In 1856, an eighteen-yearold student named William Henry Perkin, working in a makeshift home laboratory in London, accidentally synthesised the world’s first synthetic dye while attempting to produce quinine. He called it mauveine. Queen Victoria wore it to her daughter’s wedding. The colour became a craze. Britain had every conceivable advantage. The discovery, the inventor, the empire as a captive market, and the Industrial Revolution’s factories still humming. By 1914, Germany produced roughly ninety percent of the world’s synthetic dyes. BASF, Bayer, Hoechst, all German. Perkin’s own factory had shut decades earlier. Britain invented the technology. Germany took the industry.
The United States and China are replaying this drama at civilisational scale. The conventional framing of the US-China technology competition is a race about who gets to the finish line first. This is the wrong image. China has not been running America’s race. It has been running a different one, aimed not at discovery first, but at production, scale, and control of the foundational layers that advanced economies depend on to function. The distinction matters enormously. The country that invents something is not necessarily the country that controls it.
Consider a few numbers. Both the United States and China have now crossed the one-trillion-dollar threshold in annual research and development expenditure. In purchasing power terms, China has pulled ahead. Between 2003 and 2007, the United States led in more than ninety percent of critical technologies. China led in roughly five. By 2025, according to the Australian Strategic Policy Institute’s Critical Technology Tracker, those proportions have almost inverted. China now leads in nearly ninety percent of such areas. In the Nature Index (which tracks publications in the world’s most selective journals) China posted a seventeen percent advantage over the United States in 2024. Chinese entities filed roughly 1.8 million patent applications that year, the United States filed 603,191. They mark a structural shift.
The shift is not, however, uniform. China’s business sector contributes about eighty percent of the country’s research and development spending, but barely one percent of that business investment goes to fundamental research. The comparable American figure is six percent. There is a gap between application and discovery, and Beijing knows it. Xi Jinping said as much in April, urging the country to strengthen basic research and raise its capacity for “original and disruptive innovations.” The five-year plan running from 2026 to 2030 designates businesses as the “main entity” of innovation, a delegation of decision-making authority that is historically novel for China’s state-directed system. DeepSeek’s R1 model, built entirely by domestically trained Chinese talent and the first mainstream large language model to undergo full peer review in a leading academic journal (in Nature, 2025), is early evidence that the gap is narrowing. The point is not that China has perfected its innovation system. The point is that it is correcting its weaknesses with some urgency, while the United States is not.
Where China has structured itself deliberately, the United States has drifted. Federal research and development investment, which peaked at 1.86 percent of gross domestic product in 1964, had fallen to 0.66 percent by 2021. The number is still falling. The CHIPS and Science Act of 2022 gestures toward correction, but a coherent and integrated industrial strategy (the kind that connects basic science to manufacturing to deployment) remains elusive. When manufacturing migrates, process engineering follows. When process engineering follows, the feedback loops that generate the next generation of innovation degrade. Daron Acemoglu and Simon Johnson have written about how the Industrial Revolution was driven by people who built things, not merely people who theorised. A country whose people stop building loses, over time, the capacity to advance what it once invented.
The talent problem compounds everything. Seventy percent of the top US-based AI researchers were born abroad. Sixty-five percent of the leading American AI companies have at least one immigrant co-founder. These figures are not accidental; they reflect the structural choice America made after the Second World War to remain open to the world’s best minds. That choice is now being unmade with some energy. International student enrolment at American universities fell by an estimated thirty to forty percent in 2025. The H-1B fee hike, the climate of generalised scrutiny, and the China Initiative’s legacy are, individually, defensible as security measures. Together, they amount to what one researcher has called the “stockyard paradox”: securing the assets while hollowing out the system those assets depend on.
The cultural dimension is the most underexamined. Ninety-three percent of Americans report some level of mathematics anxiety; many comfortably self-identify as “not a math person.” Academic excellence in China carries social prestige and functions as a primary pathway to mobility. China will produce roughly 12.22 million university graduates in 2025 alone, more than double the total degrees awarded at all levels in the United States. The National Science Board’s Science and Engineering Indicators now shows China surpassing the United States as the leading producer of STEM doctoral degrees globally. These are not just numbers about inputs. They are numbers about the culture within which innovation is nurtured or suppressed. The competition, in this sense, is not primarily about chips or subsidies. It is about whether a society decides that mathematical excellence is as admirable as athletic excellence.
There is a further irony that Perkin would have recognised. The United States invented the internet, the transistor, GPS, and the mRNA vaccine. It built the firms (Anthropic, Google, OpenAI) that remain, for now, at the frontier of artificial intelligence. But a country that hollows out its research base does not merely fall behind. It progressively loses the capacity to benefit from its own discoveries, because absorbing and applying cutting-edge knowledge requires trained people, functioning institutions, and active participation in global scientific networks. You cannot freeride on the science of others if you have dismantled the machinery to evaluate, translate, and use it.
The question is not whether China is winning. It is whether the United States is still competing on the terms that once made it dominant, openness, sustained public investment, and the recognition that innovation is not a moment of discovery but an ecosystem sustained across decades. The 1964 peak in federal research spending was not a coincidence. It came six years after Sputnik, that small Soviet satellite whose beep concentrated an entire civilisation’s mind.
The data suggest another such signal has arrived. It is just quieter than the original.
Aditya Sinha writes on macroeconomics and geopolitics. Shaurya Pandey is a Technology & Strategy Consultant
