A Tale of Two Ecosystems: Why China Has Not Yet Surpassed the US in Original Innovation

"China risks repeating the experience of Japan or the Soviet Union—advancing to a certain stage only to stall, with any temporary lead fading as quickly as it appeared." — Wang Mingyuan

Today’s edition opens with an introduction from Rogier Creemers, an Assistant Professor in the Law and Governance of China at Leiden University. His research focuses on China’s digital technology policies and on the country’s growing role in global digital affairs. He is also a co-founder of DigiChina, a joint initiative with Stanford University and New America. Very grateful to him for his contribution to this newsletter. — Thomas

US-China technology rivalry is often viewed through the metaphor of a race. This can be attractive: it is a simple narrative within which a winner prevails at the end. Take the space race for example: the goal, as John F. Kennedy put it, was to put someone on the moon and return them safely to earth by the end of 1969.

In the current moment, however, a series of follow-up questions arises immediately. In an AI race with China, for instance, where is the finishing line? What is the prize? How do we know what “winning” means, and how do we get there? Unsurprisingly, this question has animated commentators around the world, including Wang Mingyuan, the author of the article below.

For quite some time, much focus has gone to indicators representing the cutting edge of technological innovation: quantitative assessments of Nobel prizes, published or cited research papers, granted patents and R&D investment, or specific examples of scientific and technological breakthroughs, in fields such as quantum computing, semiconductor manufacturing and AI. Yet this is only part of the story. Even if such indicators are a reliable proxy of innovative capacity – and there is ample evidence to suggest that numbers are often massaged – there is a significant difference between realising an impressive but one-off achievement and broad-based technological progress. In space terms: it is one thing to put someone on the moon, another thing to organise weekly cruises to the Sea of Tranquillity.

In Wang’s view, technological power also depends on technological diffusion: the capability to industrialise R&D output and apply it in social and economic domains broadly. This requires a welcoming ecosystem in which new applications are facilitated through supporting infrastructure and targeted investment, and can achieve commercial success. It is also not something that can be pulled out of a hat overnight, but requires broad, coordinated and diligent efforts.

Yet Wang’s article suffers from one major weakness, which it shares with uncountable similar analyses. It sees national economies as atomised, self-contained units, and pays little more than lip service to the global interdependencies required to support high-technology industries. Wang has little to say about what is necessary to retain access to global markets, resources, capital and talent, which provide the economies of scale that justify huge investments in technological advance. And this is, of course, the key political question: how does China – and the rest of the world – grapple with a fundamental shift in global economic clout, income streams and concomitant geopolitical strength?

— Rogier Creemers

Key Points

1. China has substantially closed the gap with the US in innovation, but since much of its progress has been driven by rapid economic growth, the pace of catch-up is now slowing as growth decelerates.

2. While China leads or rivals the US in its quantity of R&D personnel, research output and patent production, it lags in basic research and cutting-edge technologies, as disruptive technology still emerges from the US.

3. Government R&D spending is roughly on par with the US, but bureaucratisation and corruption in the R&D system have significantly reduced the effective utilisation of research funding.

4. Furthermore, while China’s vast consumer market enables rapid technological scaling, limited household purchasing power due to lower incomes constrains demand for high-value innovation.

5. Despite trailing in these areas, China’s strength in innovation stems from the organisational and diffusion capacity of its enterprises that have readily adopted and improved new technology since reform and opening-up.

New Year Special: 12 months for the price of 6.
Offer available to current subscribers until 15 January 2025.
To enjoy early access to select content and full access to Sinification’s archive, consider becoming a paid subscriber. As a non-profit, Sinification depends on your support.

6. Hence, the US is currently targeting China’s supply chains in order to undermine the capacity of its industrial system for commercialisation and diffusion, rather than targeting cross-border flows of basic research.

7. Sacrificing China’s advantage in diffusion to chase technological breakthroughs risks repeating the Soviet Union’s error of producing “surplus technology” misaligned with market demand.

