However, the recent sharp rise in r suggests a heightened risk of debt unsustainability, one we worry could be exacerbated by the Trump administration’s trade and fiscal agendas.

Why this time could be different

The recent jump in r partially explains why, as Summers puts it, the U.S. may be moving from the green-light region to the red-light region in terms of debt sustainability.

Recent policy changes such as tariffs and the heavily deficit-financed budget proposal are pushing each variable of the sustainability equation in the “wrong” direction. We expect the real growth rate to slow, at least in the near term, due to President Trump’s trade war; numerous nonpartisan forecasts expect the recently legislated budget to raise both interest rates and primary deficits.

Growth rates: The Trump administration’s trade war has led to widespread markdowns in near-term growth expectations and increases in inflation expectations.[17] Sweeping tariffs could hit g in the form of reduced consumer spending due to higher costs and lower business investment due to heightened uncertainty regarding the tariffs’ on-again, off-again implementation and their levels, which as of this writing are, at about 16 percent, the highest since 1936.

Mark Zandi, the chief economist of Moody’s Analytics, has simulated the macroeconomic impact of three separate tariff scenarios, from pre-2025 levels of effective tariff rates (Scenario 1 in Figure 4) to a very high tariff regime consistent with the more extreme threats from the administration (Scenario 3; see Appendix for scenario details).

Figure 4 shows that even a moderate tariff regime can be expected to slow the growth rate over the next few years relative to a pre-tariff-hike baseline, whereas much higher tariffs lead to a dramatic deceleration of real growth. Using a different model, the Yale Budget Lab finds (Figure 5) the actual 2025 tariffs to date (similar to Zandi’s Scenario 2) lowering the level of real GDP by close to 1 percent by the middle of 2026.

Figure 4: Annualized Real GDP Growth Rates Under 3 Tariff Scenarios 2025–28

Figure 5: Real GDP Level Effects of Tariffs in 2025, Relative to Baseline

If these forecasts are correct, tariffs will have a negative impact on g, which is especially concerning given the recent and rapid rise in r. There are, of course, nuances to consider. If we reasonably assume that tariffs generate a one-time hit to the economy, analogous to a one-time tax hike (in this case, a sales tax on imports), then they are likely to lower the longer-term level of GDP relative to a non-tariff counterfactual, but not the longer-term growth rate. This can be seen at the end of the time-series in Figure 5, which settles in at -0.34 percent, meaning this is how much the tariffs lower the long-run level of real GDP.

Interest rates: Turning to interest rates, upward pressures on r stem from a variety of sources. When President Trump introduced his sweeping reciprocal tariffs, it triggered a large, risk-off sell-off in equity markets. Historically, such sell-offs were matched by flights to safety in U.S. Treasuries, leading to a decline in the “safe rate” relative to riskier interest rates. But in this case, quite a different pattern prevailed: Bond prices fell (and yields rose), in what Wall Street traders call the “Sell America” trade (the U.S. dollar exchange rate also declined, also unexpected when U.S. interest rates rise). The market’s response to Trump’s sweeping “Liberation Day” tariffs moved the yield on the 30-year Treasury bond over 50 basis points in just a few weeks.

As shown in the Appendix, Zandi’s tariff scenarios predict the yield on the 10-year Treasury to increase as much as 1 to 2 percentage points under the moderate (S2) and aggressive (S3) tariffs.

A related factor putting pressure on rates in this period is the so-called term premium, which recently moved from being persistently negative into positive territory. This premium is typically defined as the extra return creditors demand for locking up their funds for longer maturities (Figure 6 below shows the term premium on 10-year government bonds). Term premiums reflect higher perceived risk, such as default risk or inflation risk, and here we see that manifested even among allegedly risk-free rates. Further evidence of such risk can be seen in the widening spreads on credit default swaps on U.S. government debt, which have trended up in recent months.[18]

