Holding U.S. debt flat would take 5%-6% real growth, CBO's Swagel estimates

In Swagel's account, growth helps less than it appears to because it also raises what the government spends. Federal spending lifts growth, faster growth lifts wages, and wages feed into Social Security benefits ^[5]. A strong economy also tends to push interest rates up, adding to the cost of carrying the debt ^[5]. His figures have nominal growth beating the assumed interest rate by about 3 points a year ^[24], and roughly 2 points of the nominal figure is inflation ^[25]. "So growth will help, but it's probably not plausible that growth alone will stabilize our fiscal trajectory," he said ^[6].

All three public thresholds sit above what the market expects. The bullish Wall Street forecasts put full-year growth at 2.5% ^[13]. That is half a point below Bessent's 3% and 2.5 to 3.5 points below Swagel's 5%-6% ^[26]. Bessent's case also depends on events going his way. "We'll get to the other side of this Iran conflict, and the underlying economy is very, very strong, and I think reaccelerating," he said at Southern Methodist University ^[21].

Productivity is the place where Swagel's numbers could turn out too gloomy. Minneapolis Fed President Neel Kashkari asked whether AI could speed up growth, and Swagel said CBO has detected an increase in total factor productivity ^[8]. He said future growth will be stronger, but that the deficit is too deep for AI-driven growth to be enough ^[9].

So far, the debt itself has done little to rates. On Swagel's coefficient, CBO's projected 20-point climb in the debt ratio by 2036 ^[22] adds about 0.3 percentage points, or 30 basis points, to long-term rates ^[23]. Long yields are already at their highest in 24 years ^[17]. Fortune puts that down to the strong economy, expected Fed hikes, oil keeping inflation high and a flood of AI hyperscaler debt competing for buyers, with the size of the debt also playing a part ^[18]. "So it's modest, but the fiscal trajectory is really quite challenging," Swagel said ^[20].

We think the debt reaches bond prices mainly through a shock. Swagel said there is "almost like a turbocharger," in which "An interest rate shock feeds into the deficit, feeds into the debt, feeds back into interest rates" ^[16]. On CBO's estimate, a decade of slow drift costs about a third of a point ^[23], while a shock keeps adding to itself through that loop. The counter-case is Bessent's: 3% growth does the job, and nobody has to touch taxes or spending ^[14]. On Swagel's numbers, that growth rate is 2 to 3 points too low ^[27]. We would be wrong if long yields kept climbing after oil, Fed expectations and AI borrowing eased. That would show the market charging more for the debt than CBO's 0.015 coefficient implies ^[19].