1 · The Current Global Public Debt Picture

Per the IMF Fiscal Monitor, global public debt has hit 94% of world GDP in 2025 (up from 92.4% in 2024) and is projected to cross the 100% of GDP mark by 2029 - a year earlier than the IMF's own April 2025 forecast. In the tail-risk scenario (5% probability), the ratio could reach 124% by 2029 (IMF Fiscal Monitor, 4/2026) (IMF Fiscal Monitor, 10/2025). Vitor Gaspar, Director of the IMF's Fiscal Affairs Department, put it bluntly: "Risks to public debt are widespread and skewed towards faster debt accumulation... Policymakers must act now to keep debt under control" (IMF press briefing, 10/2025).

One detail that gets little attention but matters a lot: fiscal space - the buffer that lets a government absorb an unexpected shock - has shrunk from over 1% of GDP a decade ago to nearly zero today, while the global primary deficit (excluding interest payments) still sits around 2.2% of GDP and is projected to narrow by only about 0.5 percentage points through 2031 - not enough to stabilize the debt/GDP ratio, let alone bring it down.

A few caveats worth reading alongside the chart above: Japan's gross debt (~230% of GDP) can't be compared directly with other countries, since most of it (~90%) is held domestically - by the central bank, commercial banks, and pension funds. Strip out the bonds the Bank of Japan (BOJ) itself holds, and Japan's net debt falls to roughly 130-160% of GDP (Conversable Economist, 12/2025). Vietnam, on this same measure, sits near the bottom of the chart - but that is precisely the point worth pausing on, because for an emerging market (EM), public debt/GDP isn't the most relevant risk gauge.

For an emerging market, the right measure isn't public debt/GDP

The public debt/GDP framework above was designed mainly for advanced economies - where the government borrows in its own currency in a sufficiently deep domestic bond market. For most emerging markets, the crisis mechanism is entirely different: researchers call it "original sin" - these countries typically can't borrow internationally in their own currency, so they're forced to borrow in foreign currency, meaning the main risk isn't a budget slowly being eaten away by interest payments, but a sudden currency devaluation that inflates the local-currency value of foreign debt overnight (Eichengreen, Hausmann & Panizza, 2003).

That's why, for an EM, analysts tend to weight foreign-currency liquidity buffers more heavily - FX reserves versus short-term external debt (the "Greenspan-Guidotti rule"), reserves versus M2 (broad money supply), and the IMF's ARA (Assessing Reserve Adequacy) index - rather than the public debt/GDP ratio alone.

2 · Where High Public Debt Actually Bites - And Who Really Pays

The shortest answer: no "government" ever pays the price for high public debt - only ordinary people, workers, savers, and generations not yet born bear it, through six fairly clear transmission channels.

The squeeze channel: a growing slice of the budget that builds nothing

This is the most visible channel, and the one playing out most clearly right now. As interest rates rise, governments pay more on the same old debt - and that money doesn't buy a single extra nurse, teacher, or bridge. It simply flows to bondholders.

In the US, fiscal year 2024 marked a milestone: for the first time in modern history, net federal interest payments ($881bn) exceeded the entire defense budget ($874bn) - and the gap kept widening in FY2025 (CRFB). The CBO projects net interest will keep exceeding defense spending through 2035, and rise to 5.4% of GDP (about 28% of total federal revenue) by 2055 (CBO, Long-Term Budget Outlook, 3/2025).

The remaining channels - briefly

Inflation / financial repression: when a government lets inflation run above bond yields, the real value of its debt automatically shrinks - but that cost falls on whoever holds bonds, savings deposits, or fixed pensions. The classic Reinhart & Sbrancia study estimates this kind of "liquidation effect" was once worth 3-4% of GDP a year for the US/UK in the post-WW2 era - a hidden tax on savers that never needed a congressional vote (IMF WP 15/7).

Currency risk: a country that can't borrow in its own currency on international markets (the BIS calls this "original sin") sees its foreign-currency debt grow heavier every time its currency weakens - precisely when its budget is already most stretched (BIS WP 1109). Argentina and Sri Lanka are the two most recent examples.

Austerity: paradoxically, "austerity" packages aimed at cutting debt often don't work as intended - the IMF's April 2023 WEO found roughly half of historical fiscal tightening episodes failed to reduce the debt/GDP ratio, because the accompanying growth slowdown shrank the GDP denominator even faster. The IMF itself has admitted it underestimated the negative impact (fiscal multiplier) of Greece's austerity packages by 100-150% relative to what actually happened (Poul Thomsen speech, IMF, 2019).

Default / restructuring: once every other channel is exhausted, creditors take a direct haircut - Greece's 2012 restructuring wiped out 53.5% of face value on €197bn of privately held bonds; Argentina's 2020 restructuring covered $65bn with a net-present-value haircut of roughly 45%.

3 · The Scariest Outcome Isn't Bankruptcy

Most large countries won't "go bankrupt" the way a company closes its doors - they can print their own currency (US, UK, Japan) or have the IMF/EU standing behind them (smaller countries in a shared currency bloc). What's far scarier, and far harder to picture because there's no specific date to point to, is gradual impoverishment: growth a little lower each year, wages rising a little slower, taxes a little higher, public services a little worse, businesses investing a little less. No single shock big enough to make headlines. Just ten years later, looking back, living standards have barely moved - while an ever-larger share of the national budget goes solely to servicing debt from the past.

This isn't just a theoretical model - it has already happened, and is happening now, in at least four major economies.

