What Everyone Gets Wrong About the Economic Problems in Europe

Dominik Leusder · London School of EconomicsNeutral

Near-frontier models like Mistral suffice for most European business needs, though its Chinese-weight foundation creates dependency risks.

Regulators and practitioners he speaks with consider near-frontier models like Mistral sufficient for the vast majority of European corporate needs. The caveat is that Mistral is launching models built on Chinese model weights because Europe lacks the compute to train its own, which creates dependency issues.

“near frontier models like Mistral are totally fine for the vast majority of business needs of most European firms... Mistral is launching models with a Chinese chassis... raises all sorts of other dependency issues, I suppose.”at 7:52

Dominik Leusder · London School of Economics

The Draghi-style narrative of Europe steadily falling behind US productivity rests on flawed data and price-adjustment artifacts.

He argues the Draghi-style claim that Europe is steadily falling behind US productivity is built on flawed data. A GDP-per-hour measure fixed at 2010 dollars extrapolates using national-accounts real growth. The US BEA adjusts tech prices more aggressively for quality, which inflates US real growth by perhaps 0.2-0.3% a year and compounds into a spurious wedge. Properly measured level snapshots show the US only slightly ahead, and that edge vanishes excluding the top 10% of earners. Europe does have real problems of its own, but they stem from poor macro policy, not an inherent productivity collapse.

“The thing that's presented as a robust stylized fact in the Draghi report... is simply no such thing. It's not established by any sort of solid economic data.”at 22:29

Dominik Leusder · London School of Economics

Europe has done poorly on its own terms for 10-15 years because of policy choices, especially macro policy.

While he disputes the narrative of relative decline versus the US, he says Europe has done incredibly poorly in its own terms over the last ten to fifteen years. He attributes this to policy choices, especially macro policy, which the US handles comparatively well. He also points to internal divergence, noting the Italian South has declined in absolute terms and the Italian economy has stagnated for twenty years, unique among G7 economies.

“Europe in its own terms has been doing incredibly poorly over the last ten, fifteen years, but that's due to policy choices, especially macro policy, which The US does comparatively very well.”at 33:09

Dominik Leusder · London School of Economics

German manufacturing is in disorderly decline, with over a million jobs lost since 2021 amid Chinese competition and the energy crisis.

Germany's industrial heartland is undergoing what looks like a disorderly decline in industrial employment. China was once a positive demand source for German high-tech goods but has flipped to a competitor that has moved up the value chain, aided by large IP transfer from Germany. Germany has shut over a million manufacturing jobs since 2021, compounded by an energy crisis and appalling European energy policy. Even the chemical industry, where employment has fallen least, is masked by labor hoarding in a dysfunctional labor market, so the true shock is understated.

“Germany... is currently experiencing, I think it's fair to say, what looks like a disorderly decline in industrial employment... Germany alone has shut over a million jobs in the manufacturing sector since 2021.”at 40:34

Dominik Leusder · London School of Economics

German chemicals are most exposed to energy costs and Chinese competition, with flat employment masking the true shock through labor hoarding.

German chemicals, a crown jewel of the industrial sector, are the most exposed to both the energy crisis and Chinese trade. Chemical export trade values are sharply down in both Germany and China, while Chinese export volumes are sharply up and German volumes down. Chinese imports from Germany are sharply down. Flat employment only reflects labor hoarding due to a dysfunctional labor market, masking the real extent of the shock.

“trade volumes for chemical have been sharply up from China and, of course, down from Germany... it's sort of masking the extent of all of this.”at 43:40

Dominik Leusder · London School of Economics

China relies on net exports for perhaps over half of GDP growth, pushing deflationary surplus capacity into open markets like Europe.

After the housing bubble burst, China intensified manufacturing subsidies and is weakening domestically, with falling demand for cars and other goods and a large gig-economy labor force. It has become reliant on net exports for GDP growth (possibly more than half), producing export-price deflation. This pushes surplus capacity into open markets like Europe and creates a political-economy obstacle to rebalancing.