8. Conversely, to let its enterprises fully realise their potential in technology diffusion and commercialisation, China needs an ecosystem-wide strategy focused on improving its business environment and rule of law.

9. Rather than focusing on duplicating “stranglehold” technologies from the West as in past state-led efforts, the priority should be to raise incomes, strengthen consumption and maintain an open and rules-based economy.

10. Technological leadership inevitably lags industrial dominance by several decades, so even with its dominance in manufacturing, it is natural for China to need 20–30 years to surpass the US in original innovation.

Name: Wang Mingyuan (王明远)
Date of birth: Not disclosed
Position: Researcher, Beijing Reform and Development Research Association, Beijing Federation of Social Science Circles (BFSSC)
Previously: Staff member and researcher, China Society of Economic Reform, including work at its magazine China Economic System Reform (2009 to 2013)
Research focus: History of China’s reform and opening up, international political economy and great-power competition
Other: Entrepreneur in Inner Mongolia and Guangdong; Columnist for the Caixin opinion section and Beijing News, and a contributor to FTChinese (FT中文网), The Paper, and Jiemian; runs the blog “阜成门六号院”
Education: BA, History, Chinese University of Hong Kong; MBA (Finance), School of Economics and Management, Tsinghua University

CHINA’S SCIENTIFIC RESEARCH IS BOOMING—BUT WHAT WORRIES THE UNITED STATES IS A DIFFERENT PROWESS
Wang Mingyuan (王明远）
Published by Phoenix on 30 October 2025
Translated and annotated by A.H.
(Illustration by OpenAI’s DALL·E 3)

I. The Structural Dynamics of the Sino-US Tech Rivalry

When the China–US tech war broke out during Donald Trump’s first term, most observers believed the outcome would become clear before long. However, the past few years have shown that, despite the “combination punches” [组合拳] of sanctions China has suffered at the hands of the United States, its capacity for scientific and technological innovation has grown against all odds, displaying remarkable resilience. This demonstrates that the United States cannot, as it once did with Japan, easily defeat its rival through a few tactical manoeuvres. Nevertheless, as the most powerful “empire” in history, it remains extremely difficult for the US to be surpassed by China in the short term. The competition between the two countries has now reached a stalemate, and it will only grow more intense in the years ahead.

I therefore believe it is far too early to judge who will ultimately prevail [Note: 鹿死谁手, literally “at whose hands the deer will die”] in Sino-US competition on the basis of the rise or fall of a given industry (whether AI or semiconductors) or the fluctuation of any single indicator. The technological rivalry between China and the United States is not a short-term or isolated contest; rather, it is a comprehensive and enduring struggle [全面的持久性竞争], one that at its core is a contest between the two countries’ social innovation ecosystems and their underlying potential.

So how can we accurately and comprehensively define an innovation ecosystem? After reviewing the World Intellectual Property Organization’s Global Innovation Index (GII), the World Economic Forum’s Global Competitiveness Index (GCI), the OECD’s Science, Technology and Industry Scoreboard, the European Union’s European Innovation Scoreboard (EIS), and the National Innovation Index Report published by the Chinese Academy of Science and Technology for Development, I arrive at the following view.

An innovation ecosystem is an organic system composed of key actors—universities, research institutions, governments, intermediary organisations, enterprises and consumers. These actors freely combine to form four major functional domains: the production domain, the application and commercialisation domain, the service domain and the consumption domain.

Accordingly, Sino-US technological competition can be broken down into a composite rivalry [组合竞争] across four capabilities: production, service, application and commercialisation, and consumption capabilities. Naming a victor [英雄, literally “hero”] based on one factor alone would be to mistake the part for the whole [盲人摸象, literally “blind men feeling an elephant”]. This article therefore compares and analyses the respective strengths and weaknesses of the two countries across these four capabilities.

II. China’s Growing R&D Strengths and Narrowing the Gap with the United States

For this issue [of R&D], I have primarily examined the output capacity of each country’s principal science and technology R&D actors—universities, research institutions, and enterprises—as this represents the most visible facet of technological competition.