Figure 6: 10-Year Treasury Yield Premium Over Time

The budget bill and sustainability

Another source of pressure on debt sustainability comes from the fiscal path implied by the 2025 budget bill recently passed by congressional Republicans, particularly regarding its impact on higher primary deficits and interest rates.[19]

• The CBO initially scored the House version of the bill as adding $2.4 trillion to the debt over 10 years, raising the debt to 124 percent of GDP from a baseline of 117 percent. CBO later added a dynamic score, which increased the bill’s cost to $2.8 trillion over 10, as any growth effects were overpowered by higher interest rates (a germane consideration for r and g dynamics).[20]
• In their dynamic score, CBO estimated that the primary deficit would increase by $2.3 trillion, or 5 percent of GDP, by 2035 (net interest costs in this score were up $440 billion).
• The Yale Budget Lab scores the Senate version of the bill (which bears the closest resemblance to the version signed by the President) will raise the debt ratio by 10 percentage points over 10 years and raise interest payments as a share of GDP by 40 basis points (in both cases, relative to CBO’s baseline).[21]
• CBO estimates that the Senate version would raise the deficit by $3.4 trillion over 10 years. Note that this assumes that temporary measures are not extended, an assumption we consider to be unrealistic. On a permanent basis, the bill would raise the deficit by over $4 trillion.[22] The comparable debt ratios in 2035 are 127 and 130 percent.

Figures 7 and 8 from Abecasis et al. from (Goldman Sachs Economic Research) provide a useful picture of the potential risks the new budget bill introduces to the fiscal path.[23] Figure 7 shows that under every interest rate assumption, including ones well below their baseline (i.e., 2 percent nominal vs. the baseline 4 percent), the debt ratio climbs to levels above the historical range. Figure 8—real debt service as a share of GDP—is one that some budget analysts have argued provides a more accurate read of sustainability.[24] The pattern for nominal interest rates above 3 percent takes this metric beyond its historical range. We will argue that such forecasts both imply fiscal adjustments that are also outside of the range of history and can thus be useful guideposts for sustainability analysis.

Figure 7: Debt Held by the Public (% of GDP)

Figure 8: Federal Real Interest Expense Projections (% of GDP)

Finally, we focus on the implications of the budget bill for s, the primary surplus from equations 1 and 2. Those equations show us that if r and g remain less favorably aligned, achieving debt stability will require moving from our recent history of primary deficits to primary surpluses.[25]

Abecasis et al. estimate, for example, that keeping real debt service as a share of GDP below 2 percent “the highest seen over an extended period in the postwar era,” would require 3.5, 3.9, and 5.4 percentage point increases in the primary balance as a share of GDP beginning in years 2025, 2029, and 2035, respectively.[26]

In historical terms, required increases of this magnitude constitute a “heavy lift.”. Figure 9 shows primary balances since the 1960s plus two forecasts of the period from 2025 to 2035: CBO’s baseline before the newly legislated budget bill and the Yale Budget Lab’s estimate of primary deficits based on the Senate version. The three dots at the end of the figure show the magnitude of the adjustment needed to stabilize the real debt service ratio as per Abecasis et al.

Figure 9: Primary Deficit, Historical Data and Projections (% of GDP)

We consider the magnitude of these distances—the delta from the expected primary deficits and the primary surplus necessary to stabilize real debt service—to be “ahistorical adjustments,” as they represent large changes in the primary balance relative to recent budgetary history. 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. This shows such an adjustment can be made, though this one involved both policy and luck. The Clinton years saw a capital gains boom, a tax increase on high-income earners in the 1993 budget, full-capacity growth, and spending restraint, including cuts on defense enabled by the end of the Cold War.

New pressures on g, r, and s, and a lack of political will to consider these risks lead us to worry that this time might be different. We should also consider, though, why this assessment might not be correct.

Why sustainability risks may not be higher

Should tariffs revert to pre-2025 levels, we expect that their negative impact on growth would be eliminated, as shown in Zandi’s Scenario 1. Though a full reversal is unlikely given the administration’s hostility toward international trade, such a scenario could occur if we witness further negative reactions to the tariffs in the Treasury market. Indeed, the administration revealed a willingness to dial back its most extreme policies in the face of the negative bond market reaction, issuing a 90-day pause in early April that was then extended again in July.

