Measures to safeguard semiconductor supplies proposed in the European Chips Act could prove to be wrongly focused, and could tip over into harmful protectionism.
The European Chips Act, proposed by the European Commission in February 2022, is intended as a framework giving the Commission more power to steer and shape the European Union’s role in global value chains for semiconductors. The impetus is the shortages of chips experienced by European industries during the COVID-19 crisis. Chips are ubiquitous in modern manufacturing. Cars, for example, contain hundreds of semiconductors. Through the Chips Act, some of which must still be approved by the European Parliament and EU countries, the Commission wants to make the EU a stronger global player in semiconductor production and minimise the risks of future supply chain disruption.
The European Chips Act could therefore have important international implications. The industry is too capital-intensive, specialised and interconnected for any one country or bloc to aim for self-sufficiency. The Chips Act is important to strengthen the EU foothold in high-tech industries, but it should be integrated into the broader European trade strategy. The EU priority should be finding ways to coordinate with partners in the value chain, while maintaining its position in the context of the confrontation between the United States and China over technology supremacy.
The proposed Chips Act has three pillars: research, development and innovation (R&D&I) policies; a new state aid exemption for cutting-edge foundries (semiconductor manufacturing plants); and measures to monitor the supply chain and intervene in crises.
The first pillar on R&D&I is the most conventional in its nature and builds on existing programmes to strengthen the EU chips ecosystem over the long-term. Europe is already prominent globally in fundamental research; the Chips Act aims to reinforce this by supporting industrial innovation. It would bring together existing (and successful) R&D programmes, including chips-related Horizon Europe projects, Digital Europe and the Key Digital Technologies Joint Undertaking under the umbrella of a new Chips for Europe Initiative.
As part of this initiative, the European Commission wants to create new ‘open’ R&D&I infrastructures. They would encourage cooperation between firms and benefit smaller companies, which are less likely to benefit directly from state aid.
It remains unclear how much new funding is available to these new initiatives. The Commission has said it wants overall to “mobilise more than €43 billion of public and private investments”. The Chips for Europe Initiative itself would have an €11 billion budget, but only €3.3 billion of this would come from the EU itself, by redirecting funds already committed through Horizon Europe and the Digital Europe Programme, some of which were earmarked for chips anyway. A ‘Chips Fund’ will leverage EU and European Investment Ban funding to raise €2 billion in equity financing for start-ups in the sector. For the remainder of the €11 billion, the Commission provides only the legal framework for countries and firms to invest.
The biggest part of the public investments will likely come from another EU instrument, the Important Project of Common European Interest (IPCEI) on microelectronics, through which EU countries can support industrial R&D projects. IPCEIs have certain shortcomings in transparency and governance but are nevertheless a fitting – and the only – EU tool to channel public and private funds towards industrial goals. Venture capital markets in the EU are less developed than in the US to support innovation in the high-tech sector. State subsidies support some projects but developing equity financing in the EU is a more long-term solution.
A new approach to supply chain management
With the second pillar of the Chips Act, the Commission wants to increase capacity in the most concentrated and capital-intensive stage of chip production: fabrication. To achieve a goal – set out already in 2021 – of doubling European fabrication capacity, the EU needs to attract foreign investment, especially for the latest generation of chips for which there are no European producers. The Chips Act would allow EU countries to grant subsidies for manufacturers willing to build cutting-edge ‘mega-fabs’ in the EU. Given the global subsidy race around foundries, this would be an expensive endeavour.
Because a European ‘mega-fab’ would apply existing cutting-edge technology, it would not meet the conditions for a standard industrial subsidy project (IPCEI) which can only finance R&D&I. Therefore, the Chips Act would establish a bespoke state-aid exemption. The European ‘first-of-a-kind’ rule that would be established by the Chips Act would allow subsidies for foundries implementing cutting-edge technology not yet present in the EU but present elsewhere. The Commission wants EU countries to address these applications for aid as rapidly as possible with an urgency that supposedly precludes any impact assessments.
Building capacity at home is only part of the EU’s answer to ensuring more resilient supply chains. The third pillar of the Chips Act foresees monitoring of the sector and would establish tools to intervene in times of crisis. This could include ‘joint procurement’ by the Commission on behalf of EU countries and industries, requiring foundries that benefitted from state support to supply European customers first. Export controls could also be envisaged. This toolbox is largely taken from the pandemic playbook, where the Commission procured medical goods and vaccines for EU countries and introduced export authorisation requirements. That production should serve EU customers first also justifies EU public support. However, the justification for these interventions is much weaker for chips – in the interest of private companies – than for medical products in the interest of public health.