The first dimension of comparison is the scale of R&D personnel and investment. This is the foundation of a country’s R&D capacity and often determines both the completeness of its industrial system and the extent of its coverage in emerging technological fields. For example, in 1990s Japan, the number of R&D personnel and amount of investment were over half that of the United States. Today, these two figures have fallen to around 20% of US levels, leaving Japan with insufficient capacity to invest in emerging industries [Note: The percentages of Japan’s R&D investment relative to the US are supported by current OECD data, but the figures on R&D personnel could not be independently verified as the US does not appear to have published data on total R&D personnel in the 1990s]. As a result, Japan now has virtually no presence in areas such as the digital economy, artificial intelligence, and renewable energy.

In terms of R&D workforce and investment, China is the world’s fastest-growing country. In 1991, China had only 660,000 people working in R&D—around one-fifth of the number in the United States. By 2024, that figure had risen to 7.24 million, surpassing the estimated five million in the US [Note: These figures could not be fully verified, as the United States does not appear to have published data on total R&D personnel in the 1990s].¹

At present, the United States has about 3.5 million researchers with doctoral degrees, compared with roughly 1.4 million in China [Note: This data could not be verified]. Considering that China currently has around 610,000 doctoral students in training, compared with only 178,000 in the United States, China’s stock of highly qualified engineers is expected to surpass that of the US within the next decade. [Note: The US figure could not be independently verified—according to the National Centre for Science and Engineering Statistics (NCSES), in 2023, there were 152,610 US citizens and permanent residents in full-time doctoral programs in the fields of science, engineering, and health].

The quality of education in China is often criticised. Yet if given sufficient employment opportunities in the future, these systematically trained young people can still be transformed into high-end human resources.

Looking at R&D expenditure, China has doubled its spending in the space of just seven years, while the US has grown at only around 4%. According to data from China’s Ministry of Science and Technology (MOST), China’s R&D expenditure currently stands at around 496 billion USD, whereas the United States’ is 713.2 billion USD.² Although a gap remains, China is expected to surpass the United States within the next five to ten years. Moreover, according to the European Commission Joint Research Centre’s Industrial R&D Investment Scoreboard (2014 to 2023 editions), the United States has consistently accounted for around 800 of the world’s 2,500 highest-spending R&D companies, while China’s share has risen from 199 to 679, suggesting that it may soon overtake the United States.

The second dimension of comparison concerns research output. China has already surpassed the United States in several basic output indicators. In 2024, according to the Nature Index, which tracks 145 leading scientific journals, China contributed 32,122 papers, compared with 22,083 from the United States—146% of the US figure (in 2015, China’s share amounted to only 37% of the US total). Chinese scholars published 845,000 papers indexed in the SCI [Science Citation Index], while US researchers published 483,000. In the patent domain, data from the World Intellectual Property Organization show that China filed 70,160 PCT (Patent Cooperation Treaty) international patent applications in 2024, compared with 54,087 filed by the United States.

Evidently, as China’s economic strength has grown, we have already largely surpassed the United States in terms of scale and basic research output—a historical turning point that occurred around 2018 to 2022. However, two points merit attention:

First, China’s advantage lies primarily in the quantity of general research output. In basic research and cutting-edge technologies, it still lags far behind the United States. For example, according to the Research Fronts 2023 [2023 研究前沿热度指数] report from the Chinese Academy of Sciences and Clarivate, among frontier research in eleven major scientific and technological domains, China surpasses the United States in only agricultural sciences, chemistry, and materials science. In all other areas—especially medicine, mathematics, and physics—China trails the United States by a substantial margin. In global frontier science and technology R&D, the United States [ranks first in] 53.9% [of the research frontiers], while China [leads in] 24.2%. In practice, disruptive innovations still mainly originate in the United States.

Second, the high growth of China’s research expenditure is reliant on the country’s rapid economic expansion. As China’s economy slows, so does the pace of growth in research funding, which in turn affects its potential to pull ahead. China risks repeating the experience of Japan or the Soviet Union—advancing to a certain stage only to stall, with any temporary lead fading as quickly as it appeared [昙花一现的覆辙].