“it's become quite reliant on export... Net exports for annual GDP growth... more than half of net... GDP growth might depend on net exports”at 42:05

Dominik Leusder · London School of Economics

China is resisting yuan appreciation, further contributing to global trade imbalances.

Beyond industrial policy, China is resisting any appreciation of the yuan, which further contributes to global trade imbalances. Europeans are slowly considering countering China on revaluation, but the yuan remains undervalued in effect.

“China, in addition to industrial policy, is also really resisting any appreciation of the yuan, and that is further contributing to the imbalances.”at 52:11

Dominik Leusder · London School of Economics

Europe won't build a frontier AI model soon and shouldn't try; it can benefit as a diffusion economy.

Europe will not build a frontier model soon and shouldn't try. As with past tech booms, where European consumers pocketed welfare gains through lower prices despite entrenched US moats, Europe can benefit as a diffusion economy. It should focus on tech-stack sovereignty and on near-frontier models like Mistral that suit most business needs. He flags a dependency risk, because Mistral builds on Chinese model weights given Europe's lack of compute, but remains sanguine on the AI race overall.

“Europe will not be able to create a frontier model anytime soon. And I think it's a good case that it shouldn't try to do so... So far, I'm pretty sanguine on Europe when it comes to the AI race.”at 7:13

Dominik Leusder · London School of Economics

US healthcare spends roughly 50% above predicted levels due to high unit prices from rent extraction, distorting GDP comparisons with Europe.

The US spends about 50% more per person on healthcare than its income level predicts, roughly $1.7-1.9 trillion a year in excess, and gets worse outcomes. The cause is high unit prices driven by rent extraction from hospitals, insurers and intermediaries, not high wages from tech productivity. This inflates US GDP and consumption figures, distorting comparisons with Europe, and it benefits the top 10-20% while imposing large costs on everyone else.

“The US spends around 50% more on health care per person than you would expect relative to what advanced countries actually spend... the difference is... prices, really.”at 26:15

Dominik Leusder · London School of Economics

Near-frontier European AI models can serve most corporate needs; Europe's real competitive risk is manufacturing versus China, not AI.

Critics claim Europe is renting the future and will miss the capex-generating generative AI technology. He argues near-frontier models are not bad and can serve most European corporate needs, and doesn't see a decisive competitive race in AI for Europe. The real competitive risk lies in legacy and high-tech manufacturing vs. China.

“I don't think that the sort of near frontier models are really bad and that they can service most of the needs of European corporates. But I don't think that there is a competitive race specifically in the area of AI.”at 40:04

Dominik Leusder · London School of Economics

Industrial decline from the external trade shock is boosting Germany's far right and hurting incumbents in regional elections.

Industrial decline from the external trade shock is producing real electoral consequences in Germany. The far right is doing very well in regional elections and incumbents poll poorly, as the far right frames industrial decline as an incumbent and green/energy policy failure, though evidence suggests the decline is mostly driven by the external trade shock.

“the far right is doing very, very well in these elections and the incumbents are polling very poorly because the far right has managed to frame the issue of industrial decline... around a failure of the incumbents and the failure of, you know, green policies and energy policy”at 47:50

Dominik Leusder · London School of Economics

The EU's Industrial Accelerator Act will likely be too little, too late; worst case is a full EU-China trade war.

The EU's Industrial Accelerator Act, with 'by Europe' provisions to protect the auto market, is slow, divided between France, Germany and exporters, and trapped by 27-country veto bargaining that regresses to the lowest common denominator. Outcomes will be too little, too late. China retaliates in an institutionalized, rapid way (e.g., blacklisting European industrial firms the day after EU sanctions). The likely scenario is a weak EU measure met with a minor Chinese concession that doesn't change the industrial shift; the worst case is a full trade war.

“I think a likely scenario is that the EU does something eventually next year that because of these problems in finding decisions, will be too little... The worst scenario is a full blown trade war.”at 53:41