Italy: three decades of near-zero growth

Italy is the clearest case. Average real GDP growth has declined decade over decade: 1.54%/year (1990s) → 0.54%/year (2000s) → 0.23%/year (2010s). Per-capita GDP fell 5.9% between 2010 and 2019. Most striking: Italy is the only EU country with net-negative real wage growth since 1990 - as of 2023, real wages were still 3.4% below their 1991 level, while over the same period real wages rose 30.9% in France and 30.4% in Germany (OECD data via InTrieste). Italy's public debt now sits around 137% of GDP, and interest payments already consumed 8.9% of total government revenue in 2023 - the highest in the Eurozone/G7. Economist Ashoka Mody (Princeton, formerly IMF) describes this as a story of productivity and investment, not a bond-market collapse: Italy spends only 1-1.3% of GDP on R&D versus 3.5-3.8% in Sweden, its university graduation rate is 18% versus Sweden's 34%, and roughly 9% of Italians with a university degree had emigrated as of 2010.

The UK: the post-2008 "lost decade"

The Resolution Foundation calculates that if UK wages had kept following their pre-2008 trend, per-capita income today would be about £11,000 a year higher - a "missing pay" gap of 37%. Median household real disposable income (excluding pensioners) in 2018-19 was no higher than in 2007-08; income for the poorest 20% in 2018-19 was still no higher than in 2004-05 (Resolution Foundation). Today, public-debt interest payments (~£110bn, 8.1-8.3% of total government spending) already exceed the defense budget and equal nearly half the health/social-care budget - and the Office for Budget Responsibility (OBR) warns that if long-term rates stay elevated, debt/GDP could reach 140% within fifty years, with interest costs at 5% of GDP - the highest in over 70 years (Resolution Foundation).

France: interest costs are quietly overtaking the education budget

France's public-debt interest costs rose from €36.2bn (2020) to an expected €59.3-66bn (2026) - now one of the largest line items in the state budget, closing in on the education budget (€88.6bn). The Cour des Comptes (France's national audit body) warns the figure could hit €100bn by 2029 as old, low-rate debt matures and gets refinanced at much higher rates (Euronews, 12/2025). France's debt/GDP has climbed to roughly 118-119%, and the back-to-back government collapses of 2024-2025 (Barnier, then Bayrou) both centered on disagreements over how to close the deficit.

4 · Why Markets Can Stay Patient for Years - Then Suddenly Lose Faith

What's notable is that the mechanism behind this habit isn't mysterious at all - it has a name in economic theory: multiple equilibria in sovereign risk pricing. Economist Guillermo Calvo, in a classic 1988 model, showed that at the exact same debt level, the interest rate markets demand can jump discontinuously between two states: a "good equilibrium" (low rates, debt seen as sustainable) and a "bad equilibrium" (high rates, where the high rate itself makes the debt unsustainable) - and the switch between the two can happen purely because investor expectations shift, with no change required beforehand in the country's actual economic fundamentals. In other words: the same debt/GDP number can be both "fine" and "a crisis," depending on whether the market currently believes or doubts - and the line between the two states can be crossed in a single afternoon.

Research by Paul De Grauwe and Yuemei Ji (LSE) applies this mechanism to the eurozone, explaining why Greece, Italy, or France are more vulnerable than the US, UK, or Japan to this kind of self-fulfilling crisis: countries sharing the euro have no central bank of their own to act as lender of last resort, so once investors panic and pull capital out en masse, that panic alone - with no new fiscal trigger required - is enough to push the whole system into the "bad equilibrium" (De Grauwe & Ji, CEPR/VoxEU). That's also why the UK in 2022, despite having its own central bank and printing its own currency, wasn't fully immune either - once confidence in fiscal discipline breaks down, even a country with full monetary sovereignty can be pushed toward the bad state, just more slowly, and with intervention tools (like the BOE) to pull it back faster than eurozone members can.

A separate study by the Federal Reserve (IFDP) goes further, modeling what happens after that shock: a self-fulfilling crisis doesn't just produce a one-off yield spike - it can be followed by a prolonged period of stagnation (Federal Reserve IFDP 1370). Gradual impoverishment and a sudden loss of confidence, then, aren't two separate risks - a sudden confidence shock can itself be the trigger for exactly that kind of prolonged stagnation, rather than an isolated event after which everything simply returns to normal.

A concrete example: the UK, September 2022

The BIS's 2026 annual report, revisiting the UK's 2022 gilt shock, estimates that roughly half of the bond selloff following the budget announcement went beyond what actual fiscal fundamentals could justify - meaning it was a genuine repricing of confidence/liquidity risk, not pure interest-rate mathematics. This is the clearest real-world example of the mechanism above: UK government bond yields traded sideways at low levels for years despite rising public debt, then one seemingly small event - a budget judged reckless - flipped the entire risk pricing within a few trading sessions.

The sequence of events in September 2022 - when Britain's bond market turned into exactly that "steamroller" - unfolded as follows:

France has just gone through a milder version of the same mechanism: on November 28, 2024, French 10-year government bond (OAT) yields matched Greek yields (3.03%) for the first time in history (Bloomberg); by September 9, 2025, right after Prime Minister Bayrou's government collapsed, OAT yields even surpassed Italy's - the first time since the eurozone debt crisis. Credit rating agencies reacted accordingly: Moody's, Fitch, and S&P all downgraded France through 2024-2025, while in the US, Fitch cut its rating from AAA to AA+ (8/2023) and Moody's downgraded to Aa1 in May 2025 - the first time the US lost its last remaining AAA rating among the three major agencies.

None of this is new, either. In 1994, when the US bond market concluded the Clinton administration was loosening fiscal discipline, the 10-year Treasury yield jumped from 5.3% to 8.0% in just 13 months - an estimated $1 trillion in losses for bondholders. Clinton adviser James Carville's famous line: if he were reincarnated, he'd want to come back as the bond market - because "you can intimidate everybody."