III. China’s Key Advantages in Industrial Organisation and Technology Diffusion

From the standpoint of great power competition, it is the capacity for technology diffusion and industrial commercialisation [产业转化]—more than scientific research output—that determines victory or defeat. For example, during the US-Soviet rivalry, although the Soviet Union was not far behind the United States at the initial stages of R&D in fields such as the internet, computer terminals, and telecommunications, its extremely weak capacity for application and commercialisation meant that it never secured a real position in these industries. Nor could it develop the kind of general-purpose technologies capable of driving a new technological revolution across society.

The same is true of today’s technological competition. The United Kingdom, Canada (which has produced 28 Nobel laureates), and Australia are all countries with high scientific creativity, but their industrial organisational capabilities are comparatively weak (with the latter two affected by resource dependence), leaving them far behind in industrial competition. Japan and South Korea, by contrast, are relatively low in scientific creativity but possess strong capacity for industrial commercialisation, and thus hold an advantage in industrial competitiveness.

I believe that application and commercialisation capability is reflected in three dimensions: the commercialisation rate of invention patents, industrial organisational capability, and technology-diffusion capability.

In terms of patent commercialisation rates, China currently stands at roughly 39% [in 2023], compared with about 60% in the United States and the European Union and 51% in Japan [Note: The figures for the European Union and United States could not be independently verified, while Japan Patent Office (JPO) estimated the domestic patent utilisation rate to be 51.8% in 2020 and around 54.2% in 2023].

China’s lower overall commercialisation rate is primarily due to the extremely low commercialisation rates of public universities and research institutions, which are approximately 11% and 4% respectively [Note: Wang may have reversed the two values—according to data from the China National Intellectual Property Administration (CNIPA), the patent industrialisation rate [产业化率] in 2022 was approximately 3.9% for universities and 13.3% for public research institutions]. By contrast, the patent commercialisation rate among Chinese enterprises reached 53% [in 2024], roughly on par with that of developed countries. In particular, the commercialisation ratio of research achievements among China’s leading private enterprises has now reached a level comparable to that of top American firms.

With regard to industrial organisational capability—the speed and scale at which innovations are converted into commercial applications—the secret to America’s rise lay in its upending of traditional industrial organisation models, which produced capabilities far stronger than those of Europe. However, after maintaining this advantage for more than a century, the United States has gradually been overtaken by China in this respect. Aside from a few high-profit industries such as semiconductors and pharmaceuticals, the United States now falls short of China.

We find that China’s industrial organisational advantage grows stronger for more emerging industries. For example, of the 20 best-selling electric vehicles worldwide in 2024, 16 were produced in China, while only two came from the United States. Among the world’s top 10 mobile phone manufacturers, nine are Chinese [Note: This claim is difficult to verify without a clear definition of “manufacturer”, but data from major market research firms such as Counterpoint and IDC indicate that at least two of the world’s top five smartphone makers in 2024 were non-Chinese]. China could increase its automobile production capacity by 20 million units in just five or six years [Note: The timeframe of Wang’s projection is unclear. For reference, China’s vehicle production increased by roughly six million units between 2019 and 2024, per data from the NBS]. Companies such as Xiaomi and Pinduoduo have taken only nine years to grow from start-ups to Fortune Global 500 companies. All of these examples demonstrate that China currently possesses the strongest industrial organisational capability in human history.

Another aspect that is often overlooked and poorly understood is a nation’s capability for technology diffusion. Every technology is originally created for a specific purpose. But if a vibrant ecosystem for diffusion exists, its use scenarios will diversify, allowing it to become widely applied across different domains and enabling it to develop into a general-purpose technology (GPT) capable of driving societal technological revolutions.

