New Analysis reveals Chinese Nationals now outnumber British Students in key STEM Postgraduate Courses at Britain's Top Universities
Chinese scientists are transforming Britain’s elite institutions.
There are now more Chinese nationals than British nationals studying physics, engineering, chemistry, mathematics or computer science at a postgraduate level at Britain’s top five universities for those subjects. In engineering, there are roughly twice as many Chinese nationals as British at that elite level (see below).
The trends expressed in the graphs above may have important ramifications for Britain’s prosperity and security but have emerged without receiving much attention outside of higher education circles.
This blog post is divided into five parts, as follows:
Analysis and presentation of the data itself.
Discussion of the financial situation in the higher education sector and STEM.
Discussion of the government’s ‘research security’ measures, and presentation of new data obtained through Freedom of Information (FOI) about their operation.
Discussion of the broader geopolitical context and the revolution in science and technology currently taking place in China.
Conclusion.
This is a factual blog, the intention of which is to inform and prompt debate about an important subject, not to attribute blame for anything or promote preconceived policy recommendations.
1. The data
UKCT’s data was commissioned from the Higher Education Statistics Agency (HESA). UKCT has published several dozen graphs and further notes on the data at this link and will consider sharing further data with anyone who would like to see it.
The data breaks down the total number of people from each country studying STEM subjects at each university, and at every level (from undergraduate to doctoral research). It covers the academic years 2019/2020 through to 2023/24. We are currently in the process of procuring the latest data for 2024/25 and will post new information derived from it in due course.
The ‘level of study’ categories are as follows: First degree, Other undergraduate, Master (Research), Masters (Taught), Postgraduate (Research), Postgraduate (Taught), Doctoral (Research) and Doctoral (Taught). The data covers all British universities. It is differentiated by subject within the Common Aggregation Hierarchy (CAH) framework, specifically, at CAH level 1 for Physical sciences (CAH07, this includes chemistry), Mathematical sciences (CAH09), Engineering and technology (CAH11) and Computing (CAH11); and at CAH level 2 for Bioscences (CAH03-01, a category which regrettably for STEM analysis includes CAH03-01-09 Hair and Beauty Sciences, but thankfully excludes CAH-03-02 Sport and Exercise Sciences.)
Across STEM as a whole, the raw number and proportion of Chinese nationals at the top five STEM institutions (treated here as Cambridge, Oxford, Imperial, UCL and Edinburgh, which tend to top Times Higher Education lists of top universities for various STEM subjects) increased significantly over the period 2019 to 2024, whilst the proportion of British nationals declined. This was true both at postgraduate levels and at all levels.
At Russell Group universities too, across STEM as a whole, the raw number and proportion of Chinese nationals increased over the period 2019 to 2024, whilst the proportion of British nationals declined. This was true both at postgraduate levels and at all levels. In engineering at Russell Group universities, there were nearly twice as many Chinese as British postgraduate students.
As illustrated above, it is evident from the data that, overall, there are proportionally more Chinese nationals at top five and Russell Group institutions than across the sector as a whole; and that there are proportionally more Chinese nationals at postgraduate than at undergraduate levels.
Engineering is the only subject which is dominated by Chinese nationals. It is the STEM subject studied by the largest number of people overall.
At the top five institutions for the subject (Oxford, Cambridge, Imperial, Manchester and UCL), there were in 2023/24 nearly twice as many Chinese as British nationals at postgraduate level, with Chinese making up 41% and British only 23%. The percentages were the same in the broader Russell Group. By contrast, Indian nationals made up just 4% at the top five and 8% in the Russell Group.
Looking beyond the Russell Group to all universities, but continuing to focus on the postgraduate level, there were roughly the same number of British as Chinese or Indian nationals.
In the interests of brevity, this blog does not discuss the data in further detail. Anyone wishing to learn more should contact the graph supplement or ask UKCT for access to the data. Some further graphs are displayed below.
2. The state of the sector
This change has taken place largely without comment or debate as to what it means for the public interest, except – perhaps – within the sector where this data is unlikely to surprise many.
The sector is in financial crisis, and international students pay more than home students. Indeed, they subsidise them. Most home STEM students cost more to teach than they pay in fees. The Russell Group has suggested that “the UK’s pipeline of STEM graduates is at risk as a result of continued underfunding from government. STEM subjects are inherently expensive to deliver and without appropriate funding they could become financially unsustainable”.