5 · If This Continues

Kristalina Georgieva, the IMF's Managing Director, opened the IMF's October 2025 Annual Meetings with two blunt words: "Buckle up" - "uncertainty is the new normal, and it is here to stay" (IMF, 10/2025). The BIS's 2026 annual report goes further, describing a new "fiscal-financial nexus" taking shape: near-record public debt combined with higher interest rates is squeezing government fiscal space, making government bond selloffs "more frequent and more severe" - and the probability of a market-stress event is roughly 10 times higher when public debt is high versus low (BIS Annual Economic Report, 6/2026).

Add it all up, and the picture isn't a pre-announced doomsday, but a trajectory: fiscal space steadily draining, an ever-larger slice of the budget going purely to servicing the past, growth slowing bit by bit - and above all, a market tolerance threshold that nobody knows the exact location of until it's been crossed.

6 · The Long-Run Cost: Who In Society Pays Most

Every figure so far has been macro - GDP, debt ratios, bond yields. But high public debt always ends up landing on specific people, in a fairly unjust way: it forces central banks and governments into one of two exits, and both tend to widen the gap between rich and poor - just as another technological shock, AI-driven automation, is arriving.

Exit 1: print money, keep rates low - who benefits first

When a central bank runs quantitative easing (QE) or keeps rates low to make government borrowing easier, the newly created money doesn't flow evenly into the economy - it flows into asset markets (stocks, bonds, real estate) first, pushing those prices up, before gradually spreading to wages and consumer prices. Because asset ownership is concentrated among wealthier households, this mechanism - which analysts outside the mainstream call the "Cantillon effect," though mainstream central banks use the more neutral term "distributional effects of quantitative easing" - tends to enrich asset-holders first, while wage-earners wait longer to see any benefit, if they see one at all.

The Bank of England (BOE), in a study published in 2012, acknowledged as much itself: its asset purchases (QE) had by then boosted the value of UK households' stock and bond holdings by roughly £600bn - and the wealthiest 5% of households held 40% of total non-pension financial assets, meaning they captured the largest share of that gain (BOE Quarterly Bulletin, 2012 Q3). In the US, Federal Reserve data for the COVID-era QE period (2020-2021) is even starker: US households gained more than $18 trillion in wealth in a single year, roughly 80% of it from asset repricing (stocks, real estate) rather than new income - and the wealthiest 1% captured 32.1% of that national wealth gain, while the bottom 50% received just 3.3% of it, even though their percentage gain was faster (Fed, FEDS Notes, 8/2021).

Exit 2: keep rates high for longer - who loses on the other side

If instead a central bank is forced to keep rates high for longer - because massive government borrowing needs are crowding out capital markets, or to contain inflation caused by the earlier money-printing - the burden shifts to the opposite group: those who need to borrow, rather than those who already hold assets.

In the UK, homeownership among 25-34 year-olds fell from 55% (1990) to just 31% (2022-23); renters in the same age bracket spend up to 31% of household income on housing, versus 12% for those still paying off a mortgage and just 5% for outright owners - a fairly clear quantitative answer to who pays more in a high-rate environment (Resolution Foundation, "Housing Hurdles," 12/2024). Small businesses share a similar fate: an NFIB (US) survey found short-term borrowing rates for small businesses ranging 8.3-8.7% in 2025, far above the yield on large, investment-grade corporate bonds (~4.8%). A World Bank study (2024) on public debt's crowding-out effect on private investment in developing countries concludes bluntly that the crowding-out effect "falls disproportionately on small and medium enterprises, domestic firms, and non-exporters" - precisely the group with the fewest alternatives (World Bank Policy Research WP 10786, 2024).

AI: possibly the solution, possibly a new fault line

The hardest variable to forecast in this entire picture is AI - and it could pull in two opposite directions. In the optimistic scenario, if AI genuinely lifts productivity broadly, nominal GDP growth could outpace borrowing costs - the one formula that has ever let a country "grow out of" its debt without a crisis. In the pessimistic scenario, AI displaces labor faster than the economy creates new jobs, pushing a large share of workers to the margins exactly when the budget has the least room left to support the transition - because most of it has already been eaten away by debt interest.

The pessimistic scenario isn't without precedent. The industrial revolution of the late 19th and early 20th centuries produced a similar shock of labor displacement and wealth concentration, and many economic historians see it as one of several indirect contributors to social unrest, the Great Depression, and eventually the two World Wars - though this is a historical analogy at the macro level, not a simple causal chain, and historians still debate how much weight this factor deserves relative to political and geopolitical causes. What's more certain: which direction wins out largely depends on whether governments still have the fiscal space to support that transition - which loops right back to the public-debt problem this piece is about.

The burden on future generations: part of why birth rates are falling too

"The burden on future generations" isn't just an abstract line in a diagram above - economists have an entire method for quantifying it, called generational accounting (developed by Alan Auerbach, Jagadeesh Gokhale, and Laurence Kotlikoff starting in the early 1990s), which measures the lifetime gap between taxes paid and benefits received for each generation. In Japan - the country with the highest public debt/GDP in this piece - studies using this method estimate future generations will bear a lifetime net tax burden 2.7 to 4.4 times higher than the current generation (generational accounting research, Japan). In the US, economist Laurence Kotlikoff estimates the "fiscal gap" (the present value of all future spending-minus-revenue shortfalls) at roughly 7.8% of GDP based on the latest CBO figures - closing that gap immediately would require a permanent ~47% increase across every federal tax.

