"China is digging out of a crisis. And America’s luck is wearing thin." — Ken Rogoff

Summary

• China is in a deep crisis and stays there “for five or ten years,” per Rogoff: real estate is a third of the economy by some measures, tier-three cities carrying 60% of income have “the feel of death,” and falling prices plus rates being pushed to zero are “signs that demand has been crushed.” He doesn’t believe the official 5% growth print — and his projection that China gains only ~1 point a year on US nominal GDP implies it may never overtake the US, far more bearish than even the pessimist consensus of 125-150% of US GDP by 2040.
• America’s debt endgame is inflation, not default or a Greek-style crisis — “I’m talking 10–20% inflation over a period”; the recent bout knocked ~10% of GDP off the debt and “we might need more next time.” When it recurs, markets will say “You are not to be trusted,” rates rise, debt compounds faster — then austerity, per Churchill: Americans “do the right thing after we try everything else.”
• Fed independence is overtrusted — despite the Supreme Court, “I think they’re dreaming. There are so many ways Congress and the President could override the Fed,” especially under a declared wartime or “war-on-pandemic” shock. Meanwhile the 10-year at 4.5% and 30-year near 5% are “going to drift up.”
• Long rates are more likely to go up, and AGI may not rescue the fiscal picture — AI’s energy and capital demand plus global debt, remilitarization, and climate push real rates above the 2012–21 zero average. “Nobody ever defaulted or had high inflation because of arithmetic… They do it because of political pressures” — and fast AGI “would make the populism phenomenon we’re facing now seem like nothing.”
• Tradeable call: foreign equities beat dollar equities; when the dollar is really high, the euro should go up — Europe has catch-up room, remilitarization helps its tech and currency, and it’s “the team that doesn’t have as many injured players.” The dollar’s “gentle decline” predates Trump — “it would’ve happened with Harris winning.”
• The Taiwan tell is in the rails, not the reserves: China is moving into gold, holds ~$2 trillion of Treasuries via proxies (double the official figure), and is building its own payment mechanisms because “they can’t live without being able to pay suppliers” — Russia did the same before invading.
• Financial crises scar permanently — Japan is maybe 50% poorer per person than the no-crisis counterfactual, the US “probably 15% lower” from 2008 even now; the Plaza Accord mea culpa (“I have ruined our country”) flipped Rogoff’s own view: “financial liberalization needs to be done gradually.”
• The meta-thesis: America has been lucky as well as good — rivals’ blunders did much of the work — and “I worry our luck is wearing thin.”

Deep dive

1. Xi dismantled the technocracy — and both superpowers got less competent

• Rogoff’s inside view from lecturing at the Party’s training school: Chinese technocrats “actually asked really raw questions,” the school allowed anything to be said, and leaders “listen to everybody… way better than we are at hearing a hundred different views.” Xi Jinping changed that since 2013 — “pushed out that system and moved more toward loyalists, people who are less technocratic.” His well-connected contacts had promised “He’s going to be Ronald Reagan… we’re going to liberalize.” He didn’t, and growth slowed markedly on his watch.
• The 2016 China Development Forum story: in front of top leaders and tech elites, Rogoff said “you’re going into a classical housing crisis problem. Your catch-up is over. Your demographics don’t look good… power is becoming very centralized.” He figured “you only live once” — and braced for jail when leaders approached, only to hear “Professor Rogoff, we very much appreciated your remarks.” They invited him back — to “a tiny room instead of the big hall.”
• The strategic worry: “I’m less impressed by them now… we’re not as competent [either]. The average quality at the very top, I think, has gone down. And China’s not as competent either. That’s a recipe for having bad things happen.”

2. Overbuilt supply, crushed consumer — the anatomy of China’s crisis

• The seeds: Hu Jintao’s 2010 stimulus created local-government debt; land sales let local governments start and fund construction companies, and they left the system running as a stimulus program. Xi inherited it but kept it going.
• The result in tier-three cities — which generate 60% of Chinese income: “amazing roads, amazing real estate, amazing housing. But the feel of death in those cities… The Soviet Union was very good at building cement factories and steel plants and railroads. But they’ve run their course.” Green energy, AI, EVs? “That stuff’s still tiny compared to infrastructure and real estate” — real estate is a third of the economy by some measures.
• Dwarkesh’s corroborating anecdote as told: a town of half a million outside Chengdu with a huge train station and a freshly built Buddhist temple, concentric shrines “for like eight turns… There was just nobody there. It was like me and three other white people.” Rogoff: young people don’t want to live there, and the jobs aren’t there.
• The core imbalance: savings maybe 45% while US consumption pushes 70%; “a lot of China is living on $200-a-month kind of incomes,” with no social security or health system pushing precautionary saving — and housing, the only real savings vehicle besides crummy bank deposits, is collapsing. “There’s no magic bullet to make them grow at 5%. By the way, that is the official number but I don’t think they’re anywhere near that.”