The steam engine, for instance, was invented by the French, yet for several decades it served merely as equipment for hauling artillery. It was British engineers who extended its application to smelting, textiles, and transport—transforming it into the driving force behind humanity’s First Industrial Revolution. Similarly, the internet began as a tool for information sharing within military and research institutions, but over the past thirty years it has spread to become a ubiquitous foundational technology.

In technological competition, the decisive factor is therefore often not who is first to develop advanced technology, but who is more capable of technology diffusion and industrial commercialisation. Over the past decade or so, China has gradually acquired one of the world’s strongest capabilities for technology diffusion.

For example, driven by companies such as Baidu, Alibaba, and Tencent, China has become the country with the highest internet penetration rate in the world. [Note: The World Bank estimates that 92 percent of China’s individuals used the internet in 2024, placing its penetration rate behind countries such as the United States, Russia, Singapore, Saudi Arabia, and the UAE]. Companies like DJI have broken the long-standing pattern in which drones abroad were largely confined to military use, and they are now gradually transforming them into indispensable tools for logistics, agriculture, and emergency relief. In the emerging field of artificial intelligence, Chinese internet platforms have also helped popularise application scenarios for this new technology. Recently, platforms such as Tencent Yuanbao and Baidu Wenku have actively integrated DeepSeek, allowing this foundational model to move rapidly from specialised domains into the hands of over a billion users.³

Why, then, does China possess such formidable application and commercialisation capability? Many people attribute it to the vast size of China’s market, but I would argue that while market scale is a necessary condition, it is not a sufficient one. If market size were decisive, why did China not undergo an industrial revolution before the 1990s? And why does India’s industrial development still lag behind today?

What brought about China’s revolutionary transformation in this capability was the emergence, after reform and opening-up, of a cohort of enterprises and entrepreneurs with the world’s strongest pioneering spirit and organisational capability. They became the crucial link between technology and commerce, between laboratories and consumers—connecting key resources [资源要素] from across China and setting off the “chemical reaction” [产生化学反应] that enabled the dramatic transformation of China’s innovation capacity and industrial competitiveness.

Thus, although China is not currently a centre of cutting-edge technological R&D, it can nonetheless possess the strongest industrial competitiveness. Consequently, the focus of current US sanctions against China is on supply chain measures aimed at undermining China’s capacity for industrial commercialisation—rather than, as in the McCarthy era, restricting the movement of scientists and preventing the formation of scientific research capacity. Likewise, both Joe Biden’s and Donald Trump’s efforts to promote the return of manufacturing fundamentally reflect an attempt to revive the United States’ capability for [technology] application and commercialisation. All of this illustrates the critical importance of this domain in the competition between the two countries.

IV. Funding Inefficiencies in China’s Innovation Ecosystem

The main actors in this area are governments, financial institutions, and intermediary organisations. When discussing service and support capacity, people generally think of the government’s ability to mobilise resources to support scientific research. According to the National Science Foundation’s 2024 Research and Development: U.S. Trends and International Comparisons report, the US federal government provided 159.8 billion USD in R&D funding in 2022, while the Chinese government spent approximately 622.6 billion yuan (equivalent to 89.5 billion USD). Of this, 381.4 billion yuan was direct appropriations [to government-affiliated research institutions]; considering that most Chinese universities are public, the 241.2 billion yuan they spend on R&D also largely comes from government support [Note: The China R&D spending data comes from the NBS].

Moreover, China’s fiscal expenditure on science and technology in 2022 was 1.1128 trillion yuan (approximately 159.9 billion USD). By this measure, the two countries are broadly on par, with Chinese government R&D spending slightly exceeding that of the United States.

However, entrenched problems in China’s research system, such as bureaucratisation and corruption, have significantly reduced the effective utilisation of R&D expenditure. According to a 2011 survey by the China Association for Science and Technology, only about 40% of research funds were actually used for the projects themselves, with large amounts of funding being diverted elsewhere [Note: The survey with the data cited here, the first National Survey Report on the Status of Science and Technology Workers 全国科技工作者状况调查报告), was published in 2004, not 2011].