Just as this future burden becomes clearer, birth rates in most of the high-debt countries covered here are also falling to record lows - all far below the 2.1 replacement rate: Italy at 1.18 (a record low, below even its old 1995 trough), Japan around 1.15, Spain at 1.10, Greece around 1.26, the EU average at 1.34 - the lowest since 2001; South Korea, despite a recent uptick, remains the world's lowest at 0.75 (ISTAT, Eurostat, Korea.net). Direct surveys show economic reasons play no small part: a Pew Research survey (US, 2024) found 36% of adults under 50 who don't plan to have children cite "can't afford it" as a reason; an ISTAT fertility-intentions survey (Italy, 2024) found roughly a third of respondents cited economic reasons, with housing costs - the same topic discussed above - recurring among the top policy-support priorities.

Whatever the exact cause, the reverse consequence is much clearer: low birth rates make the public-debt problem harder on their own. The OECD projects old-age dependency ratios (people over 65 per 100 working-age people) will reach 70.7 in Greece, 76.6 in Italy, 80.0 in Japan, and 84.5 in South Korea by 2054. Across Europe broadly, the number of working-age people per person aged 65+ will fall from roughly 3.4 today to just about 2 by 2050. Fewer taxpayers per person requiring support means a heavier debt-servicing burden on each remaining worker - exactly the point the IMF has flagged: official public-debt figures "may understate the true fiscal challenge" once demographic pressure is factored in (IMF Blog, 10/2024). It's a closed loop: high public debt contributes (in part) to lower birth rates, and lower birth rates in turn make that debt harder to repay for those who remain.

7 · A Framework for Debt: Ray Dalio's Four Levers

If high public debt is a problem, how have countries escaped it before? Ray Dalio - founder of the hedge fund Bridgewater, who spent years studying roughly 48 major debt crises over the past century-plus - offers a fairly tight framework for answering that, laid out in "Principles for Navigating Big Debt Crises" (2018) and, most recently, "How Countries Go Broke" (2025) (Dalio, 2018). Worth stating upfront: this is a practitioner's framework from an investor, not a peer-reviewed academic model formally endorsed by the IMF or BIS - economist Kenneth Rogoff (co-author of "This Time Is Different") calls it "broadly right" but notes it "doesn't cite the academic research that came before it and reached similar conclusions" (Rogoff, book review, 2025). Even so, the core mechanism the framework describes - financial repression, inflation eroding debt - matches the more rigorous academic research (Reinhart & Sbrancia) cited above, making it a useful, accessible lens.

According to Dalio, a government has only four levers to reduce its debt burden - the first two are "deflationary and recessionary," the last two are "inflationary and stimulative":

History has produced all three outcomes:

Applied to today: Dalio is grading the US too

Dalio doesn't stop at theory - he's applied this framework to the US's own current public debt, proposing a "3% solution": bringing the federal deficit from around 6% of GDP today down to 3% of GDP, through a balanced mix of all four levers - roughly 4 percentage points of GDP in spending cuts, 4 percentage points in higher tax revenue, and 1 percentage point from lower real interest rates - which, by his calculation, could leave debt/GDP about 17% lower after 10 years than the current trajectory (TIME, 2025). The proposal was publicly echoed by then-Treasury Secretary Scott Bessent.

But Dalio also offers a notable warning: he doesn't believe the US will default outright - "there won't be a default, the central bank will step in, print money, and buy back the debt" - but unlike the 1930s, when money-printing happened in the middle of a deep recession (hence its "beautiful" effect), this time he worries it will be "printing money into a bubble" - a late-cycle dynamic he considers far more dangerous (Fortune, 10/2025).

8 · Zero/Negative Public Debt: The Singapore Case

If the seven sections above describe the mechanics of who pays as public debt climbs, the reverse question is worth asking too: has any country never stepped into that trap at all? Yes - but the right number to look at matters. Net debt (gross debt minus the financial assets a government holds - reserves, sovereign wealth funds, state-enterprise stakes) is the measure that reflects the real burden, not gross debt - the same logic used to re-read Japan's numbers in section 1. By that measure, a small handful of countries sit at zero or negative net debt - but most of that group are oil states: Norway (Government Pension Fund Global - GPFG - above $2 trillion, 5/2026, roughly 4x GDP), Kuwait (KIA ~$853B, roughly 4.5x GDP), the UAE, Brunei. The mechanism behind this group is fairly simple in essence: an exceptional revenue source (oil and gas) plus a law that bars spending most of it - in Norway's case the 1990 Petroleum Fund Act plus the handlingsregelen rule, which caps annual withdrawals at roughly 3% of the fund's expected long-run return (NBIM) (Meld. St. 7, 2025-2026) - but this isn't really an institutional lesson a country without natural resources can start from.

The far more interesting case, precisely because it doesn't rest on that advantage, is Singapore.

Three legal locks, not a coincidence

Singapore's headline gross debt looks alarming - around 168-173% of GDP (2024, combining SGS, SSGS, Savings Bonds and RMGS) - higher than Italy or the US in the section 1 chart. But unlike those countries, almost none of it funds spending: the Government Securities Act 1992 requires roughly 99% of the proceeds from issuing these instruments to be invested, not spent directly (AMRO, "Much Ado About Nothing?"). SSGS - the largest component - are non-tradable bonds issued specifically to the Central Provident Fund (CPF), the mandatory retirement savings scheme for workers; RMGS transfers a portion of MAS's foreign reserves to the government for longer-horizon investment.