Of course, if the Republican budget bill were to unleash significant growth through supply-side improvements in productivity and labor force growth, that could be favorable for g and s.[27] Notably, even conservative-leaning budget analysts, and certainly most nonpartisan ones, do not take this view. The Joint Committee on Taxation (JCT) found that the House version of the budget bill would boost the 10-year GDP growth rate from 1.83 percent to 1.86 percent, an increase of only 3 basis points; CBO’s estimate was not far off, at 4 basis points.

One growth score stands out from the rest: that of President Trump’s Council of Economic Advisors (CEA), who estimate that the bill would increase the level of real GDP by just under 5 percent, raise real average annual GDP growth by more than 1 percent, and raise the level of real investment by 7-10 percent.[28] For many reasons, we take little solace in these estimates. First, their estimates of the positive dynamic feedback (growth effects) from the bill are extreme outliers relative to non-partisan modelers. As CRFB points out, “CEA’s claim of $2.1 to $2.3 trillion of dynamic feedback is nearly 2.5 times as large as the next highest estimate, more than 50 times as high as the average of outside modelers, and the opposite sign of CBO and others that conducted a comprehensive analysis of the bill.”[29] 

One reason for this disparity is that CEA bases its forecast on the claimed success of the Tax Cuts and Jobs Act, although the TCJA fell short of initial estimates—for instance, recent analysis by Gravelle and Marples find that equipment and structure investment was largely unaffected by the TCJA.27 CEA’s outlier estimates are also driven by unfounded assumptions regarding capital-formation elasticities and labor-force multipliers.[30]

The CEA bases its forecast on the claimed success of the Tax Cuts and Jobs Act, although the TCJA fell short of initial estimates—for instance, recent analysis by Gravelle and Marples find that equipment and structure investment was largely unaffected by the TCJA.[31] CEA’s outlier estimates are also driven by unfounded assumptions regarding capital-formation elasticities and labor-force multipliers.[32]

Okay, the risk is elevated. What should we do about it?

While we argue that “this time is different,” ultimately, we do not know where the key sustainability variables, namely r and g, are headed (we’re fairly confident that primary deficits will be growing). It is clear, however, that the likelihood of debt service requiring historically unprecedented adjustments in the primary balance is higher now than in recent decades.

The question therefore becomes what can we do about this. There exists an extensive literature that addresses this question, from which we can borrow ideas to avoid future crises.

Before brainstorming new tools, however, we must examine past attempts to solve these problems—and understand why such tools have not worked. For instance, while the debt ceiling was originally envisioned to ensure fiscal sustainability, today it’s weaponized by partisans to hijack the budgeting process, raising risk premia and making debt service more expensive. And while Gramm-Rudman’s enforced spending constraints were once thought of as a useful tool, history has shown them to be subject to “clumsy and unworkable sequestration procedures”[33] consistently violated by politicians on both sides of the aisle. Rating agency downgrades, while on occasion providing some useful signals, have also been consistently ignored in the budget process.

In general, “lines in the sand”—limits to deficits or debt ratios—have at least two major shortcomings. First, because they do not evaluate variables discussed in this brief, they risk invoking austerity when, based on r – g, primary balance adjustments are not required. Second, unless constructed otherwise, they are insensitive to the business cycle, muting appropriate fiscal responses to crises. Rules based on hard caps to deficits and debt ratios, such as the Maastricht Treaty rules in Europe, can be sub-optimal for these reasons.

As we think about new approaches to debt sustainability, there are at least three challenges to designing good policies to prevent debt crises. We must 1) define specific “break-glass moments” that force a response, 2) calibrate fiscal adjustments that respond appropriately to these triggers, and 3) work with policymakers to set up binding responses to these crises, or at least secure their commitment to respond with necessary adjustments when such triggers are reached.