Furthermore, a 2022 audit conducted by the Guangxi Audit Office on nine regionally administered universities found problems in the use of research funds across 69% of projects [Note: The audit report does not appear to mention this specific data point, but did find that some universities had problems with idle surplus funds due to improper budgeting, with 22 projects in two universities having a budget variance of up to 69.24%]. Consequently, there is likely to be a substantial gap between China and the United States in terms of direct R&D expenditure.

Innovation is a high-risk activity—one that requires a well-developed financial support [system]. The United States has the world’s most advanced venture capital and securities financing markets. According to data from Zero2IPO Research, venture capital and private equity investment in the United States totalled 7.4602 trillion yuan in 2024, whereas the figure for China was only 603.6 billion yuan—just 8 percent of the US level. Even at its peak in 2017, China reached only 26 percent of the US figure.

In securities financing, 272 companies successfully went public in the United States in 2024, raising a total of 40.7 billion USD, while 228 Chinese firms raised 23.4 billion USD through IPOs.

China has also made notable improvements to its business environment in recent years. According to the World Bank’s [Ease of Doing Business] rankings, China rose from 91st place in 2012 to 31st in 2020, while the United States remained consistently within the top six. Nevertheless, Chinese enterprises still lag far behind their US counterparts in areas such as tax friendliness, cross-border trade convenience, and the ease of obtaining business services.

The rule of law index is another key factor. According to the World Justice Project’s Rule of Law Index, the United States ranks around 20th globally, while China stands around the 80th position [Note: This is largely true prior to 2020, but in the years since, both countries have dropped in the rankings, with the United States hovering around the 26th position and China the 95th]. Above all, in the area of intellectual property protection, the United States has the world’s most developed systems for intellectual property protection, commercial valuation, and patent agency services. By comparison, according to the International Intellectual Property Index, China ranks 25th among 55 assessed economies, placing it at a moderate level.

Therefore, compared with the previous two areas, China has clear shortcomings in service guarantees, which limits the ability of entities to fully realise their potential in R&D and technological application and commercialisation. It is also worth noting that, according to the World Intellectual Property Organization’s GII analysis and evaluation, both China and the United States have experienced a downward trend in institutional services and innovation environments over the past decade or so. Under these circumstances, China should seize the opportunity to accelerate reform and narrow its gap with the United States.

V. The Weakness of a Large Market With Low Purchasing Power

China’s strength lies in the fact that it possesses the world’s largest single market in terms of number of consumers. A large single market can better help companies survive the most difficult stages of early development. For example, if a technological innovation requires five million customers to make it steadily through the start-up phase, it would need only a 0.35% market share in China, whereas in South Korea it would require 15% [Note: Wang’s 0.35% figure for China is calculated using China’s total population; applying the same method to South Korea’s population of roughly 51 million results in a market share of about 10%, not 15%]. The difficulty of scaling up in the two cases is vastly different.

Furthermore, as firms expand [in China], they do not need to reorganise their sales models, reconfigure their product development strategies, or adapt to different government regulations, significantly reducing operational costs. This explains why today’s major technology companies are concentrated primarily in China and the United States, and why they are exceedingly rare in countries with populations below 50 million.

However, China’s consumer potential has yet to be fully realised, and there remains a considerable gap between Chinese and American purchasing power. In 2023, China’s total private consumption expenditure amounted to 703 million [sic] USD, whereas the United States recorded 18.5 trillion USD [Note: Wang likely intended ~7 trillion USD for China’s consumption; according to the NBS, China’s total private consumption expenditure in 2023 was 49 trillion yuan, approximately 7 trillion USD]. Thus, although the overall economic output of the two countries is not vastly different, their consumption capacities differ sharply. The per-capita purchasing power of Chinese consumers is equivalent only to that of Americans in roughly 1972.

Looking at the price of major consumer goods, China sold over 23 million new cars, maintaining its position as the world’s largest automobile market and surpassing the United States by seven million units.⁴ However, the average price per car was only 24,000 USD, compared with 47,000 USD in the US. Consequently, China still falls far short of the United States in total automotive consumption expenditure.