The second lock sits in the Constitution itself: Article 148/148A designates "past reserves" (accumulated before the current term of government) as untouchable without the assent of the elected President, given after consulting the Council of Presidential Advisers; Parliament can only override a presidential refusal with a two-thirds majority. This is exactly the mechanism triggered during COVID-19: President Halimah Yacob successively approved raising the drawdown ceiling to S$49.3B (6/2020), and according to the Ministry of Finance's retrospective figures, the total actually drawn over 2020-2022 came to roughly S$42.9B (MOF Singapore). In other words: when a crisis hits, Singapore spends by drawing down savings it already has, through a public approval process - not by issuing new debt into the market the way most other countries did over the same period.

The third lock is asset management: three separate institutions - GIC (the reserves investment fund, ~$936B, 3/2025), Temasek (the state-enterprise holding company, net portfolio value S$434B, 3/2025) and MAS (official foreign reserves, ~$416B) - together exceed $1.5 trillion, roughly 3-4x Singapore's GDP. Fitch puts the net international investment position (NIIP) at around 150% of GDP - and it's this figure, not gross debt, that all three rating agencies (S&P, Moody's, Fitch) cite when keeping Singapore at the top AAA tier (IMF Article IV Singapore, 7/2025). (Fitch itself uses a narrower measure for gross debt alone - counting only tradable SGS, issued to develop the domestic bond market - and arrives at roughly 40% of GDP for FY2025-26; the gap with the 168-173% figure above is a definitional difference, not an error, the same kind of caveat already flagged for Japan/China in section 1.)

What a positive net balance sheet actually buys

Set against the entire "who pays" machinery described throughout this post - inflation eroding savings, financial repression, austerity - a negative net-debt position buys precisely the opposite. First, crisis firepower without new borrowing: the COVID example above is direct proof - no new debt issued into the market, no exposure to rising yields at the exact moment spending needs peak. Second, near-immunity to the "sovereign multiple-equilibrium" dynamic from section 4: with no large net debt stock needing continuous market refinancing, there's almost no channel for investor confidence to reverse abruptly. Third, pension and aging-related obligations are pre-funded rather than deferred to the future: CPF shifts retirement obligations onto mandatory individual accounts rather than leaving a hidden (pay-as-you-go) liability for the future budget to carry - the same logic Norway achieves through GPFG, just smoothing oil wealth instead of mandatory savings (cf. section 6). One caveat worth separating out: this model doesn't equate to "monetary independence" in the usual sense - MAS runs policy through an exchange-rate band (S$NEER) rather than interest rates, so that's a different axis of benefit, not an automatic reward that comes bundled with negative net debt.

Not a free formula - and not easy to copy

This model has hidden costs. CPF, seen from the worker's side, effectively individualizes retirement risk rather than having the state absorb it collectively: a Wharton Pension Research Council study estimates roughly 40-45% of CPF members fail to hit the "Full Retirement Sum" target, largely because of withdrawals for housing during their working years (Wharton PRC) - a clean government balance sheet is partly the result of pushing that risk down to households. Even the oil-fund group isn't a risk-immune buffer: Norway's fund holds ~49% of assets in equities, and CEPR/DNB research finds that a 10% oil-price move correlates with roughly a 3.0% move in Norwegian equities specifically - meaning the same oil shock threatens both future revenue and the current portfolio's value at once (CEPR VoxEU); and without a hard legal lock like Norway's, a resource fund can erode fast - Botswana is the clearest case, its Pula Fund shrinking from $1.8B (2018) to just $142M (8/2024) as the budget deficit hit 9% of GDP (Bank of Botswana). In other words: both routes - a resource-free city-state's constitutional discipline, or oil wealth tightly locked in by law - are hard to replicate wholesale for any high-debt country from sections 1-7 through political will alone. But Singapore remains the more compelling case of the two: it needed no oil well at all to prove that the burden of public debt isn't a law of physics, but the result of an institutional choice - one that could always have been made differently.

9 · Is War the Solution?

One argument that comes up often: US debt/GDP fell sharply after World War 2, and postwar growth in Germany and Japan ranks among the fastest of the 20th century - so wasn't war, however brutal, an effective "reset" for a system weighed down by debt, bureaucracy, and inequality? That's a question worth taking seriously rather than waving away - and the answer, once each channel is examined on its own, is more complicated than either side usually presents.

First, two phases need separating: while a war is actually underway, debt doesn't fall - it explodes. Britain entered World War 1 with public debt at a historic low (~29% of GDP, 1913-14) and ended the war at 135-143% of GDP (1918-19); Germany entered with debt under 10% of GDP and ended at the equivalent of ~126% of GDP, not counting the 132 billion gold-mark reparations bill imposed under the Treaty of Versailles. In World War 2, US debt rose from ~42% of GDP (1941) to 106-119% of GDP (1945-46); UK debt rose from 135% to roughly 250-270% of GDP. And the intergovernmental debt from World War 1 (the US lent Britain and France roughly $9.5-11bn) plus Germany's reparations bill were never repaid through growth either - they were resolved almost entirely through default: after successive restructurings (the Dawes Plan 1924, the Young Plan 1929) collapsed under the pressure of the Great Depression, when Hoover's debt moratorium expired in 1934, 18 debtor nations - including France, Belgium, and Germany - simply stopped paying. War itself is a tool for creating debt first - whatever "reset" follows comes later, and arrives through very different routes depending on who wins.

World War 2: winners and losers had entirely different outcomes

This is the most commonly misunderstood part. For the US and UK - victors with their economies intact - quantitative research by Acalin & Ball (NBER 2023) decomposes the drivers precisely: strip out budget surpluses, surprise inflation, and the Fed holding rates below inflation (1942-1951), and US debt would have fallen only from 106% to 74% of GDP, instead of the actual 23% of GDP (1974) - meaning most of the debt reduction came from financial repression and budget surpluses (the same Reinhart & Sbrancia toolkit cited above), not "growth from war." In principle, any government could use that same toolkit in peacetime.

