Deal or Defence: Europe Reads China's Economic Security Turn

Opening

Welcome to today's China report. Today it's China and Europe, and specifically the question Brussels says it has to answer within weeks: cut a trade deal with Beijing, or start defending the single market on its own. Two reports, both from late September twenty twenty-six. One from the Mercator Institute for China Studies in Berlin, on what Chinese economic security doctrine actually is. One from the German Marshall Fund, on what Europe would have to bring to the table.

Two is all we got this week, which is thin, and a quiet month in the literature is worth noticing on its own. And one thing plainly up front: the MERICS report isn't really about Europe. It's about China's own machinery, and Europe shows up in the closing section. It earns its place because it describes the doctrine the German Marshall Fund piece is reacting to. But I won't stretch it into claims it never made.

MERICS on China's economic security turn

The argument is that Beijing has switched from defence to offence. Economic security used to mean making China harder to hurt. Now, they say, it also means making everyone else easier to hurt. They organise it as four verbs: promote, protect, partner, punish. Promote is industrial policy for foundational technology. Protect is export controls, graphite for batteries, rare earth magnets, and restrictions on people, on AI engineers and rare earth technicians. Partner is locking in inputs and markets without depending on any one supplier. Punish is the new one: three State Council orders from spring this year, numbers eight three four, eight three five and eight three seven, covering supply chain security, countering foreign extraterritorial jurisdiction, and outbound investment screening. Plus an amended Foreign Trade Law in March that permits export restrictions specifically to build up a domestic industry. That last one is the tell, it's not defensive language at all.

The logic chain is short. Xi treats control of value chains as a security requirement rather than an economic preference. So the tools get built. So as institutional capacity grows, willingness to use them grows. That last link is load-bearing, and the evidence for it is thinner than the rest: the commerce ministry's export control bureau has run at least three recruitment rounds since twenty twenty-two, and twenty Japanese entities went onto control lists in June after the Japanese prime minister's Taiwan remarks.

What it rests on is overwhelmingly Chinese regulatory text. State Council orders, ministry notices, the fifteenth five-year plan. That's the right way to read a doctrine and the wrong way to measure an effect. So when they need an effect number, they borrow one: up to six hundred and fifty billion dollars of exports, about twelve percent of G7 manufacturing exports, exposed to Chinese market share gains by twenty thirty. That's from Rhodium Group and the US Chamber of Commerce in May. The Chamber is an interested party and MERICS doesn't tell you how the number was built.

Here's where I went and looked. MERICS asserts leverage constantly and never sizes it. Someone has. The Observer Research Foundation in Delhi, close to the Indian government's line, read it that way, cites a twenty twenty-five study by the Korea Institute for International Economic Policy, which is Korean government funded, estimating that a rare earth supply disruption would cut Korean secondary battery exports by about eleven percent and auto parts exports by about twenty-four percent. That's a real number and it supports MERICS. It also shows you what MERICS chose not to do.

On forecasts: almost everything here is undated. "Willingness to act is rising." The only claim with a clock on it is the borrowed one.

The doctrine reading is careful and I think it's broadly right. My challenge is the chapter on third countries. MERICS says China's export surge is blocking industrialisation in middle-income countries, and the evidence offered is that only about twenty percent of countries under Chinese customs data have a trade surplus with China. That's not evidence of blocked industrialisation. That's the accounting signature of importing capital goods. In April, a seminar at ISEAS in Singapore heard the Lowy Institute's lead economist argue the opposite reading of the same flows: this shock is mainly intermediate inputs and capital goods, ASEAN and China have expanded third-market exports at broadly similar rates, and ASEAN's production structures weren't hollowed out, the region added domestic value while leaning harder on Chinese inputs. One seminar isn't a refutation. But MERICS prints a bilateral balance and calls it obstruction, and economists in the affected region are reading it as upgrading. What settles that is value-added trade data of the kind IDE-JETRO in Chiba builds for exactly this purpose. Headline balances can't.

GMF on the EU's October deadline

Gunnar Wiegand's piece is about a calendar. The EU's goods deficit with China: three hundred and six billion euros in twenty twenty-four, three hundred and sixty billion in twenty twenty-five, heading for four hundred billion this year. Von der Leyen calls it a billion euros a day and a tipping point. Trade and investment consultations opened on the twenty-ninth of June, a ministerial lands in Beijing in the second week of October, and the European Council meets on the fifteenth and sixteenth. Deal, or unilateral measures.

The argument runs like this. China is under real domestic pressure and needs export markets, so it can be squeezed, but only if the EU turns up with instruments rather than communiqués. He names three. Anti-dumping and anti-subsidy duties, which are slow and case-by-case, and he expects them used less. Sectoral safeguards, adoptable within months and blockable only by reverse qualified majority, that's the practical one. And a new diversification instrument. Germany is the swing weight and it's moved: Volkswagen's chief executive and IG Metall both want measures now.

Look at what the pressure case actually rests on. Construction investment down seven point two percent, property down nearly twenty, urban unemployment at five point three, industrial production up five point two. Those are Chinese official statistics, the numbers published by the government he's arguing is on the back foot. And the subsidy case rests on a July twenty twenty-four speech by a US Treasury undersecretary: subsidies around five percent of GDP, Chinese firms receiving three to eight times what OECD firms do, and almost sixty percent of their global market share gains explained by subsidy. That last figure is doing enormous work in this piece, and it comes from a political speech.

The forecast is the sharpest either report makes, and it resolves in about three weeks. Deal or no deal by mid-October.

Now the problem. His own section on whether China can be persuaded answers no. Dual circulation is official policy and has been since twenty twenty. The Fourth Plenum's priorities are technological autonomy and national security with nothing on global imbalances. And any concession to Brussels sets a precedent with every other trading partner. He argues all of that well. But if it's true, the deal branch of his deal-or-trade-war choice isn't actually live, and the piece isn't describing a choice, it's describing a deadline with one exit. He never reconciles those two sections. The second gap is stranger. He costs what the EU can do to China in detail and never costs what China does back. There's no retaliation column anywhere in the instruments section, in a piece whose entire thesis is that resolve has to be priced before you show it.

What to watch

The nearest thing on the clock is that Beijing ministerial and the European Council right behind it. If concrete product-level commitments come out of it, hybrid cars is the one Wiegand flags, then the deal branch was live and his own scepticism was overdone. If it's a communiqué instead, watch whether a sectoral safeguard actually gets tabled, because that's the instrument that can move in months rather than years. And if Beijing answers with the punishment toolkit MERICS catalogued rather than with ordinary tariffs, that's the first real test of whether capacity really does become willingness.