Billions of dollars in taxpayer funds will be offered up on the altar of European economic sovereignty. The EU has swung rapidly from naive globalism to the indiscriminate pursuit of economic self-reliance due to the shock of the epidemic and Putin’s war.
Even when it flagrantly breaches the Ricardian competitive advantage principle, the EU insists on producing its own semiconductor chip supply, lithium gigafactories, and clean hydrogen domestically.
Intel has secured €10bn of subsidies from the German authorities to build two semiconductor ‘fabs’ in Saxony-Anhalt able to manufacture advanced chips down to the frontier of two nanometers (2nm) – largely duplicating a plant that already exists on EU territory in Ireland.
This may be glamorous, but Germany does not need such cutting-edge chips to build cars or to supply its manufacturing base. It needs routine chips that are cheap and globally abundant, albeit subject to boom-bust cycles and occasional scarcity much like commodities.
The Intel subsidy amounts to over €3.3m for each of the 3,000 permanent jobs expected at the site. It is the biggest foreign investment project in German history and perhaps the most foolhardy. Negotiations are running in parallel with Taiwan’s TSMC for another site near Dresden, undoubtedly involving public funds of equal scale.
German economists from across the spectrum are united in a blizzard of criticism. “It is sheer madness. They will struggle to generate any value added,” said Justus Haucap, a competition expert from Dusseldorf University.
“Why are we handing out presents to profitable companies?” asked Reint Gropp, head of the Halle Institute for Economic Research (IWH).
The plant may prove to be the last hurrah of the silicon chip era. It relies on the extreme miniaturisation of silicon wafers, a technology already nearing its limits that may soon be leap-frogged by the next generation of compound semiconductors. It is akin to the strategy of Germany’s car industry in relying for too long on refinements to the combustion engine when it was about to be blind-sided by electrification.
The UK company Paragraf, a Cambridge University spin-off, is already producing 2D graphene chips for sensors that are one-atom thick, a thousand times faster than silicon wafers, and use 10,000 times less energy. It is a technology so sensitive that staff require special protection in electronic intelligence to prevent Chinese cyber-theft.
The British government is spending far less in headline sums on its £1bn semiconductor strategy but the money is surgical, targeting ultra hi-tech clusters in areas where the UK is already a world leader.
It is a bet on the future of advanced compound chips, but also on the crucial niche of design. Some 95pc of the processors in the world’s smartphones are designed by ARM in Cambridge.
The UK strategy is nothing like the EU’s €43bn chips act, which aims to raise Europe’s share of global chip output from 8pc to 20pc by the end of the decade (they love targets in Brussels), and which actually requires €500bn in total investment to be plausible.
It is the semiconductor equivalent of the EU’s Common Agricultural Policy, and will probably result in the same pathological misallocation of resources.
Olaf Scholz’s coalition in Berlin hopes that Intel’s €30bn project – with a smaller sister site nearby in Poland – will lead to a Silicon Saxony able to match the great global chip hubs, with an economic ecosystem of supporting universities and tech contractors.
Prof Gropp said it is really just a big factory: “It won’t be the centre of decision-making and there won’t be any important research activity.”
TSMC’s founder Morris Chang describes America’s push for chip sovereignty as a “very expensive exercise in futility”.
One wonders what he thinks of Europe’s bid to catch up from further behind and without cheap energy or a reservoir of specialised workers. Labour costs in Taiwan’s chip industry are 40pc lower and the country has a nexus of technical colleges feeding into the fabs.
Critics say the EU has drawn entirely the wrong conclusion from the supply-chain rupture during the pandemic, when carmakers cancelled their chip orders, only to discover later that makers of laptops and home electronics had snapped up all the available supply to meet lockdown demand.
There was no disruption in global chip supply apart from a brief problem in Malaysia.
It is true that half the world’s advanced chips of 10nm or below come from Taiwan, which could be stormed by the Chinese navy at any time, and these chips are becoming necessary for AI accelerators, supercomputing, G5 mobile, or smart weapons systems.
This supply chain is clearly unsafe but the problem is being rectified rapidly. The question is how much money the EU wants to spend replicating plants in the US, Korea, Japan, or indeed Ireland, and thus contributing to the next global chip glut.
The EU is adopting the same blunderbuss strategy in pursuit of battery sovereignty. Is it really worth trying to claw back market share for lithium-ion EV batteries with lavish subsidies for gigafactories when Chinese companies such as CATL are already well ahead and are switching to zero-carbon plants faster than Europe can talk about it?
The wiser course would be to focus all efforts on the next generation of solid state and other advanced batteries with five times the energy density and range. It is these that will conquer the world market in the 2030s.
It is the same folly with the EU’s €500bn hydrogen strategy. It makes no sense to ship green ammonia to Europe from halfway around the world in oceanic tankers, losing 80pc of the energy in the process.
Green hydrogen should be produced at scale in spots with the world’s cheapest renewable energy, such as Namibia, Morocco, and the Gulf, or Chile’s Atacama Desert and Australia’s Pilbara, and it should be turned into fertilisers or green steel on site before being shipped.
Europe cannot compete in green hydrogen production at a global level and should not try to do so. What it can plausibly do (by the 2030s) is to harness free surplus power from its wind and solar parks at peak times to produce hydrogen for local needs, either for energy storage to balance the grid, or to displace dirty hydrogen in nearby chemical plants. That makes sense.
Industrial strategy has its value. It is imperative at times. But there is a mad frenetic element to Europe’s quest for fortress autonomy across the board. It smacks of over-reaction. EU superstate enthusiasts are exploiting the emotion of the moment to push an ideological agenda.
The UK is well-advised to avoid this numbers game. The best industrial strategy is good education in STEM subjects (science, technology, engineering, maths), nurtured by well-funded research bodies, under the rule of tort law, and all sitting on a bed of low effective taxation for business and investment. Let the market do the rest.