Setting triggers

Blanchard introduces a framework of stochastic debt sustainability analysis (SDSA) to identify periods when the probability of a debt shock is heightened. This process requires modeling the trajectory of (r – g) based on its historical distribution and plugging in assumptions for primary deficits based on forecasts n years out. This yields a probabilistic likelihood of explosions over the n-year window. Using this framework, budget analysts can model how various fiscal rules and policy approaches would impact the probability of future debt crises.

The Blanchard process is similar to the one applied by Auerbach and Yagan, whose 2024 paper begins with the assertion that “America’s fiscal path is likely unsustainable.”[34] The paper quantifies the assertion that in the last two decades, Congress has been significantly less responsive to the deterioration of the fiscal outlook than in the previous two decades.[35] They define debt sustainability by assuming that:

… there is a threshold level of the debt-GDP ratio that is not plausibly sustainable. That is, we assume that if the United States crosses a very high debt-GDP threshold within the next 100 years, the debt sensitivity of the interest rate on government debt would rise due to especially high default risk, further compounding explosive debt dynamics and leading to default.[36]

While a 100-year target is not politically salient to policymakers who are known for heavily discounting the future, by setting the exercise up this way, Auerbach and Yagan make a different point, one that strikes us as potentially helpful: avoiding (very) long-term unsustainability requires only relatively mild deficit reduction today (a 0.5 percent to 1 percent reduction in the deficit/GDP over the next decade). The sooner we start, the smaller the reduction needed.

We could also develop an “ahistorical adjustment” trigger based on Abecasis et al.—recall the outlier dots in Figure 9. In this case, if interest rates and real debt service are rising without corresponding increases in growth, leading to this type of pressure on the debt or debt service ratios, we would look at how large an adjustment in primary deficits is required to stabilize these trends. If that adjustment is in the upper range of historical movements from primary deficit to primary surplus, it would trigger appropriate action. Of course, as in any area, it would be the job of economists and policymakers to calibrate these models to ensure that they respond appropriately to fiscal issues, maintain credibility with policymakers, and avoid unnecessary fiscal austerity.

Calibrating appropriate actions

What would appropriate responses to these triggers look like? Most budget analysts—as in the work by Auerbach and Yagan or the CBO in fiscal space presentations—enumerate the percentage points of deficit reduction implied by the sustainability analysis. That’s the right place to start, but a more robust analysis includes more specificity about revenue increases and spending cuts, as well as optimal timing for those actions given growing debt pressures. Orszag et al. (2021) propose a set of ambitious measures to automatically avoid debt unsustainability, a framework they call “semiautonomous discretionary fiscal architecture.” This includes debt maturity extensions and automatic adjustments to meet the financing needs of the big entitlement programs.

From ideas to action

The challenges to tackling debt sustainability are of course not solely analytical, but also political. For decades, it has been easy and politically advantageous to cut taxes, spend freely, and ignore the rising debt. Such inaction was far less costly during the long period wherein r<g. So the first thing to do, assuming rates don’t fall back to their earlier, declining trend, is for budget analysts to make sure policymakers understand the extent to which the debt sustainability situation has shifted. Though their advocacy might be ignored, experts must nevertheless try to explain why this time is different.

Of course, politicians are most incentivized to act in times of crisis. A market shock, spiking bond yields, a failed Treasury auction, or some other fiscal crisis could be highly influential and force policymakers to act quickly, as was the case with the TARP following the market crash in 2008 (though it is not incidental that the TARP was deficit financed!).

Such a response plan would need to be not only informed by the best research but also legible, credible, and useful to members of the political class. Economists and budget experts must continue to refine the emerging tools of stochastic debt sustainability analysis, automatic fiscal adjustments as per Orszag et al., and r – g dynamics, translating these insights into actionable plans.