The strategic question
All this raises the broader question about strategy. The Chips Act is intended not only as an answer to recent shortages, but to the broader challenge of competitiveness and interdependence in high-end technologies. It marks a shift towards a more hands-on approach to industrial and trade policy.
The Chips Act would see the EU joining a global subsidy race. With the goal of €43 billion in public and private investment, the EU wants to match amounts spent by China ($150 billion over 10 years) and the US ($52 billion over 5 years). Estimates of support provided to the industry by the US, China, Japan, South Korea and the EU amount to $721 billion, or 0.9% of 2020 global GDP. This is bound to create major competition distortions. The EU has prioritised limiting the effect of foreign subsidies. The fact that it is now entering a global subsidy race in high-tech is a sign of the failure of multilateral subsidy control – and since most of the competitors are likeminded partners, also of policy coordination.
Furthermore, investing public money in fabrication capacity for high-end chips is risky. Investments in innovative foundries do not always succeed: the technologies are tricky to master. To be profitable, factories need high utilisation rates. Only three companies in the world currently produce cutting-edge logic chips (TSMC, Samsung and Intel), with Intel and Samsung yet to master the latest generation, notwithstanding billions invested. This high concentration at the top gives these three firms huge negotiating power when countries compete to attract them. They are already well financed, with tens of billions of dollars in annual capital expenditure. Since the Chips Act was proposed, Intel has announced plans for up to €80 billion of investment in Europe, but there is no disclosure of support from EU countries the firm might be negotiating.
State aid to this sector should also be balanced carefully with European demand for such chips. The EU currently consumes very few cutting-edge chips as inputs to production. Cutting-edge chips are most important for product such as smartphones and computers, but EU industries (including automotive) rely more on trailing-edge chips, of which each car needs hundreds. Shortages affecting the auto industry would be better addressed by increasing capacity in lower-end rather than cutting-edge chip manufacturing. The COVID-19 crisis also revealed vulnerabilities in global supply chains through concentration and bottlenecks. Diversification of supply of high-end chips outside of Taiwan is important, but diversification of supply of lower-end chips away from China is just as important.
Though global demand for chips will undoubtedly increase in the coming years, the current shortages have prompted governments and firms to invest in fabrication, making overcapacity in the future not unlikely. For instance, prices for some types of memory chip, which best fit the definition of a commodity, are expected to decline by 16% in 2022 as supply growth outstrips demand. The chips industry is subject to boom-and-bursts cycles. The Commission has not clarified which market failure the Chips Act is supposed to address with massive government support, or how obtaining an (at best) modest market share in cutting-edge logic chips would actually increase EU geostrategic leverage.
On supply management, implementing emergency measures would not fit the reality of supply and demand dynamics in the high-end chips sector. The chips that would be produced by an EU ‘first-of-a-kind’ facility are not commodities that can be reallocated. High-end logic chips are produced by foundries according to customer specifications, which vary between industries and buyers, and the manufacturing of such chips can take up to 26 weeks. A request to divert production would be inconsistent with an industry that cannot swiftly adapt production lines – as confirmed by the 2021 shortages.
Governments currently make up 1% of global demand for chips. If the Commission were to act as a buyer, it would have to choose to which companies to allocate the scarce supplies, while making the situation worse for everyone else. This problem can be much better solved by firms reviewing the trade-off between lean inventories and supply sustainability. The private sector should adapt to the risks on global value chains witnessed during the pandemic – it is starting to do so.
Such measures raise risks of adverse economic implications and political messaging. Protectionist measures in a sector in which the EU is currently dependent on imports could create a precedent which would be to the detriment of the EU if copied by other players. Chip shortages were not the result of export bans but of a multitude of economic factors. Where export control measures have been introduced in the chips sector, it has been because of geopolitical rivalries and has not targeted the EU. While the rationale for EU export controls – economic activity or national security – remains unclear, their signalling is clearly protectionist. The EU has more to lose than others from ‘beggar-thy-neighbour’ policies in the industry.
The need for a more muscular trade policy would be better achieved through general trade defence tools than chip-specific export controls. The EU should push back against uncoordinated protectionist reactions triggered by heightened supply chain risks and uncertainty, since the COVID-19 crisis and Russia’s aggression against Ukraine. This moment should be turned into an opportunity for sustainable trade: the balancing act is between creating global coordination structures among likeminded partners for strategically important products, while retaining policy independence from players such as the US.
Poitiers, N. and P. Weil (2022) ‘Is the EU Chips Act the right approach?’, Bruegel Blog, 2 June