Governments on both sides of the Atlantic have resolved that they will not let their economic relations with China become — or remain — a geopolitical dependency that Beijing could use for its own strategic purposes. While this fear had been growing for years, helped along nicely by an increasingly unabashed Chinese openness that economic dependencies on it were precisely what it was seeking, Russian president Vladimir Putin’s weaponisation of energy demonstrated how real the risk was. That political shock is a main driver of global politics today, as the latest G7 summit shows. That is despite Europe’s surprisingly creditable performance in weaning itself off Russian gas at much less economic cost than many had feared — including in Germany, as I highlighted in my column this week.Western companies, however, are starting to squeal. A striking example is Nvidia, the US chipmaker that is a global leader in AI-suitable semiconductors. In an interview with the FT, chief executive Jensen Huang warned that US policies to limit US chip sales to China threatened “enormous damage to US companies” because it removed the incentive to invest in capacity:“If the American tech industry requires one-third less capacity [due to the loss of the Chinese market], no one is going to need American fabs [semiconductor factories], we will be swimming in fabs . . . If they’re not thoughtful on regulations, they will hurt the tech industry.”The importance of the Chinese market is front and centre of European business leaders’ minds too. In another FT interview, the chief executive of industrial conglomerate Siemens said that it was “not an option” to lose Chinese market share, which accounts for 13 per cent of the group’s total revenues. Even as his company is scouring south-east Asia to find alternative manufacturing locations, Roland Busch argues that sales to China drive its overall innovation and growth.No western companies better illustrate this dependence on China than Volkswagen Group, which makes half of its profits in the country. It is investing massive amounts in Chinese manufacturing operations in collaboration with Chinese partners, potentially including software company Huawei. The strategy is to produce cars “in China, for China”, which in the case of electric vehicles could well become producing cars in China for import into Europe. Many a business leader is watching both the size of China’s market and the growth of its own industrial companies and worry if they can afford not to try to embrace both as much as they can.It is understandable that executives of such companies look with trepidation at the thought of a “decoupling” between China and western economies. (Do read my colleague Peter Campbell’s excellent analysis of their conundrum.) But the more acute risk may be something close to the opposite of what they think, namely, that the Chinese economy is not all that it’s cracked up to be.This week, statistics showed that German exports to China in the first four months of this year had dropped 11.3 per cent on the previous year. Since the EU as a whole saw a small rise, that meant the German share of EU exports to China shrank. The reason could presumably be that Germany trades intensively in industrial goods where Chinese competitors are doing better (cars) or that were particularly hit by high energy prices (chemicals).But the bigger point applies to European and American companies in general: what if the Chinese market is not going to be as big as it was? It has become clear that the post-pandemic rebound in demand-driven growth that everyone expected when China lifted Covid-19 restrictions is not materialising. In addition, the long-term headwinds to China’s growth are making themselves felt: the population has peaked, the bill for bad investments in real estate is coming due and the country faces the familiar “middle income trap” — the difficulty of shifting from poor-to-middle-income economic growth based on adding more capital to middle-to-high-income economic growth based on greater productivity. A good summary of the pessimistic outlook is a note from Capital Economics, which points out that the demographic outlook has got a lot worse in just a few years. It estimates that China’s economy will never outgrow the US’s when measured at market exchange rate: the convergence will top out at 90 per cent of the US economy’s size in 2035. If so, much of the expected market growth at the root of the western companies’ angst about “missing out” on China may be illusory. Even some of the growth to date may have been based on illusory wealth.We don’t know the answer to this question. But surely it behoves western companies’ governments to think not just about the risks of untying themselves from a China that expands strongly, but also those of tying themselves to one that does not. What are those risks? Think of three different things you could mean by “depending” on the Chinese market. First, a western company could be making money from exporting things to China. Second, it could be making money from producing things in China for sale to that market (or indeed for export to other countries). Both of these create profits for the company, but only the former also employs people, pays salaries, invests physical capital and generates learning-by-doing in Europe or the US. The third dependence could be on technology. So far, this dependence has gone the other way: China has allowed western companies in on the condition that they share their technology with local joint venture partners. But in some areas where China is at the technological frontier — battery and EV technology may be one — it could be increasingly relevant, especially if western companies decide to put their research and development resources there. And in an indirect sense, a western company’s perceived need to succeed in China could motivate it to innovate more at home.In most of the public debate in Europe, the discussion seems to take for granted that we are talking about the first kind of relationship. It’s certainly the one that brings the most tangible benefits to the “home” economies. But conversely, it’s the one that makes you most dependent geopolitically, because the fallout from losing an export market would be felt so widely at home. In contrast, the second mode of “dependence” is really only a matter of money — which is important, for sure, but benefits a much smaller constituency at home (and big European and US companies have a global shareholder base anyway). But that also means that if politicians see straight, this is a lesser source of geopolitical dependency insofar as it should be politically less traumatic to take a hit to western companies’ operations “in China, for China”. As for the third “dependency”, so long as technology still flows mostly from the west to China, it is really just a way of deepening the other two kinds of dependencies.Politicians and their technocratic servants — and, above all, the populations they serve — would benefit from cutting through this fog and being clear about which companies “depend on China” and in which way. Losing a profit source is one thing; losing a lot of voters’ jobs and earnings is another. If a company’s “in China, for China” profits in effect cross-subsidise operations at home, that would be good to know. So here is a suggestion: you could enforce a clearer separation between the legal entities of big western companies’ China-based and home operations with strict transparency and publicity requirements for the financial flows and technological transfers/licensing between them. A bit like with big financial companies’ “living wills”, the idea would be to require companies to make plans for a restructuring that would split off their China operations entirely — that is to say, end common ownership, financial cross-subsidies and technology transfers. Knowing what this would do to the home operations’ finances and sales would be a great help in judging how vulnerable they are to a weaponisation of economic relations in a geopolitical crisis. And knowing this, in turn, would raise the pressure to reduce that vulnerability before it was exploited. Free Lunch readers on AIMany of you were unconvinced by my decidedly unpanicked assessment of artificial intelligence in last week’s Free Lunch. The main objection was that I underplay what is really new about AI — namely, that the technology becomes a decision-making subject in its own right. One reader took me to task for using the term “AI algorithm” since it doesn’t make sense to compare an impenetrable neural network with a decision rule that can transparently be broken into sequential steps. Point taken — I should really have said just “decisions taken by AI” (although I think it could still make sense to talk about AI algorithms in the sense of algorithms that require user input, and where that input is provided by AI instead of a human decision). But it doesn’t change my broader point about good policy, which I argued should ensure that any AI decision-making should be assigned to humans who can be held legally (and morally?) accountable for it. Rather than tell AI developers and company executives how they should design and use it, the prospect of jail time for harmful effects is more likely to instil caution. Other readablesThe German economy adapted much better to the loss of Russian gas imports than doomsayers had predicted, I point out in this week’s column. We should draw the broader lesson that the European economy is more adaptable than policymakers, and corporate lobbies, give it credit for.If we want to get serious about sanctions evasion, we need to crack down on secrecy jurisdictions, says the FT Editorial column.A very modest idea is making the rounds in the EU for letting Ukraine benefit from immobilised Russian assets: pass on profits from the securities settlement house Euroclear’s reinvestment of cash left idle by coupon payments and redemptions Russian owners have had to leave unclaimed because of sanctions.Numbers newsThe UK’s latest inflation numbers are terrible through and through.
You are seeing a snapshot of an interactive graphic. This is most likely due to being offline or JavaScript being disabled in your browser. More