For Germany and Japan - the defeated powers - the story is entirely different. Germany's 1948 currency reform (Reichsmark to Deutsche Mark) wiped out roughly 90% of old debt and savings, and it was then cut by another 50% on remaining foreign debt under the 1953 London Agreement. Japan erased most of its wartime government debt through a 1945-49 hyperinflation episode (wholesale prices rose roughly 90-fold) - effectively a disguised default. Both cases share something the US/UK case doesn't: a fully collapsed economy and terms dictated by an occupying power, not a policy choice made by a sovereign government.

The pragmatic case: war as a system reset

Setting the moral question aside for a moment, there are four channels through which war - especially the kind that ends in defeat or occupation - has genuinely produced an economic reset, and all four have concrete numbers behind them, not just speculation.

1. Resetting institutions and bureaucracy. Mancur Olson's argument (1982) is fairly direct: prolonged peace lets organized interest groups - cartels, unions, trade associations - accumulate power and distort the allocation of resources, growing more sclerotic the longer it goes ("the British disease" being Olson's classic contrast case). Military defeat and occupation wipe out that network of vested interests in one stroke. The supporting evidence is fairly striking: West Germany grew at roughly 8% a year through the 1950s (the Wirtschaftswunder), and Japan at roughly 9-10% a year from 1955-1973 - both rising from a rubble of institutions as much as physical capital.

2. Innovation under pressure. The concentrated military R&D spending of wartime produced genuine technological leaps that later spread into civilian use: radar and jet engines (WW2), the first electronic computer (ENIAC, originally built for artillery calculations), mass-produced penicillin, synthetic rubber, and later the network that became the internet's precursor (ARPANET, during the Cold War). Wartime R&D budgets routinely dwarfed what markets would voluntarily spend on high-risk research - "hardship forces the leap" has real grounding in this case.

3. Compressing inequality (the Great Compression). This is probably the clearest quantitative evidence of the four. Thomas Piketty shows Europe's wealth-to-income ratio fell from roughly 6-7 times national income (1910) to just 2-3 times (1950) - a narrowing of the wealth gap that no peacetime tax policy in modern history has matched at anything like that scale. If wealth inequality is part of what high public debt is compounding today (Section 6), this is exactly the kind of "clear the board" outcome the pro-war case points to.

4. A growth and demographic jolt. War spending pulled the US out of the Great Depression at a speed no peacetime policy ever managed - unemployment fell from ~25% (1933) to 1.2% (1944) as the federal deficit hit 27-29% of GDP. What followed was the postwar baby boom - over 70 million American babies born between 1946 and 1964 - expanding the labor force and the future tax base right as postwar debt was being paid down. If falling birth rates are making the debt math harder in most countries today (Section 6), a demographic shock along the lines of post-1945 - through whatever mechanism - is exactly the counterweight that's currently missing.

Checking the evidence: all four channels get weaker in isolation

All four arguments above have real numbers behind them at the phenomenon level - but when researchers isolate each channel from the rest of the postwar context, most don't hold up as a "policy tool" compared with using that same toolkit in peacetime.

Institutional reset: a 1988 study in International Organization found Olson's variables explained only "a negligible part" of cross-country growth differences; Barry Eichengreen argues the opposite - that strong (not erased) corporatist institutions were actually the engine of Europe's postwar "golden age."

Technological innovation: the wartime R&D channel (radar, ENIAC, penicillin) is real, but it's a different mechanism from "creative destruction" through bombing - and it's the latter that gets misapplied most often. Studies using Allied bombing intensity as a city-level "natural experiment" (Davis & Weinstein, AER 2002, on Japan; Brakman et al., 2004, on Germany) both find the most heavily bombed cities simply returned to their pre-war size trajectory, without ever overtaking it; Tamás Vonyó (2018) argues destroyed housing actually slowed recovery by a decade. In other words: military R&D budgets produce real breakthroughs; destroying civilian infrastructure doesn't - two distinct mechanisms that get conflated.

Compressing inequality: correct on the outcome, but the mechanism behind it wasn't orderly redistribution - it was direct physical destruction, inflation, and the collapse of overseas assets. Both rich and poor lost wealth, just at different magnitudes. Measured on cost, it's a vastly more expensive way to compress inequality than peacetime tax and redistribution policy achieving the same result.

The demographic jolt: the post-1945 baby boom happened only in victorious countries with economies still intact - it's not an automatic consequence of war. Today nearly every high-debt economy faces falling birth rates rather than a population boom, including countries that have recently been through conflict - the demographic lever that once amplified the 1945 formula has no built-in mechanism guaranteeing it reappears.

Broad, country-level evidence also paints a considerably less favorable aggregate picture than these four selected examples: an IMF study (Novta & Pugacheva, covering 115 conflicts across 145 countries over 75 years) finds conflict causes prolonged output/investment losses that don't fully recover even a decade later, and pushes countries into financing budgets with inflation - making the debt problem worse, not better; another study covering 160 countries from 1955-2015 (Thies & Baum, Cato Journal, 2020) similarly concludes war is generally a drag on net growth, exceptions notwithstanding. On "military Keynesianism" specifically - the claim that the sheer scale of war spending is itself the stimulus - Barro & Redlick (QJE, 2011) estimate the wartime defense-spending multiplier at just 0.4-0.7, below 1: civilian spending (infrastructure, healthcare, education) at an equivalent scale is likely to produce a stronger stimulus without destroying anything - as Paul Krugman famously joked, a fictional "alien invasion" would end a recession just as fast, and Bastiat's classic "broken window" rebuttal applies just the same.