STEM graduates are key to innovation, economic growth and strategic capability, especially in an era of supply chain insecurity and geopolitical risk. In the UK, STEM graduates are too scarce, with widely reported vacancies in high tech industries and difficulties in finding sufficiently skilled and qualified candidates. “In 2022 the [the Department for Education] found that over 40% of vacancies in the manufacturing, information and communications, and construction sectors remained unfilled due to skills shortages. A 2022 report by the Institution of Engineering and Technology said there was a shortfall of more than 173,000 workers across the STEM sector…”
In parallel, the number of teenagers taking key STEM subjects at A-level has increased, as has the proportion of teenagers applying for STEM undergraduate degrees. Sadly, public demand for STEM degrees is not being met due to the lack of funding. Per the Financial Times, “UK applications for so-called high-cost undergraduate degrees have risen 14 per cent since 2019, the first available year of data, outstripping the 5 per cent increase in enrolments over the same period…”
As well as teaching, research funding continues to be squeezed, and this too impacts innovation and the wider economy.
3. Research security
During the ‘golden era’ of UK-China relations from roughly 2014 onwards, the British Government sought to expand scientific exchange between the UK and China. The blossoming of partnerships that resulted has been partially mapped, for example in a study by RAND.
This expansion of collaboration even extended to the establishment of several programmes based at British universities which fed directly into Chinese military modernisation. An example is illustrated below.
From roughly 2019 onwards, these kinds of partnerships received greater levels of scrutiny. The British government has put in place or strengthened multiple new pieces of legislation, bureaucracies and regulatory measures to enhance ‘research security’, which intends to address the risks associated with dual-use or strategic technology and knowledge being transferred from the British higher education ecosystem to countries including the People’s Republic of China, where government ‘military-civil fusion’ (军民融合) policies risk it being put to use for the military or cyber warfare. UK-China Transparency has obtained some new information about these measures through Freedom of Information (FOI) requests sent by volunteers and contacts of the organisation.
The most important measures are the following:
The work of the Export Control Joint Unit (ECJU) within the Department for Business & Trade (DBT). The ECJU primarily deals with the export of goods and information in the private sector. UKCT has established through FOI (view the data here) that the ECJU’s staff number increased from 82 in 2019 to 123 at the beginning of 2024, and its budget from £4.8m to £7.1m.
The establishment in 2021 of the Research Collaboration Advice Team (RCAT) within the Department for Science, Innovation & Technology (DSIT). UKCT has established through FOI (view the data here) that RCAT increased its provision of advice from 350 “pieces” of advice in 2023 to 1457 in 2024 and has increased its staffing from 14 to 19 people.
The establishment of the Investment Security Unit (ISU), stemming from the National Security & Investment Act 2021. ISU now sits within the Cabinet Office. It publishes lengthy annual reports , the latest of which for 2024-25 describes receipt of “1,143 notifications, an increase from 906 notifications received in the previous reporting period”, again suggesting a growing footprint, although there is no mention of staffing or budget.
The Academic Technology Approval Scheme (ATAS), which is a Foreign, Commonwealth & Development Office (FCDO) scheme whereby incoming postgraduate STEM students and researchers coming from certain countries, including China, are security screened. ATAS applications have soared since 2020. UKCT’s FOI regarding ATAS is discussed further below.
The National Protective Security Authority (NPSA), which is part of MI5. It provides “campaigns, training, guidance, and advice”.
According to a 2023 report from the Association of Research Managers and Administrators (ARMA), new research security requirements are proving difficult to fulfil across the sector: “issues of ownership and accountability are common, in-house processes are often burdensome and myriad actors and stakeholders offering guidance in this space only serves to cause confusion and this makes it difficult to keep up to date.” The Russell Group and UniversitiesUK have called for a special fund to improve their capacity.
Britain’s new ‘research security’ measures have nonetheless had impact. Since 2018, formal collaboration based on institutional agreements between Chinese organisations and British universities tends to have decreased, and research tied to such agreements and with obvious military uses has mostly come to a stop. However, on the informal level, when it comes to ad hoc collaboration between two or more scientists outside of the scope of a binding agreement between two organisations, collaboration on research with obvious military use continues. An example can be found here.
