Search interesting materials

Showing posts with label author. Show all posts
Showing posts with label author. Show all posts

Saturday, November 21, 2026

Who owns frontier technologies?

by Mohd Shadab Danish and Megha Patnaik.

Governments around the world are subsidising frontier technologies such as chips, computer hardware and software, and the systems used to train and run artificial intelligence (AI). In June 2026 the European Commission proposed a Chips Act 2.0, as part of what it calls a "Technological Sovereignty Package". Other large economies are doing the same. The United States set aside US\$52.7 billion for chip manufacturing and research under the CHIPS and Science Act of 2022. India's semiconductor mission had approved 12 chip projects by June 2026, with planned investment of about Rs.1.64 lakh crore.

Governments promote these programmes to improve technological sovereignty, a term that now frames much of industrial and innovation policy (Edler et al, 2023). Technological sovereignty is the idea that a country can rely on the technologies its economy and its security depend on, without depending on one foreign supplier or government that controls them and could cut them off. Two recent events highlighted the importance of technological sovereignty. In 2021 a shortage of chips stopped car production around the world. An estimated 70% of the microprocessors used in cars' electronic control units came from a single chipmaker, TSMC in Taiwan. Since October 2022 the United States has restricted exports of advanced chips and chip-making equipment to China, cutting Chinese firms off from technology they had relied on.

Measuring technological sovereignty

Technological sovereignty is hard to measure because it has several parts. A country needs the capability to develop critical technologies. Capability is usually tracked with R&D spending, patent counts and research output. The Australian Strategic Policy Institute's Critical Technology Tracker, for example, ranks countries by high-impact research in 44 critical technologies. A country also needs reliable access to the inputs these technologies are made from, such as chips, components and critical minerals, without depending on one or a few external suppliers. The European Commission measures this dependence from trade data, and in 2021 identified 137 products in sensitive sectors for which the European Union depends heavily on foreign suppliers. A country also needs control over the technologies used at home, which depends on who holds the rights to them. Control has had the least attention. The OECD's statistics on the foreign ownership of inventions come closest, but they follow where an invention is made rather than where its rights apply (Guellec and van Pottelsberghe de la Potterie, 2001). A country can host many innovative facilities built by foreign firms, and its firms and inventors can file many patents at the national patent office, but the rights to the most advanced technologies used in the country could still be largely held abroad.

Measuring who owns frontier technologies

In a new paper, Measuring technological sovereignty: Who owns frontier technologies? (Danish and Patnaik, 2026), we propose a measure of technological sovereignty based on ownership. For country $c$, with $R_c$ frontier patents granted to residents and $N_c$ granted to non-residents at the country's own office,

\[ \text{Ownership}_c = \frac{R_c}{R_c + N_c} \]

Ownership is the share of frontier patents granted by a country's patent office that go to residents of that country. The share tells us how much of the frontier technology protected in a country is controlled by its own residents.

The data used are the counts of grants to resident and non-resident applicants by office, year and field of technology, published by the World Intellectual Property Organization (WIPO). We take Computer technology as the frontier. It covers computing, data processing, memory and machine learning. It has been the largest of the 35 technology fields worldwide in every year since 2004. We compute the measure for 53 economies over 1996-2022. When we rank countries we keep the offices that grant at least 100 frontier patents a year. The members of the European Patent Convention patent mostly through the European Patent Office, so we treat them as one bloc.

The measure counts the local subsidiary of a foreign firm as a resident, so it could overstate domestic ownership. We checked the measure against two other sources. Among the 13 largest patent offices, countries rank similarly on our measure and on OECD data on the foreign ownership of domestic inventions, with rank correlations of 0.56 to 0.70. For six offices we traced each Computer-technology patent granted in 2019 to its ultimate parent company. Counting patents by the parent's country lowers the share by less than five percentage points at every office. Host economies therefore own slightly less than the measure shows.

Fact 1: Japan, Korea and the US patent a lot and own most

Table 1: Ownership and world share, large patent offices

Notes: Data are from WIPO patent statistics for 1996-2022. The table covers the 11 large patent systems, the offices that grant at least 100 Computer-technology patents a year, with the members of the European Patent Convention as one bloc. Ownership is the share of Computer-technology patents granted at a country's own office that go to its residents, averaged over the years in which the office granted such patents. World share is the country's share of all Computer-technology patents granted to residents worldwide, ranked among the 50 economies whose offices granted such patents. Frontier grants per year is the average number of Computer-technology patents granted at the office each year.

Table 1 compares ownership with the usual way of ranking countries, by their share of the world's frontier patents. Japan, Korea and the United States rank high on both. They are in the top four by world share. Their residents hold most of the frontier patents granted at home: 81% in Japan, 68% in Korea and 63% in the United States. At Europe's patent offices taken together the figure is 44%.

Fact 2: Canada, India and Australia patent a lot but own little

Canada, India and Australia rank in the top 15 of 50 economies by world share. That ranking mostly reflects the size of a country's patent system. Across 50 economies, the rank correlation between world share and the total number of patents an office grants is 0.87. For ownership the rank correlation is 0.08. Yet their residents hold only 14%, 12% and 8% of the frontier patents granted at home, the lowest shares among the large patent systems. The pattern is the same outside computing. In each of eight technology fields, including digital communication and pharmaceuticals, their residents hold between 1% and 25% of the patents granted at home.

