10 Feb 2022

The European Chips initiative

Industrial policy at its absolute worst



Nobody needs convincing that semiconductors and electronic chips are essential for modern industry and for the green and digital transitions. But this does not necessarily mean that the EU should actively support this sector. Many sectors are important but the cold hard truth is that the state cannot support everyone.Industrial policy means setting priorities. Moreover, public intervention is justified only if the market doesn’t function as it should, such as in the case of nascent or ‘infant’ industries.  However, the European chip industry is definitely not an infant. It has received financial support in the past but then lost its way, no longer a source of cutting-edge innovation within the sector.

Investing public funds can be justified if a certain sector cannot finance itself through market means (this was the case for aircraft manufacturers in the 1970s, hence the public start-up aid for Airbus).  This is not the case with the chip industry, as its leading firms have no problem sourcing the huge amounts of capital required to build new foundries or to support the intensive research going into chip design and the development of cutting-edge machinery that pack evermore smaller transistors onto a single chip.

Taking this background into account, it is difficult to find any reasonable justification for the European ‘Chips Act’, recently unveiled by the European Commission.

Financing and management: A mirage and a mess waiting to happen

EUR 3.3 billion (about 0.5 billion per year over seven years) is earmarked for the initiative, obtained mainly by redirecting funds from other parts of the EU budget.  The only new funding is EUR 250 million, or about 8 % of the total. 92 % of the EUR 3.3 billion would thus come from cutting funding to other research or digital initiatives.  The Commission does not even attempt to make the case that funding for the Chips Act will yield a higher return than the general research cut elsewhere (which includes cuts to advanced computing and AI).

With only EUR 250 million in ‘fresh’ money out of EUR 3.3 billion, it is highly misleading to suggest that the Chips Act will have a major impact when a single foundry can cost billions.  As usual, the proposal hedges its promises with the, by now, standard formula that the Chips Act ‘will mobilise more than EUR 43 billion of public and private investments’. The key words here are ‘mobilise’ and ‘private and public’, both vague enough for the Commission to claim credit for any investment (mostly private in this sector) that is even remotely related to the initiative.

Additionally, part of the proposal’s budget will go towards setting up entirely new bodies, including a European Semiconductor Board composed of Member State representatives, opening the door for blatant horse trading over the distribution of funds. This is in clear contrast to Horizon Europe, where Member States have no say on which project proposals win funding – in short, only excellence matters. And unfortunately, it is Horizon Europe that will face cuts to pay for the Chips Act.

The European Semiconductor Board is thus an invitation for Member States to advance their own national interests. The Commission of course encourages Member States to coordinate their chips research but this is wishful thinking – national funding always follows national priorities.

Security of supply? 

A key aim of the Chips Act is to contribute to security of supply via so-called ‘Integrated Production Facilities’ and ‘Open EU Foundries’.  These facilities could be ordered to follow ‘priority rated orders’ when a shortage is declared.  Moreover, the Commission could even impose export controls (which would be hard to reconcile with WTO rules).

The main qualification of these ‘European’ facilities is that they are a ‘first-of-a-kind facility’. However, chips come in many varieties. The number of transistors per chip is not the only variable that determines their value or strategic importance. Consequently, it is not clear which kind of foundry should be supported: those with the smallest transistors, those with lower energy consumption or those specialised in certain applications?

There is also no clear definition of what constitutes a ‘shortage’, only that it could be declared if there are ‘serious disruptions in the supply of semiconductors leading to significant shortages, which: (a) entail significant delays or significant negative effects on one or more important economic sectors in the Union, or (b) prevent the supply, repair and maintenance of essential products used by critical sectors.’  This is vague enough to cover any price increase which inconveniences politically powerful sectors, and the Commission can declare a shortage without giving a justification via an implementing act.

Market forces should be allowed free rein in the chips industry. Critical sectors should pay higher prices for chips as they can also increase their own prices (if their products are really that critical). The most efficient producers will always hesitate to become ‘Open EU Foundries’ if that means they must fulfil politically motivated orders when prices are high, i.e. when they would be otherwise generating their highest profits.

There’s a better alternative

All in all, the European Chips Act combines the worst features of industrial policy – insufficient resources, mostly obtained by cutting valuable research elsewhere, and a vague security of supply mechanism, all managed by new permanent bodies where Member States will fight each other over the distribution of funds.

The main argument for the Chips Act seems to be that other major economies provide huge state support to the industry.  The US has the CHIPS for America Act, with proposals of financial support of USD 52 billion.  China has the official goal to domestically produce 70 % of chips needed for its own consumption (which is unlikely to be met).

There are two reasons though why the EU does not need to follow in their footsteps. First, it’s ridiculous to copy the mistakes of others. Second, the rate of return in this sector for Europe is likely to be very low, precisely because others are supporting the same sector so lavishly.

The smart approach for Europe would be to not participate in this subsidy race. Within the wider chips sector, Europe does have important expertise in specific manufacturing equipment. Financial resources should be concentrated in this area, perhaps also in the innovative design of new chips. This would be a far better use of scarce EU funds, definitely more desirable than setting up cumbersome bureaucratic machinery with the sole aim of ensuring a certain percentage of global production is relocated to Europe.