Similarly, China purchased over 270 million mobile phones in 2023, but only around 70 million of these were high-end models priced above 600 USD, accounting for less than 26 percent of total sales. In contrast, around 86 million high-end phones were sold in the United States, accounting for as much as around 60% of the market [Note: The US mobile phone data could not be independently verified].

This disparity in consumption capacity means that Chinese enterprises are naturally inclined toward developing low-cost, high-volume products, which limits their ability to invest in high-value-added, top-tier products. In contrast, the stronger private consumption power of Americans naturally incentivises firms to develop products at the upper end of the value chain.

For this reason, simply relying on the government to introduce industrial support policies or increase R&D funding will not fundamentally resolve the problem of enhancing China’s industrial competitiveness in the upstream segments of the value chain. It is essential to raise household incomes and strengthen consumer purchasing power. If 200 million people in China had annual incomes above 300,000 yuan, many of the country’s innovation bottlenecks would be addressed in one fell swoop [纲举目张、迎刃而解].

VI. China’s Long-Term Path to Technological Leadership Requires an Ecosystem-Wide Strategy

First, as the comparisons above show, the key determinants of the innovation ecosystems in China and the United States each have their own strengths and weaknesses. Overall, China is either gaining—or has already secured—an advantage in research output and commercialisation, but it still has clear disadvantages on the service and support side and on the consumer side.

Since reform and opening-up—and especially after joining the WTO—China’s innovation ecosystem has undergone fundamental growth. The most remarkable achievement of reform and opening-up is not simply the increase in GDP, but the establishment of a vibrant, market-driven innovation ecosystem. This marks the essential difference between China’s rise and that of the Soviet Union.

At present, China’s innovation competitiveness is drawing ever closer to that of the United States and has already reached roughly 70% of the US level. However, the pace of catching up is slowing. Whether China can reach the historic turning point of surpassing the United States remains a fifty-fifty possibility.

Whether China or the US will have the last laugh ultimately depends not so much on government interventions—which can have significant short-term and localised effects—but on which side can uphold an order grounded in the rule of law, openness, and inclusiveness, thereby allowing the potential of its innovation ecosystem to be realised in the most enduring way. This is what underpins the confidence of society as a whole, as well as of international talent and capital, as the experience of the past decade has already demonstrated.

Second, when examining the internal elements of the innovation ecosystem, we can see that the United States’ development is relatively balanced across its various components, reflecting a mature innovation ecosystem. China, by contrast, exhibits both pronounced strengths and weaknesses. From the perspective of the key actors, China’s strengths are largely associated with individuals and enterprises, whereas its weaknesses—such as technological innovation efficiency, the business environment and consumer capability—are directly or indirectly linked to institutional factors.

Therefore, optimising China’s innovation ecosystem and enhancing its innovative capacity depend on building a modern, service-oriented government, improving the rule of law, strengthening the vitality of the financial system, and reforming income-distribution mechanisms—thereby unleashing the potential of its innovation actors.

I believe that innovation ecosystems follow the short-board theory [短板理论]: if China cannot remedy its weaknesses, they will eventually constrain the effectiveness of its strengths.⁵ In particular, as the pace of China’s market expansion slows sharply—or even contracts—the constraints imposed by institutional and consumer shortcomings on innovation growth are becoming increasingly apparent.

Third, China needs to adopt a holistic, ecosystem-based perspective in its strategic thinking about competition. Since the onset of the China–US tech war, a prevalent view at home has been that China can prevail simply by mobilising all its resources to achieve breakthroughs in a few “stranglehold” [卡脖子] areas, as it once did in developing the “Two Bombs, One Satellite” [两弹一星] [projects].⁶ In other words, the focus of competition has been placed on the government and on boosting scientific research output.

Contemporary China–US tech competition ultimately comes down to the global influence of products and services. Relying solely on the early stages of the innovation chain and on expanding R&D capacity is of very limited significance. It is more essential to strengthen the capacity for industrial commercialisation, together with the service and consumer capacities that support it.

