Over the past several issues we have followed one thread from different angles. In Mr Bond Market Always Wins we argued that bond investors, not politicians, will ultimately set the limits of fiscal policy. In Productivity or Reckoning we asked whether a productivity boom could arrive in time. In Who Buys the Debt? we looked at the shrinking pool of natural buyers for government bonds. And in Slow, Then Sudden we tested the claims being made about AI adoption against the evidence.

This issue steps back and asks the simplest question in public finance: how much is growth actually worth? Governments across the developed world are now making hard choices on spending because the price of their debt has risen sharply. The debate tends to centre on what to cut. We think it should centre on what to grow.

The Arithmetic of a Hard Choice

The starting position is well documented. CBO's February 2026 outlook puts US federal debt held by the public at US$32.1 trillion, or 100.6% of GDP, in fiscal year 2026. The deficit is projected at US$1.9 trillion. Net interest costs alone are expected to reach US$1.04 trillion this year, rising to US$2.1 trillion by 2036. On CBO's numbers, debt reaches 120% of GDP within a decade, well past the previous record of 106% set in 1946.

The problem with fixing this through spending restraint alone is that most of the budget is hard to touch. Interest must be paid. Social Security, Medicare and Medicaid are politically entrenched and grow with an ageing population. Defence is rising, not falling. That leaves the third lever, the one that requires no one to be made worse off: a larger economy. When GDP grows, the same tax rate collects more revenue, and the existing debt shrinks relative to the economy carrying it.

The Experiment: 2% Versus 4%

We ran a ten-year comparison for the United States. One economy grows at 2% a year in real terms — roughly the pace CBO expects. The other grows at 4%. We then applied the extra tax revenue from faster growth to the national debt and compared where each economy ends up. The method is deliberately simple, so the result is driven by growth and nothing else. Federal revenue held at 17.5% of GDP in both scenarios. Non-interest spending follows the same dollar path. The effective rate on federal debt rises from 3.3% today to 4.0% by 2036 — and lower debt means lower interest, so that saving compounds.

Ten-year outcome (to 2036)2% growth3% growth4% growth
Nominal GDP in 2036US$47.4TUS$52.3TUS$57.6T
Cumulative tax revenueUS$69.9TUS$74.0TUS$78.3T
Extra tax vs 2% growth—US$4.1TUS$8.5T
Interest saved vs 2% growth—US$0.5TUS$1.0T
Debt held by the public, 2036US$57.8TUS$53.2TUS$48.4T
Debt to GDP, 2036122%102%84%
Annual deficit, 2036 (% of GDP)7.1%4.5%2.2%

Over ten years, the 4% economy collects US$8.5 trillion more in tax than the 2% economy, without a single change to tax rates. That is the cost of slow growth: money a slower economy simply never raises. The dividend is back-loaded — in the first year the gap is only about US$110 billion, but by year ten it is US$1.78 trillion a year, more than the entire federal interest bill today.

US$8.5TExtra tax revenue at 4% vs 2% growth over ten years — without any change to tax rates
38 ptsDebt-to-GDP swing: 122% at 2% growth versus 84% at 4% growth by 2036
20 ptsWhat the first extra percentage point of growth is worth in debt-to-GDP terms over a decade

The Power of Just One Per Cent

Four per cent real growth sustained for a decade is ambitious. The more important finding may be the middle column. At 3% real growth — one point above the slow-growth case — the US collects an extra US$4.1 trillion in tax over ten years. That is enough to hold debt to GDP almost perfectly flat, at around 102%, for the entire decade. It does not solve the debt problem. But it stops it getting worse, and it does so without the political cost of deep spending cuts or major tax rises. The first percentage point of additional growth is worth roughly 20 points of debt to GDP over ten years — a larger fiscal improvement than almost any consolidation package currently under serious discussion in Washington.

Faster growth can mean higher interest rates, which would erode part of the dividend. But the relationship that drives debt sustainability is the gap between the interest rate and the growth rate. A two-point increase in growth would need to be matched by a two-point rise in the average cost of the entire debt stock to cancel out, and because most debt is refinanced gradually, that would take many years to feed through. And has 4% ever happened? Yes. In the late 1990s the US economy grew at more than 4% a year in real terms for four consecutive years. Combined with the 1993 tax increases and spending restraint, that growth delivered four consecutive federal surpluses. Debt held by the public fell from close to 48% of GDP in 1993 to around 31% by 2001.

Can AI Deliver the Growth?

If growth is the only painless exit, the obvious question is where it comes from. Population growth in the developed world is slowing. Labour force participation is constrained by ageing. That leaves productivity — and the most credible source of a productivity step-change in the coming decade is the adoption of artificial intelligence. The forecasts span a wide range: Goldman Sachs estimates up to 1.5 points a year added to US productivity growth with widespread adoption, while Acemoglu at MIT sees productivity gains of under 1% in total over ten years. Mainstream estimates suggest something between 0.2 and 0.5 points at the peak — valuable, but not a rescue.

Two observations matter. First, forecasts of general-purpose technologies have a long history of underestimating the eventual impact while overestimating the speed of the early phase. The internet looked like a disappointment in productivity data until, briefly, it did not. Second, adoption speed is not fixed — it is a function of policy. Energy availability, grid connection queues, permitting, data centre approvals and regulatory clarity all determine how quickly AI moves from pilot to production. A government that treats AI adoption as a fiscal strategy, rather than a technology question, has the ability to pull the growth dividend forward.

So, What's Around the Corner?

Growth is not one option among many for indebted governments. On the numbers, it is the only option that does not involve asking someone to accept less. One point of extra growth stops the US debt ratio rising; two points reverse it.

AI is the most credible source of that growth over the next decade, but its contribution is not guaranteed. It depends on how fast economies are allowed to adopt it — and that depends on energy, infrastructure and policy. The signals we are watching: token usage and enterprise AI deployment rates, US productivity data from 2027 onwards, data centre power connection queues, and the term premium on long-dated Treasuries. If the first three accelerate, the growth scenario becomes our base case. If the last rises while the others stall, the reckoning moves closer.

Disclaimer

This publication is provided for general informational purposes only and does not constitute financial, investment, legal or tax advice. Views expressed reflect the judgement of Monard's Investment Committee as at the date of publication and are subject to change without notice. Past performance is not indicative of future results.

Sources: Congressional Budget Office, The Budget and Economic Outlook: 2026 to 2036 (February 2026); Goldman Sachs Research; Penn Wharton Budget Model, The Projected Impact of Generative AI on Future Productivity Growth (2025); D. Acemoglu, The Simple Macroeconomics of AI (2024). Scenarios modelled by Monard from CBO's February 2026 baseline.