Conclusion

It is much easier to document a worrisome fiscal outlook than it is to design actions to appropriately get ahead of potential future crises. Even if a perfect solution was available, politicians operate on the timescale of election cycles: Arguments about leaving our grandchildren with high debt levels do little to move them. And in recent years, while the interest rate was below the growth rate (r<g), persistent deficits did not pose a risk to debt spiraling unsustainably.

The point of this brief, however, is that at least for now, updating Blanchard’s work with more recent data (Figure 3) shows that r is now equal to g and that primary deficits are larger than they’ve been and are growing more negative. Moreover, the Trump administration’s trade war is likely to lower at least near-term g, while the Republican budget bill is likely to raise r and even more so to lastingly raise primary deficits.

The implications of these changes recall the end of the 1995 Ball et al. paper titled “The Deficit Gamble,” wherein they compare disregarding an unfavorable fiscal outlook to homeowners not purchasing fire insurance on their homes. Though homeowners can raise their near-term living standards by failing to insure against the low-probability event of their house burning down, it is a risky bet.[37] We argue that recent evidence suggests the house is more fire prone than it used to be and the probability of a fire has gone up.[38]

Government debt matters again, and if the relevant parameters continue to evolve as we and others fear they will, it will increasingly continue to do so.

Appendix

Tariff simulations

For Mark Zandi’s simulated tariff scenarios, here is how he describes the three scenarios:

S1: Return tariffs back to where they were at the start of the year, beginning in 2025q3, and keep them there.

S2: 10% reciprocal, 25% steel and aluminum, 25% auto, and 30% China, beginning in 2025q3, and keep them there.

S3: 20% reciprocal, 50% steel and aluminum, 50% auto, 50% pharmaceutical and semiconductors, and 40% China, beginning in 2025q3, and keep them there.

The figure below plots for forecasted trajectory of the 10-year Treasury yield under these tariff scenarios. Over the medium term, the interest rate is forecasted to increase by about 20 basis points under S2 and about 60 bps under S3.

Figure A-1: 10-Year Treasury Yield (%), By Tariff Scenario

r, g Dynamics

The figure below plots four debt scenarios based on Equation 1 in the text, designed to show the debt ratio’s sensitivity to different parameters. The simulations start with the debt set at 100% of GDP. Line (1) shows the pressure on the debt ratio when r > g (in this case, r – g = 2%) and the primary deficit is a constant -2 %. The next line down (2) shows that even with that same primary deficit, with r < g (r – g = -1%) the debt trajectory is much less steep. The next two lines simulate primary surpluses of a constant 2%. Even with r – g, a constant primary surplus prevents the debt from rising. Finally, with r – g and a primary surplus, the debt ratio falls.

Figure A-2: Illustrative Debt Dynamics: r<g vs r>g

Footnote

[1] International Monetary Fund (IMF); Federal Reserve of St. Louis (FRED); Authors’ analysis based on Robin Brooks, 2005.

[2] Authors’ analysis.

[3] Growth Forecast Errors and Fiscal Multipliers, IMF Working Paper (WP/13/1)

[4] The Debt Is About to Matter Again, The Atlantic May 23,2025

[5] The Budget Lab, “State of U.S. Tariffs: June 17, 2025.

[6] Abecasis, Mericle, Phillips: US Daily: "Fiscal Sustainability: The Cost of Inaction," Goldman Sachs Economic Research, June 17, 2025.

[7] We later label this an ahistorical adjustment (AA) trigger.

[8] Niall Furguson argues that “any great power that spends more on debt servicing than on defense risks ceasing to be a great power.” While he provides compelling examples of this correlation, he does not speak to the unsustainability criteria which is our focus. Ferguson's Law: Debt Service, Military Spending, and the Fiscal Limits of Power, Hoover Institution History Working Papers, February 21, 2025.

[9] Countries that borrow in their own currencies, such as the United States and Japan, can clearly maintain higher debt-to-GDP ratios than those that borrow in foreign currencies. More in-depth analysis of this phenomenon lies outside the scope of this piece.

[10] Economist Robin Brooks has often featured similar figures. There are, of course, many reasons such yields will differ between countries, including currency differences, different inflation outcomes and expectations, different growth and productivity expectations, and so on.