Without commercialisation, technological progress simply leads to surplus technology [技术的过剩]. The Soviet Union of the 1970s and 1980s is a classic example: it had strong scientific research but lacked strong industry, and in the end, its patents served little purpose beyond functioning as evaluation criteria for the professional titles of engineers.

In summary, compared with the traditional industrial era, the global innovation ecosystem has undergone a dramatic transformation. It is essential to adopt new ways of thinking and establish an ecosystem-based perspective [生态系统观]. Focusing narrowly on a single actor or function while neglecting the others is, in fact, damaging to the ecosystem and can only lead to a decline in innovative capacity.

Fourth, the migration of a global technological centre is a slow process that lags behind the migration of industrial centres. China must remain composed and avoid impatiently seeking quick success [急于求成]. After the United States became the world’s industrial hub, it took another forty years before it became the global centre of science and technology (1894 to 1934). Even if China can maintain its current manufacturing advantages, it will still require at least two to three decades of accumulated [experience] before it can hope to surpass the United States in original innovation capacity.

Therefore, in strategic terms, China should draw on the experience of how Britain overtook France and how the United States surpassed Europe: it must first further consolidate its capacity for industrial commercialisation before [moving to] enhance its capacity for original innovation. Only by holding on to its strengths can China make up for its weaknesses.

There is now also a view in society that China’s shortcomings in core original innovation are the fault of its enterprises being too strong in application and commercialisation, with negative perceptions particularly [directed at] large corporate platforms. From this, some argue that to strengthen their core originality, firms’ [focus on] commercialisation and application must be restricted.
This line of reasoning is entirely muddled and runs counter to basic principles of modern industry. It is like claiming that your poor Chinese-language grades are caused by being too good at mathematics and therefore banning you from studying maths so you can focus solely on Chinese. This only results in losing on both fronts [两盘结空].

China’s insufficient capacity for original scientific innovation has primarily arisen from the accumulation of a comprehensive set of historical factors that have left its [resulting foundation] too weak. Any view that pits original innovation against commercialisation and application is playing directly into the hands of the competitor. We must cherish our strengths; they are our most valuable capital, not the source of our shortcomings [而不是罪魁].

READ MORE

1. The estimate of 660,000 R&D workers in China in 1991 is broadly consistent with the OECD MSTI series; however, the number of “researchers” in China that year was 471,400, compared with 792,028 in the United States. For 2022, the most recent year with comparable data, the US reported roughly 2.7 million FTE R&D personnel, not five million.

2. Wang appears to be comparing R&D expenditure from different years. While this exact US figure can’t be verified, it corresponds more closely to 2022, whereas the Chinese figure aligns more closely with 2024 levels (approximately 3.61 trillion yuan according to the National Bureau of Statistics (NBS)). NCSES has not yet published US R&D expenditure for 2024; its estimate for 2022 is 761.6 billion USD in constant 2015 PPP-adjusted dollars, and its nominal estimate for 2023 is 940 billion USD.

3. Tencent Yuanbao is an AI assistant app that is integrated into Tencent’s wider ecosystem, including WeChat, and Baidu Wenku is a document and content creation app.

4. Wang appears to be comparing China’s 2024 passenger vehicle sales with total US light vehicle sales. The actual gap may be much wider: according to the US Bureau of Economic Analysis, 15.9 million light vehicles were sold in the United States in 2024, of which just under three million were passenger vehicles, the remaining 12.9 million being “light trucks”. While some of these may be used as passenger vehicles, it is unlikely all of them are.

5. The short-board theory, also known as the barrel principle (木桶原理), Cannikin’s Law or Liebig’s Law of the Minimum, posits that a system’s overall capacity is determined by its weakest link. It is a commonly used analytical metaphor in Chinese policy discourse, and Wang has invoked it in earlier writings on China’s scientific and technological development.

6. This refers to the development of China’s atomic and hydrogen bombs (first tested in 1964 and 1967 respectively) and the launch of its first satellite in 1970.