2026-08-13 ATS briefing¶

The Next Global Economic Crisis Could Be Made in China¶

The arithmetic of Chinese overcapacity has finally caught up with its politics. For two decades the arrangement worked because it was mutually convenient: Beijing offloaded manufactured goods at prices no competitor could match, and the rest of the world enjoyed the disinflationary benefits of cheap imports while outsourcing its industrial base. That bargain has now inverted. With a trade surplus approaching $1.2 trillion in 2025 — three times the growth rate of global goods trade — China has simply run out of customers willing to absorb the output. The political tolerance for deindustrialisation has evaporated, and protectionism is rising precisely as Beijing's "dual circulation" strategy requires open markets to function.

What makes this crisis qualitatively different from previous rounds of trade friction is the mechanism driving it. This is not a story of market competition but of state-directed credit that has severed the link between investment and return. Nearly 30 percent of Chinese industrial firms operate at a loss, rising to 34 percent in the priority sectors of "Made in China 2025." Local governments prop up unprofitable factories to preserve employment and tax revenue; state banks roll over debt for insolvent borrowers. The Chinese themselves have a word for it — neijuan, involution — describing firms that invest more to produce more to export more at negative margins, trapped in a race that cannot stop and cannot slow down.

The contradiction is concrete: Beijing's growth model requires ever-expanding export markets, but the very scale of its surplus — now greater than Germany and Japan's combined at any point in the 1980s — has destroyed the political conditions for accessing them. The leadership recognises the problem, hence the 2025 anti-involution campaign and the 15th Five-Year Plan's rhetorical shift toward consumption. Yet the fundamental orientation remains unchanged: suppress domestic demand to maximise industrial output. The result is an industrial machine that cannot be rebalanced without triggering the domestic reckoning Beijing fears most, but cannot keep running without triggering the global one Froman describes. For the rest of the world, the question is whether the shock arrives as a managed slowdown or a collapse — and whether the United States, the only actor with the institutional capacity to stabilise the system, will choose to do so.

The AI Growth Paradox¶

Source: Foreign Affairs

The magic of the AI growth story is that it lets Washington postpone an accounting it has no intention of doing. Rogoff's framing is useful precisely because he refuses the usual terms of the debt debate: the question is not whether $40 trillion is "unsustainable" — a word that has meant nothing for two decades — but what happens when the growth that is supposed to make the debt manageable arrives with a price tag attached.

The mechanism is straightforward. AI-driven productivity gains would, in theory, expand the tax base. But the same boom that generates those revenues will also push up real interest rates, as capital competes for the investment opportunities the technology creates. The ten-year inflation-indexed Treasury yield has already moved from roughly zero in the 2010s to over 2.4 percent. Servicing the debt becomes more expensive at exactly the moment the government is tempted to borrow against anticipated future windfalls. Rogoff's historical point lands with force: the post-war debt reduction worked because financial repression held rates artificially low, an implicit tax on small savers. That mechanism is unavailable now, and the "secular stagnation" orthodoxy that justified endless borrowing has been quietly retired.

What Rogoff does not say, but the material invites, is that this is a class question dressed as a technical one. The debt is not an abstraction; it is the accumulated record of tax cuts and spending decisions made by governments that never faced a political cost for them. The "magical thinking" of the 2010s was not an intellectual error but a rational response to a structure in which the wealthy could borrow cheaply, the state could spend freely, and the bill was deferred. AI now offers a new deferral mechanism: if the technology delivers, the problem solves itself; if it does not, the crisis arrives with interest rates already elevated and no fiscal ammunition left. The paradox is that the more credible the AI boom becomes, the more dangerous it is — because it licenses the very overspending that will turn a productivity miracle into a sovereign debt event.

China’s Biggest Weakness on AI¶

Source: Project Syndicate

The Chinese state is trying to have it both ways with AI, and the strain is showing. Courts rule that automation is not grounds for dismissal, regulators suspend robotaxi licences after a Wuhan failure, state media lectures companies on their duty to protect workers. Each intervention is a small admission that the social costs of AI cannot be left to the market. Yet none of them addresses why those costs are so acute in the first place.

Frey's argument is that China's thin welfare state is the decisive constraint. A state that cannot credibly promise that displacement will not mean destitution must instead block the displacement itself. The result is a brake on the very productivity gains the AI race is supposed to deliver. The court ruling and the licence suspension are not acts of Luddite caution; they are the price of legitimacy when the safety net is a family savings account and a local government's forbearance.