All three are open economies with large markets. Foreign technology firms sell and produce in Canada, India and Australia. They also patent their technology at each country's patent office to stop domestic firms from imitating it. If anything, these shares overstate domestic ownership, because the local subsidiaries of foreign firms count as residents. In Canada, counting patents by the country of the parent company lowers the resident share from 10% to 7% in 2019, mostly because of subsidiaries of American firms. In India the same correction lowers the share by 1.2 percentage points, mostly because of subsidiaries of Korean firms.

Fact 3: China and India now own more of their frontier technology

Figure 1: Resident share of frontier patents, 1997-2021

Notes: The share of Computer-technology patents granted at each country's own patent office that go to its residents, as a three-year moving average. Data are from WIPO patent statistics.

The ownership shares for China and India show that ownership can rise quickly (Figure 1). Chinese residents held 29% of the frontier patents granted in China in 1996-1999, and 91% in 2020-2022. India's share rose from 7% in 2017 to 28% in 2022, as grants to Indian residents rose tenfold between 2016 and 2022. Japan, Korea and the United States stayed above 60% throughout. Canada and Australia stayed below 25%.

Fact 4: Ownership is higher with more R&D and better institutions

Many of the subsidised plants described at the start belong to foreign firms. One of the largest American awards, of up to US\$6.6 billion, went to the Arizona subsidiary of TSMC, a Taiwanese firm. Japan agreed to pay up to 476 billion yen towards a TSMC plant in Kumamoto. India's first approval went to Micron, an American firm, for a plant at Sanand. In each case the host country becomes a production site in a value chain run by a foreign firm.

Development policy often treats participation in global value chains as a route to technology transfer (World Bank, 2020), and Javorcik (2004) argues that foreign affiliates raise the productivity of their local suppliers. We test whether joining value chains brings ownership, using World Bank data on participation and comparing each economy with itself over time.

Across all economies we find no clear link between participation and ownership. In middle-income economies, ownership is lower when an economy supplies more inputs to other countries' exports. The association is large. When an economy's role as a supplier grows by a typical amount (one standard deviation), its resident share is about 17 percentage points lower. The fall comes from more foreign patents rather than fewer domestic ones. With the same increase, grants to foreign applicants rise almost fourfold, while grants to residents stay about the same. The pattern is consistent with lead firms keeping the intellectual property while suppliers get the production contracts (Gereffi et al, 2005).

In the same economies, ownership is higher where domestic capability is stronger. A typical increase in R&D intensity (R&D spending as a share of GDP) is associated with a resident share about 37 percentage points higher. A typical improvement in institutional quality, measured by the World Bank's governance indicators, is associated with a resident share that is also about 37 percentage points higher. China's rise in ownership came with a sustained rise in domestic R&D, from about 0.9% of GDP in 2000 to 2.6% in 2022.

Implications for industrial policy

The findings have three implications.

  1. Governments should add ownership to the way they measure technological sovereignty. Governments already track capability, through R&D spending and patent counts, and dependence on foreign suppliers, through trade data. The resident share of frontier patents adds a measure of who holds the rights to the technology used at home. Patent offices already publish the data needed to compute the resident share every year. Subsidy programmes are now judged by the investment, plants and jobs they bring. A rising resident share would also show whether they build domestic ownership.
  2. Governments that aim at technological sovereignty should fund design and research as well as production. A plant built by a foreign firm adds production at home, but the patents on its technology stay with the firm. Funding for design and research by domestic firms and universities creates patents held by residents. The European Commission is putting about EUR 500 million into a European Chip Design Platform, and India's design-linked incentive scheme supports chip design by Indian start-ups and small firms.
  3. Middle-income economies are more likely to gain ownership through domestic R&D and better institutions than through supplying foreign firms. In these economies, ownership is higher where R&D and institutions are stronger. Ownership has fallen as these economies became larger suppliers to firms abroad.

References

Danish, Mohd Shadab and Megha Patnaik. Measuring technological sovereignty: Who owns frontier technologies? CESifo Working Paper No. 13140 (2026).

Edler, Jakob, Knut Blind, Henning Kroll and Torben Schubert. Technology sovereignty as an emerging frame for innovation policy: Defining rationales, ends and means. Research Policy (2023).

Gereffi, Gary, John Humphrey and Timothy Sturgeon. The governance of global value chains. Review of International Political Economy (2005).

Guellec, Dominique and Bruno van Pottelsberghe de la Potterie. The internationalisation of technology analysed with patent data. Research Policy (2001).

Javorcik, Beata Smarzynska. Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages. American Economic Review (2004).

World Bank. World Development Report 2020: Trading for development in the age of global value chains. World Bank (2020).


Mohd Shadab Danish is faculty at the Indian Institute of Management Raipur. Megha Patnaik is at LUISS University and CEPR, and is cofounder of LEAP Insights and India Productivity Lab.