3. The bearish projection: China gains 1 point a year — maybe never overtakes

• The measurement caveat first: official growth 1980–2012 was almost 10%; on purchasing-power terms “just over 7%.” Xi-era official figures of 6-7% may really be maybe only 3.5% — and “historically they have given numbers which are accurate on average… that’s gotten less and less true in the Xi Jinping era.”
• The forecast: Dwarkesh put China at roughly 75% of US nominal GDP; Rogoff thought it was a little lower and was actually going to say 75% in 2030. At one point in 2024 it was around two-thirds, exchange-rate volatile; he expects China to gain “about a percent a year on us, maybe.” Dwarkesh’s pushback — that means they never actually get bigger, despite 4x the population, which is extremely bearish even versus China pessimists. Rogoff holds: “It’ll take a long time.” Goldman-style extrapolations “were proven wrong… Economists at least consider ourselves terrible at that.”
• Dwarkesh’s sharper pushback on PPP: for war, doesn’t cheaper shipbuilding and cheaper soldiers make PPP the right lens? Rogoff concedes fully: “You’re absolutely right. They just crush us in shipbuilding” — ~50% of the global market, commercial-military symbiosis — and the US mistake is building everything itself instead of importing from allies like Korea. “We’re way ahead in your department, tech… If that were to dissipate, it would certainly hurt.”
• Ground truth from Dwarkesh’s China trip: VCs there are depressed — “even if you invest in the next Alibaba, who’s to say the government doesn’t cancel the IPO?” Rogoff: “They’re in trouble… for five or ten years they’re going to stay in trouble.”

4. Watch the rails, not the reserves, for Taiwan preparation

• On reserves: China is “definitely moving more and more into gold” — euros and Canadian dollars don’t help “because those countries might side with us.” Not crypto yet, per a student’s paper. Officially $1 trillion in Treasuries, but a student estimate Rogoff endorses puts it at more like $2 trillion held through proxies.
• The real constraint: “They could live without their $2 trillion for a little while. But they can’t live without being able to pay suppliers” — the US disproportionately controls the pipes of global payments, so China is building its own rails, as “Russia actually did quite a bit in preparation for the invasion.”
• Nothing sudden: “they don’t want to be the ones to bring down the markets and cause a global crash.” And it’s not just China — Europe’s central bank digital currency is “moving quite a bit faster than I thought it would,” partly for international payments.

5. Rogoff changed his mind — the Plaza Accord may have caused the crisis

• His stated reversal, worth keeping verbatim in spirit: he and Carmen Reinhart date Japan’s crisis to 1992, seven years after the 1985 Plaza Accord, so he long dismissed the causal link — “but over the years, particularly recently, I’ve started to think I was wrong. These things unfold slowly… It was a huge mistake for Japan to agree.” The general law: “Financial repression is bad. But financial liberalization needs to be done gradually. If you do it too quickly, you get a crisis.”
• The scene that carries it: at the Plaza Accord’s 10th anniversary in Tokyo, the 1985 Bank of Japan head bowed and apologized — “I have ruined our country. I did this. I take responsibility.”
• China learned the lesson early: Jiang Zemin, on Rogoff’s exchange-rate advice — “That’s what the United States told Japan. Look what happened to Japan.”
• The counterfactual cost: Japan would be “50% wealthier per person” without the crisis (he later hedges to maybe 25-30%) — it was richer than the US at market rates in the late 1980s, richer than any European country; “they’ve moved to the bottom of the rung now.” The crisis “blew up their business model” of export-led growth just as China rose.

6. Financial crises aren’t recessions — they scar for decades

• The This Time Is Different vindication: he and Reinhart argued financial crises last way longer and cut way deeper than normal recessions — “we were mocked… I think the New York Times had a two-page spread saying how ridiculous everyone thought this was.” The US thought itself immune, having had none since 1933.
• The mechanism, via Bernanke’s “thought piece”: Friedman said the Depression was too-tight money — but Bernanke asked, if that were all, wages and prices would adjust in a year or two; “the Depression took 10 years. How can that be?” The credit system itself broke — lending to innovators and dynamic firms stops working.
• America’s own bill: without 2008, US income today is “probably 15% lower… It led to this political crisis that caused us to lose a lot more.” Dwarkesh’s update: “if there’s 15% lingering even after almost 20 years, then wow that’s huge.”
• The policy residue: “I would describe Treasury and Federal Reserve policy today as, ‘When in doubt, bail it out.’… As your financial sector grows, that will lead to a problem someday” — SVB was an echo.