The comparison with the West is instructive, though not in the way Frey frames it. He implies that a robust welfare state is the precondition for accepting technological change. But the US and Europe are not accepting it either — they are simply letting the costs fall where they may, and calling the resulting political backlash a separate problem. China's approach at least has the virtue of honesty: it knows that AI-driven unemployment is a political question, not a technical one. The court in April was not protecting workers out of sentiment; it was protecting the social contract that keeps the system running.

The deeper difficulty is temporal. A welfare state is built over decades, while AI deployment is measured in quarters. China cannot tax its way to a Nordic model before the robotaxis arrive. So it improvises — case by case, city by city — and each improvisation signals to investors that the state will intervene unpredictably. That uncertainty is itself a cost in the AI race, one that no amount of chip manufacturing capacity can offset. The weakness is not that China lacks the will to manage disruption; it is that the tools for managing it were never built, and the time to build them has run out.

Shipping Number of the Week , China’s iron ore mining cools as imports rise 6% and steel production weakens¶

Source: Hellenic Shipping News

The numbers here tell a story that runs against the usual logic of import substitution. China is digging less of its own iron ore — down 7% year-on-year in the first half — precisely because foreign ore is better and cheaper. Domestic mining is being outcompeted by Australian and Brazilian product, and the share of imports in total supply has crept from 50% to 57% in four years. This is not a supply crisis forcing China abroad; it is a deliberate preference for higher-grade ore that makes steelmaking more efficient and, presumably, more profitable.

But the demand side is where the tension sits. Steel production fell 3%, and the Basic Oxygen Furnace route that actually consumes iron ore dropped around 4%. Port inventories stayed elevated. So the shipping market is booming — the Platts Capesize T4 Index up 79% — on the back of a commodity whose end-use is weakening. The growth is partly a mirage of distance: Guinea, Liberia and Peru are sending ore over much longer hauls, and the Simandou project alone could eventually add 120 million tonnes a year. Freight demand rises even when volumes barely move, because the tonne-miles stretch.

The second half looks shakier. The 2025 baseline was inflated by inventory building that will not repeat, and property-sector weakness is not going to lift steel demand. What is striking is how little of this is about China's own economic cycle. The shift to imported ore is a structural decision, and the shipping industry is now hostage to it. If Chinese mills keep preferring foreign ore, the capesize market stays supported regardless of domestic steel output. If the property slump deepens and steel demand collapses, the longer sailing distances only delay the reckoning. For the dry bulk sector, the question is whether it has built its current strength on a preference that could reverse as quickly as it emerged.

The Twilight of the ICC¶

Source: Project Syndicate

The ICC’s warrants for Putin and Netanyahu mark a genuine departure from its historical role as a tribunal for the defeated. But Ben-Ami’s framing — that this boldness “may well have incited its downfall” — flatters the Court with a causal power it does not possess. The ICC did not stumble into irrelevance by overreaching; it was built to fail in a world of Machtpolitik, and its recent assertiveness merely made the architecture visible.

The Court’s original function was to administer justice after the fact, against leaders whose states had already been crushed or coerced. That was never impartiality; it was the legal arm of the post-Cold War settlement. The warrants against Putin and Netanyahu break that mould because they target leaders of states that still possess the capacity to resist — Russia militarily, Israel through its patron. The Court’s statute assumes a universal sovereign that can compel compliance. No such sovereign exists. What the ICC actually discovered is that its authority is only as strong as the great powers’ willingness to enforce it, and that willingness evaporates precisely when the warrants point at allies or adversaries with real leverage.

The deeper problem is that the Court’s legitimacy was always borrowed from the very power structure it now presumes to judge. When it acted as victors’ justice, it was protected by the victors. When it acts against them, it loses that protection without gaining any independent enforcement capacity. The result is not a tragic fall from grace but the exposure of a founding contradiction: a court that claims universality while depending on the particular interests of the strong.

For those watching the crisis of the international order, the ICC’s twilight is less a legal story than a measure of how far the post-war liberal framework has decayed. The Court’s paralysis is not a bug in an otherwise functional system; it is the system’s honest self-portrait.

Europe Failed the Ceuta Test¶

Source: Project Syndicate

The Ceuta episode is instructive precisely because it was over before it began. Seventy thousand people attempting a crossing in a single surge — the number itself is less important than what it did to the EU's internal architecture. Twenty-two member states demanded an emergency meeting, and before that meeting could convene, Italy had already suspended visa-free travel from Spain. The crisis was not managed; it was pre-empted by unilateral action. That sequence tells you everything about the actual hierarchy of power within the bloc.