The reverse direction: can debt crises help cause wars?

Historical evidence points to a real, if not straightforward, link. Keynes, in "The Economic Consequences of the Peace" (1919), warned that the reparations burden would push Germany into a spiral of deflation, unemployment, and instability, predicting the fallout would arrive "some twenty years hence" - close to the actual timing of World War 2. Barry Eichengreen ("Golden Fetters") shows how the gold standard of that era transmitted the depression globally and tied governments' hands against devaluation. Adam Tooze ("The Wages of Destruction") argues part of the Nazi regime's expansionist drive stemmed directly from fiscal pressure - a budget secretly financed through hidden debt instruments (Mefo bills) that pushed the regime to seek resources abroad. Paul Krugman, commenting on America's WW2 debt, has flatly stated the opposite of the "war erases debt" claim: "we never paid off that debt, we just kept rolling it over... the debt/GDP ratio shrank because the economy grew."

So where does the pragmatic calculation land?

The four "reset" channels above are all real at the phenomenon level - which is why the "war helps" argument keeps resurfacing every generation. But once each channel is isolated, most of its value turns out to come from the same peacetime toolkit used throughout this piece: financial repression and budget surpluses (Sections 2, 7), targeted R&D investment, tax and redistribution policy, population policy - the difference is that war bundles in physical and human destruction as an added cost, not a precondition for the result. In the losing case (Germany, Japan), the "reset" didn't come from any policy choice at all - it came from collapse and terms imposed by an occupying power, a path nobody can choose in advance and that can't be reproduced in a controlled way. Add to that the fact that a 1945-style demographic tailwind (the baby boom) doesn't appear automatically, and that today's nuclear powers face mutually assured destruction (MAD), making all-out war between major economies essentially off the table as a strategic option - the cost-benefit math leans fairly clearly in one direction: these four reset channels are worth studying and deliberately reproducing through peacetime policy, which is far cheaper and doesn't require waiting on an event capable of killing tens of millions of people (15-22 million in WW1, 70-85 million in WW2) to "earn" the outcome.

10 · Rising Developed-Market Debt: The Ripple Effect On Emerging Markets Like Vietnam

The nine sections above describe what happens inside the borders of a country with high public debt. But the debt trajectories of Japan and the US don't stop at those borders - both are the largest "borrowers" in the same global capital market that Vietnam and other emerging economies compete in. When they borrow more or pull capital back home, the rest of the world feels it through two channels: a higher global cost of capital, and sudden carry-trade reversals.

The IMF's April 2026 Fiscal Monitor - the same edition cited in Section 1 - calls the first channel the "portfolio rebalancing channel": as US Treasury supply shifts, foreign investors rebalance holdings broadly, not just in the US market - a country in the 75th percentile of foreign holdings of its own public debt sees an output decline roughly 0.2 percentage points larger after 12 months than one in the 25th percentile, following a US Treasury supply shock (IMF Fiscal Monitor, Chapter 1 Online Annex, 4/2026). The same-period IMF GFSR adds: EM portfolio debt has risen from roughly 9% of GDP (2006) to about 15% today, with 80% held by nonbank investors - far more sensitive to interest-rate differentials and global risk sentiment than traditional FDI, and quicker to pull out once carry trades unwind (IMF Blog, based on the GFSR, 4/2026). This is the same mechanism Hélène Rey named in her well-known Jackson Hole 2013 address - "Dilemma, not Trilemma": a "global financial cycle," driven largely by US monetary conditions and risk appetite, that leaves a country unable to insulate itself simply by floating its exchange rate, short of also controlling capital flows (Rey, NBER WP 21162, 2013).

Channel 2: capital repatriation when carry trades unwind - the JPY case, August 2024

The sudden version of the same story has a recent precedent. For years, near-0% JPY rates funded a global carry trade worth roughly ¥40 trillion (~$250bn) by BIS estimates - borrow cheap yen, invest where yields are higher. When the BOJ unexpectedly raised rates to ~0.25% (31/7/2024) just as a weak US jobs report landed, the carry trade unwound violently: USD/JPY fell from ~161 to 141.7 within three weeks, and the Nikkei 225 dropped 12.4% in a single session on 5/8/2024 - the worst since 1987 (BIS Bulletin No. 90, 8/2024). Vietnam wasn't at the center of it - the VN-Index still closed 2024 up 12.1% - but the interbank USD/VND rate stayed elevated at 24,200-24,600 throughout the episode, reflecting the same mechanism that can reverse as fast as it did with the yen, just at a smaller scale.

That loop is now closing back on Sections 1 and 3: the BOJ has since raised rates to 1.0% (16/6/2026, the highest since 1995) to fight inflation and yen weakness (CNBC, 6/2026), while Japan's Ministry of Finance has net-sold roughly ¥4 trillion (~$25bn) of foreign securities since the start of 2026 - meaning Japan itself is now repatriating capital, adding further upward pressure on global yields (TD Economics).