[11]This dynamic might be already in play in the United States–we will return to this conversation below.

[12] Government Debt, Limited Foresight, and Longer-term Interest Rates, Finance and Economics Discussion Series 2024-027.

[13] The arithmetic of debt sustainability shows that one way in which the debt level matters is in the size of the primary deficit or surplus needed to stabilize the debt. Perhaps counterintuitively, if r<g, then, as Blanchard points out “the larger the debt ratio, the larger the primary deficit [the government] can run while keeping the debt ratio stable!” (pg. 54). If r>g, the same fact holds but for a primary surplus.

[14] Since g, the real growth rate, is generally “small,” we can safely ignore it in the denominator, 1-g.

[15] Blanchard: “If (r – g) is close to zero, and if the government runs a large primary deficit, debt may increase for a long time and converge to a very high level: For example, if s=−3%, g=2%, and r=1%, (r – g)=−1%, debt will still converge, but converge to a high 306% of GDP. In practice, such a large increase may be impossible to distinguish from an actual explosion.”

[16] Difference Between Economic Growth Rates and Treasury Interest Rates Significantly Affects Long-Term Budget Outlook, Center on Budget and Policy Priorities, February 27, 2015.

[17]Blanchard, Figure 3.3.

[18] Though, these have been more common in consumer surveys than in market-based metrics.

[19]See, for example: Credit default swaps are back in fashion — even if the panic might be overblown, CNBC, May 28, 2025.

[20] Because Senate Republicans used the reconciliation process to pass the budget bill in their chamber (a process that requires a simple majority to pass), they did not allow Senate Democrats to participate in the debate.

[21] H.R. 1, One Big Beautiful Bill Act (Dynamic Estimate), Congressional Budget Office, June 17, 2025.

[22] The Financial Cost of the Senate-Passed Budget Bill, the budget lab, July 1, 2025. 

[23] The Senate OBBBA in Charts

[24] Information Concerning the Budgetary Effects of H.R. 1, as Passed by the Senate on July 1, 2025, Congressional Budget Office. The Senate OBBBA in Charts, Committee for a Responsible Federal Budget, Jun 30, 2025. 

[25] Note that the Abecasis et al. analysis includes expected tariff revenues.

[26] In our updated figure above, this means following the flat line at the right of the figure from Goldman Sachs.

[27] See Table 3 in their paper. The numbers in the text use the simulation wherein interest rates rise by 2 basis points for every percentage point increase in the debt ratio.

[28] In some models, faster productivity growth is associated with higher demand for capital and less savings, putting upward pressure on r.

[29] Gravelle, Marples (2025): Economic Effects of the Tax Cuts and Jobs Act, Library of Congress, 04/07/2025. 

[30] https://www.whitehouse.gov/wp-content/uploads/2025/03/The-One-Big-Beautiful-Bill-Legislation-for-Historic-Prosperity-and-Deficit-Reduction-1.pdf

[31] https://www.crfb.org/blogs/ceas-fantastical-economic-assumptions

[32] The CEA uses a user cost of capital approach assuming an elasticity of -1, meaning a 1 percent increase in cost of capital means a 1 percent drop in business investment. This approach is fleshed out a bit more in their April report on extension of TCJA provisions and in a Kevin Hassett paper from February. However, elasticities vary widely by sector, and the literature suggests that business investment is not so elastic, leading their estimates to overcount likely growth impacts.

[33] A Surplus, If We Can Keep It: How the Federal Budget Surplus Happened, Brookings, December 1, 2000. 

[34] Robust Fiscal Stabilization, NBER Working Paper Series, January 2025. 

[35] See their Figure 1, panel c.

[36] Auerbach and Yagan

[37] Ball, L., Elmendorf, D. W., & Mankiw, N. G. (1995). The Deficit Gamble (NBER Working Paper No. 5015). National Bureau of Economic Research.

[38] Danny Yagan suggested this framing.