Laïdi's framing — that hardline policies cannot address the "deeper forces" driving migration — is correct but incomplete. The deeper forces are not mysterious. They are the structural byproducts of the EU's own economic model: the systematic underdevelopment of North Africa's productive capacity, the agricultural subsidies that destroy local markets, the labour market segmentation that makes European economies dependent on a reserve army of undocumented workers they refuse to formally admit. The contradiction is not between humanitarian values and border security. It is between the EU's need for cheap, flexible labour and its political need to appear sovereign over its borders. Every crackdown produces the next surge; every surge produces the next crackdown.

What Ceuta revealed is that this contradiction is now being resolved at the expense of the EU's internal cohesion. Italy's move was not aimed at Morocco. It was aimed at Spain, and through Spain, at the Commission. When member states start treating each other as migration threats, the Schengen system — the bloc's most tangible achievement — becomes a fiction maintained by mutual suspicion rather than shared interest. The far right does not cause this dynamic; it exploits it, and it is winning because the centre has accepted the far right's premise that the problem is one of enforcement rather than political economy.

For those watching from the left, the lesson is not that the EU needs a more humane migration policy. It is that the EU's internal contradictions are now producing the kind of fragmentation that makes unified capitalist governance of the Mediterranean impossible. That fragmentation is an opportunity, but only if there is a political force capable of articulating what the EU cannot: that the movement of people across this sea is not a crisis to be managed but a demand to be met.

Why tariffs are not the right tool to bolster US production¶

Source: Al Jazeera

The tariff programme is being sold as a reindustrialisation strategy, but the internal logic is already coming apart. Trump wants tariffs to do two things at once: protect domestic producers from foreign competition and raise revenue for the state. Those goals collide. A tariff that successfully shields US industry reduces the volume of imports, which shrinks the tax base the tariff was meant to feed. The more effective the protection, the less money it collects. The administration is effectively trying to run the same policy as both a wall and a tollbooth.

The deeper problem is that tariffs treat the symptom of deindustrialisation while leaving its cause untouched. US manufacturing did not hollow out because foreign goods were unfairly cheap; it hollowed out because capital found higher returns elsewhere — in finance, in technology, in the extraction of value from already-built infrastructure. Tariffs raise the cost of imported inputs, which hits the very manufacturers they are meant to help. A steel tariff protects US steel mills but punishes US carmakers who buy steel. The policy is a transfer between capitals, not a restoration of production.

Even the economists quoted in the piece who sympathise with the goal concede the results are poor. That is telling. The tariff is not a miscalculation; it is a political gesture dressed as economic policy. It offers the appearance of state intervention while leaving the underlying allocation of capital untouched. For the working class, the outcome is predictable: higher prices on consumer goods, no guarantee of jobs returning, and a state that has spent its political capital on a policy that cannot deliver what it promises. The real question is not whether tariffs work, but why the state feels compelled to pretend they might.

Oman says oil spill from stricken tanker has reached its coast¶

Source: Al Jazeera

The Caroline Bezengi has been sitting disabled off Oman’s Al Hallaniyat Islands since June, leaking crude onto a coastline that includes a marine reserve established barely a year ago. The spill has now reached 40km of mainland beach, and the slick is concentrated around the archipelago. The tanker is part of Russia’s shadow fleet, flagged in Cameroon, owned and operated by two China-based firms, and sanctioned by the EU, UK, Switzerland, Canada and Ukraine. Nobody is responsible for it, in the legal sense that matters.

That is the real story here, and it is not primarily an environmental one. The “polluter pays” principle that underpins Western oil-spill regulation assumes a solvent, identifiable owner who can be sued into financing a cleanup. The shadow fleet was constructed precisely to defeat that assumption. A tanker built in 2001, transferred through opaque registries, crewed and insured through layers of intermediaries, is a vessel designed to externalise every cost of its operation — including the cost of sinking. The 800,000 barrels on board are not the point; the structure that made them uninsurable and unaccountable is.

The Omani authorities’ response — monitoring, containment, “priority to areas of high environmental sensitivity” — is the response of a state trying to manage a disaster whose authors have engineered themselves out of reach. Greenpeace’s warning of an “unprecedented oil disaster” is accurate but incomplete: the unprecedented part is not the size of the spill but the fact that the entire apparatus of liability has been hollowed out. Sanctions have made the tanker untouchable for Western firms, but they have not made it disappear. It just sits there, leaking, while the legal and financial architecture that would normally compel a cleanup has been deliberately dismantled by the very governments now expressing alarm.

The collision is concrete: a marine reserve, established to protect turtle habitats, versus a vessel whose entire commercial logic depends on being beyond the reach of any authority that might hold it to account. The spill is the price of that logic, paid by the Omani coast.