7. The US endgame: inflation lets off steam, then austerity

• The setup: “you have a crisis of some sort when your debt is high and your political system is inflexible. We’ve checked those boxes. Then you get hit by a shock you weren’t ready for.” Of the four exits — default, financial repression, austerity, inflation — default is out (“we can print money”), and Japan-style repression (BOJ holding ~100% of GDP in debt, a $30 trillion equivalent vs the Fed’s ~$7 trillion) is tough for a market-driven US: “We can’t force French insurance companies to hold US debt.”
• So: “the most likely thing will be inflation… 10–20% inflation over a period. We just went through that. That actually knocked about 10% of GDP off our debt. We might need more next time.” But next time “markets will be very unforgiving… They’ll look at us and say, ‘You are not to be trusted.’” Then rates rise, debt builds faster — and eventually austerity, per the Churchill line about Americans doing the right thing “after we try everything else.”
• His hedge and his jab, both preserved: this is “more likely than not, not that it’s definitely going to happen” — not the end of the world (Europe raised retirement ages in 2010-12) but “pretty unpleasant,” with the dollar franchise eroding and rates climbing as others use it less. And on language: “austerity” is what progressives call it “when you don’t” spend whatever you want — the framing “pretends there are no costs to having your debt be higher.”

8. Markets trust Fed independence too much

• The number-one market irrationality: “too much faith in the independence of the Federal Reserve.” Despite the Supreme Court ruling Trump can’t fire Powell, “I think they’re dreaming. There are so many ways Congress and the President could override the Fed, especially if they declare some kind of wartime or ‘war-on-pandemic’ situation.” And this from the man who “wrote the first paper on why you should have an independent central bank.”
• Dwarkesh’s angle — the Fed as a buck-passing device politicians secretly like — Rogoff accepts: “That’s for sure. That’s why Trump bashes the Fed… it gives him someone to blame.” He’s now hearing tech titans, not just progressives, say “Scott Bessent… he’s smarter than Powell. Why don’t we let him run things? They could.” The Turkish counterexample: Erdoğan firing central bank heads annually, inflation hovering toward 100%.
• The Fed’s secret is one barometer — inflation — and it has “managed to keep their core competency” despite mission-creep pressure (Fed working papers were “all about inequality, the environment, social justice”). The profession’s own miss: a Hoover analysis of AEA meeting abstracts found “the word inflation had not appeared until this year” in 15 years. Rogoff was “a lone voice in the wilderness” — teaching inflation “was like I was teaching them the music of Fred Astaire.” But the “intellectual market” and ruthless seminar system are now rebalancing.

9. AGI may not fix the fiscal problem — and rates are more likely to rise

• The claim: “Nobody ever defaulted or had high inflation because of arithmetic, because they couldn’t pay… They do it because of political pressures.” A productivity boom helps, but countries with growth above their interest rates still blew up — and AGI arriving fast “would make the populism phenomenon we’re facing now seem like nothing.”
• Dwarkesh’s best exchange: if AI removes downward wage rigidity — Keynes’s cornerstone puzzle of why prices didn’t fall in the Depression — should the Fed even fight deflation? Rogoff concedes, “that is a very good point… if you have these docile AI workers… and firms willing to let prices fall, then certainly you can do that.” If real rates are rising anyway, deflation stops being a technical problem — you just let rates rise a little less.
• On direction: “AGI and AI are upward pressures on interest rates” — huge energy needs, and if AI substitutes for workers it makes capital more valuable so you invest more (citing Acemoglu that it can go both ways). But AGI is “only a piece of it”: debt rising everywhere, remilitarization, climate, populism. The 10-year real rate averaged zero 2012–2021, hit -1 post-pandemic, is higher now — “for a macroeconomist, the biggest question in the world” — and he calls the rise “just a normalization,” more likely to continue up than down.

10. Exorbitant privilege is fantastic, Europe is the healthy team — and the luck is running out

• Against the it-hurts-us argument (Stephen Miran’s “clever” hollowing-out thesis gets “a little bit of truth”): the privilege is “incredibly fantastic if you owe $37 trillion… to be paying half a percent to a percent less. We’re talking about hundreds of billions of dollars” — plus surveillance through the dollar network and sanctions “in place of military intervention.” Kindleberger’s frame: America is “bankers to the world… making money hand over fist” — borrowing in safe assets, investing in risky ones.
• The concrete call: “foreign equities do better than dollar equities”; when the dollar is really high, “you should expect the euro to go up” — Europe has catch-up room, and re-militarizing “would actually be good for the euro… good for technology in Europe.” His basketball analogy as told: like the Celtics losing to the Knicks with Jayson Tatum injured — “You may not have gotten any better, Europe in this case, but if somebody’s hobbling the United States — I do think that’s going on to some extent now — you do better.”
• The book’s core warning: US dominance rode on rivals’ blunders — Japan’s forced liberalization, China “sticking to the dollar so long,” Greece let into the euro too early. Bent Larsen, asked whether he’d rather be lucky or good at chess: “Both.” “We’ve also been lucky… I worry our luck is wearing thin.” When he pitched the thesis to academics, financiers, and tech people, “They said, ‘You’re nuts.’ They didn’t want to think about it.”
• Dwarkesh’s closing reframe — worth keeping: if great powers fall into ruts this easily, it’s “like the Fermi estimate thing… it suggests that there’s some kind of filter.” And Rogoff’s rule-of-law coda: foreign investors used to know they’d get their money back in America — “That’s in doubt now. There’s no question.”