All transformational technological progress is the result of research conducted in not just one country, but globally. Recently, China has been rapidly advancing to a technology leader. While Europe should continue its research cooperation with China the case of semiconductors illustrates the existence of three pitfalls: technological dependencies, national security and values.
From the days of its foundation, the European Union (EU) was a project of transnational cooperation. European integration is rooted in the belief that collaboration increases mutual understanding and interdependence raising trust and lowering the risk of confrontation and war. It is no wonder that this includes the field of research collaboration. The EU’s “Erasmus” and “Horizon” programs are only the most famous instruments facilitating research cooperation. There is no doubt: The EU’s openness and support for research collaboration have preserved Europe’s strength in high-tech innovation and have furthered technological progress.
Over the course of the last five years, emerging and foundational technologies (EFT) have been politicized. In particular, the United States (US) and the People’s Republic of China (PRC) compete over power in the field of technology. The EU sees the need to adapt to these developments. This “geopolitical turn” of EFT is also including research cooperation. This is why the European Commission is discussing (further information also here and here) an EU-China Joint Roadmap on science, technology and innovation cooperation.
In this piece, we argue that unfettered EU-China technology research cooperation is indeed problematic. Considering the case of semiconductors, we suggest that Europe should consider at least three types of pitfalls arising from (a) national security concerns, (b) technological dependencies, and (c) differences in values. This is not to say that the EU should end research cooperation with China. We acknowledge the benefit and the necessity of such cooperation. Instead, the three pitfalls should be considered references to think about limitations of high-tech EU-China research collaboration.
The benefit of semiconductor research collaboration and the role of China – a brief stock taking
Semiconductor companies have one of the highest R&D costs of any industry. Over the past 20 years the semiconductor industry spent on average above 15% of revenues on R&D. Many fields of semiconductor research are also very costly as researchers need access to increasingly expensive fabs. Thus, a lot of semiconductor research by universities is done jointly with private companies or research organizations. Historically, the US contributes to most of semiconductor research. When looking at the three leading semiconductor conferences over the past 25 years (IEDM, ISSCC, VLSI), US institutions contributed to more than 40% of papers every year. While institutions in China contributed to less than 1% of conference papers in 2010, by 2020 this share was already at 10%. Today, China is the EU’s second most important research partner based on this conference paper analysis – right after the United States. In general, semiconductor research has become more international and cooperative over the past 25 years, not least because it takes 18 times more researchers today than in 1980 to keep pace with Moore’s Law.
Despite all these benefits of research collaboration in semiconductors and China’s growing role, we argue that the EU should be carefully reviewing its exposure to the PRC.
Pitfall 1: National security concerns
As a foundational technology, semiconductor research is often applicable not only to civilian but also military applications. In contrast to the US, China is not a security ally of the EU and its member states. This puts Europe in the situation that research collaboration with China in the field of semiconductors strengthens the military capabilities of a “systemic rival” and potential adversary. Complicating matters even further, the PRC is explicitly formulating its aspiration of what is referred to as a “civil-military fusion”. This concept is to facilitate civil research and commercial applications for military purposes.
European researchers need to not only consider the potential military applications arising from joint research but also the institutions they work with. Several public universities with a strong footprint in technology research are not just public funded but an integral part of the security apparatus. The Nanjing University of Science and Technology, for example, is part of the Ministry of Industry and Information Technology’s State Administration for Science, Technology and Industry for National Defense. The Jiangnan Institute of Computing Technology’s integration is even more direct; it works under the People’s Liberal Army Strategic Support Force. These are just two of several examples of how seemingly “normal” public research institutions are an integral part of China’s security apparatus.
Pitfall 2: Technological dependencies
Europe’s stronghold in semiconductors research largely stems from contract research done by research technology organizations (RTOs), primarily based in Belgium, France, Germany and the Netherlands. The commercialization of RTO’s innovation is happening outside of the EU in East Asia and North America – wherever the RTO’s client is headquartered. Since China’s share of semiconductor R&D is constantly growing, Chinese companies become an increasingly important customer of EU’s RTOs who increasingly depend on resource-rich non-European industry to finance their research.
To the extent that it is in the EU’s interest to preserve cutting-edge research capabilities it needs to be concerned that its preservation increasingly relies on the willingness of Chinese actors. This is politically concerning because of the party-state’s control over companies and research institutions working on foundational technologies.
Pitfall 3: Value differences
As a foundational technology, semiconductors are at the center of advancements in emerging technologies such as artificial intelligence (AI), quantum computing or autonomous vehicles. If the EU wants to strengthen “human-centric” AI, we also need to invest in AI chip research since software-hardware co-design is crucial especially in the field of machine learning (ML). Chinese companies are quickly advancing in the field of AI accelerators and ML making them a viable research partner in these areas. At the same time, China’s use of AI for social control is at odds with European values and human rights.
What shall the EU do?
Only as recently as 2014 did the Harvard Business Review publish a paper entitled “Why China cannot innovate”. Fortunately, the days of such western hubris are gone. The EU is on high alert but still in the process of identifying its weaknesses. While we do not offer a comprehensive set of tools to cope with the risks stemming from national security concerns, technological dependencies and value differences we close this piece with two broad suggestions.
First, the EU needs to scrutinize its RTO business model. Strength in research and innovation is without a doubt useful to the EU. However, it is questionable whether it is in the EU’s best interest to innovate without commercialization. This implies that addressing the above described challenges requires action that goes far beyond the narrow field of research cooperation.
Second, we do not argue Europe should give up on research collaboration with the PRC. However, the EU should strengthen transparency requirements and put in place an approval regime that considers risks from both technological and political considerations.