The Dollar Smile: why the dollar can strengthen at both extremes

The Dollar Smile framework (Stephen Jen, Morgan Stanley, 2001-2002) - the dollar strengthens both when the world panics (flight to safety) and when the US outperforms (capital gets pulled back), and weakens in the "middle of the smile" during calm, synchronized global growth, exactly the window when capital tends to flow into EMs - is covered in full in a separate piece, "When The World Shorts The USD: America Hurts, But The Rest Of The World Hurts More". Worth adding here: that framework just "broke" once, during the April 2025 tariff shock, when the dollar fell instead of rallying even as the S&P 500 dropped ~11% over three sessions - something Stephen Jen himself acknowledged in a November 2025 note, describing the US as heading "into the trough of the Dollar Smile" (Eurizon SLJ Capital, 11/2025). For Vietnam, that cuts both ways: a prolonged weak dollar is a tailwind for EM inflows, but the classic risk-off end of the "smile" can still be triggered at any time.

11 · The Bottom Line

High public debt rarely kills a major economy in a single shock - it usually squeezes gradually: crowding out the budget, eroding savings through inflation, shifting the burden to future generations (even contributing to falling birth rates), and forcing austerity that often backfires - all while quietly widening the wealth gap through either monetary-policy exit, just as AI could reshape the labor market in either direction. Italy, the UK, and increasingly France show that this "gradual impoverishment" scenario isn't a hypothesis - it has already happened, and is happening now. And even when that process unfolds calmly over many years, the bond market retains the right to suddenly demand a rate of return that matches the real risk - as the UK in 2022 and France in 2024-2025 both demonstrated. History also shows a way out isn't impossible, and certainly doesn't require war - it just requires the right balance of austerity, debt restructuring, money-printing, and redistribution, rather than relying on a single lever or an external shock. For a country like Vietnam, the story isn't about a low public debt/GDP ratio - which is indeed genuinely low - but about maintaining an FX reserve buffer thick enough relative to M2 and short-term external debt, exactly the measure that historical EM crises (Mexico 1994, Asia 1997) show to be the decisive one - not the public debt/GDP ratio, which was designed for advanced economies. And as Section 10 shows, that's no longer an abstract principle: the debt trajectories and interest-rate policies of the US and Japan are now directly shaping the cost of capital and the carry-trade flows that buffer has to absorb.

12 · A Personal Take: How Savers And Individual Investors Should Read This Picture

The eleven sections above tried to stay descriptive - data, mechanisms, historical precedent. This closing section shifts to a more specific angle, limited strictly to the household and individual saver/investor level - not what governments should do. It still holds to the disclaimer at the top: this is a macro-reading principle, not a specific asset-allocation recommendation.

A "risk-free" nominal rate no longer means real purchasing-power safety

If the primary channel for working off debt is inflation or financial repression - exactly the Reinhart & Sbrancia mechanism from Section 2, which once "liquidated" 3-4% of GDP a year in the real value of savings in the US/UK after World War II - then the nominal rate printed on a savings account or government bond is no longer a sufficient safety measure. The question worth asking about any long-term savings is the real return after inflation, not the nominal figure - especially for money left untouched for years (fixed-term deposits, savings-linked life insurance, fixed-rate pension funds).

Long-duration bonds aren't as "boring" as they used to be

The multiple-equilibria mechanism from Section 4 (Calvo) has a direct consequence for bondholders: yields can sit still for years, then jump within a few sessions - as UK 30-year gilts did, up 150 basis points in four days in September 2022, or US 10-year Treasuries, which jumped from 5.3% to 8.0% over 13 months in 1994. The BIS quantifies this in Section 5 too: the probability of a bond-market stress event is roughly 10 times higher when public debt is high versus low. For anyone holding long-duration bonds - directly, or through bond funds or insurance products - this means duration risk is more real and less predictable than during the earlier stretch of low, flat rates. It's a reason to know exactly what duration one is holding, rather than assuming government bonds are automatically a "quiet" asset.

Diversify sovereign risk, not just asset classes

Central banks - some of the most risk-averse institutions in the financial system - are already doing exactly this: official global gold reserves rose by a record 1,237 tonnes in 2025, and the World Gold Council's 2026 survey found 45% of reserve managers expect to keep adding gold over the next 12 months, with 84% believing gold's share of reserves will rise further by 2031 (World Gold Council, Central Bank Gold Reserves Survey 2026). This isn't a recommendation to buy gold - it's a signal worth noting: when the institutions with the deepest visibility into sovereign risk are actively shifting weight away from dependence on any single currency or government bond, that's a data point worth weighing against how concentrated one's own portfolio is in a single country or currency.

Don't assume the public pension system will cover the gap

Section 6 laid out two facts worth remembering: generational accounting shows Japan's future generations could face a lifetime net tax burden 2.7-4.4 times higher than today's, and old-age dependency ratios in many high-debt countries are set to climb sharply by 2054 (70-84 people over 65 per 100 working-age people). For anyone living in - or drawing retirement income from - a high-debt, rapidly aging economy, this is a reason not to assume the public pension will keep replacing the same share of income it does today. Building an independent personal retirement cushion, rather than relying entirely on a pay-as-you-go system, is a reasonable hedge against exactly this demographic risk.

A few gauges worth watching, instead of trying to predict a specific doomsday

• Long-term bond yields in one's own country - the speed of change over a few sessions matters more than the absolute level.
• The trend in the primary budget deficit (excluding interest) - a shrinking primary deficit signals recovering fiscal space; a widening one is the more telling warning sign.
• For anyone in an emerging economy like Vietnam: the ratio of FX reserves to short-term external debt - exactly the measure flagged in Section 1, not public debt/GDP, which fits advanced economies better.

No one can predict exactly when the bond market will flip from patient to impatient (Section 4) - history shows that can happen in a single afternoon, after years of calm. But what's actually within a saver's or individual investor's control isn't predicting that date - it's holding duration, sovereign/currency diversification, and retirement assumptions that can be adjusted starting today, regardless of which direction the global public-debt trajectory ultimately takes.