Indeed, a recent study of published research found that “There have been more than 8,000 joint publications and collaborations between researchers affiliated with UK organizations and researchers affiliated with People’s Liberation Army-linked research institutes (PLA-RI) on STEM topics since 2020.”
Regarding ATAS, the FCDO refused (read our request here) to divulge to UKCT the number of Chinese nationals applying for certification to study in the UK through the scheme, withholding the data in order to protect the UK’s “international relations”. The FCDO did provide overall figures, however (see below).
UKCT is also in possession of older ATAS data stretching back to 2016 (see below). This data distinguishes between “student applications” and “research applications”, whereas the data above combines the two.
Relating these figures to the HESA data presents some difficulties and no conclusions can be drawn with full confidence. The graph below illustrates STEM student numbers from China and other ATAS-requiring countries, and ATAS applicants. It suggests that ATAS was tightened up from 2021 and then began to rise commensurate with the growing number of applications from ATAS-requiring countries. The relation of this to applicants from China and geopolitics relating to China remains obscure because of the FCDO’s refusal to release data.
Graph: STEM students from China and from other ATAS-requiring countries, versus the overall number of ATAS applicants. Full notes on the above graph can be found here.
In addition, concerningly, the FCDO was unable to answer the following FOI query: “Finally, please provide overall figures for each year which distinguish between applicants requesting certification prior to obtaining a student visa, and applicants requesting certification when a student visa had already been awarded”. This was refused on the basis that “the ATAS IT system does not collect visa data”. This seems rather extraordinary.
UKCT also discovered, by examining the ATAS online form, that ATAS does not appear to collect applicant’s Chinese-character names, only their Pinyin or Latinised names. In other words, Wang Zhanjiang would be recorded simply as Wang Zhanjiang, with no record of 汪战江. This is another striking omission, because basic due diligence on an individual’s background ought to require their Chinese-character name. Repeated phone and email enquiries about this matter to the FCDO press office went unanswered.
Nor does the form, which is not specific to China, ask whether the applicant is a Chinese Communist Party (CCP) member. All members are formally obliged to “carry out the Party’s decisions, strictly observe Party discipline, guard Party secrets, be loyal to the Party, work hard, fight for communism throughout my life, be ready at all times to sacrifice my all for the Party and the people, and never betray the Party”.
These apparent omissions raise questions about the efficacy or purpose of ATAS.
In response to UKCT’s FOI, the FCDO squarely refused to state “how many FCDO staff administer the ATAS scheme, and how many have achieved C1 or HSK5 language level or above” in Chinese.
The limits of research security
Clearly, imposing research security requirements on the 150,000-odd STEM postgraduates – not to mention all academic staff – is a resource-intensive task. China-specific research security concerns are likely to be numerically preponderant and potentially challenging for linguistic and cultural reasons.
Whilst, of course, British or other nationals might and in some cases have generated a China-related research security issue, Chinese nationals may be more likely to be in a better position to profit from avoiding or evading research security measures or otherwise motivated to do so. Individuals of all nationalities bend or break rules, contravene norms and act in their self-interest. But, practically speaking, it may be easier for someone who speaks Mandarin to work with researchers in China, or take intellectual property [IP] to China and raise money for a start-up.
It is little wonder that universities feel challenged by research security requirements and that ad-hoc grassroots research collaboration between individuals at British institutions and Chinese military institutions persists, as do IP issues relating to China.
As the number of Chinese STEM postgraduates grows, especially at elite institutions, and as these individuals mature, advance in their careers, perhaps gain leave to remain or citizenship, and take up institutional leadership roles with control over resources and admissions, research security risks might evolve and multiply. By way of example, UKCT is aware of faculty leaders born in China but with British citizenship who have facilitated ATAS evasion and possible immigration offences with regard to Chinese students; we are aware also of concerns at some institutions about official material beginning to be bilingual or even Mandarin-only, about visiting staff who do not speak English invited by institutional leaders born in China, and about irregular appointment processes.
None of this is to imply that more stringent measures are desirable. UKCT makes no policy prescription. Indeed, the next section may be interpreted to suggest that those trying to devise policy approaches to support the public or national interest ought to think more creatively and constructively about how to address the concerns behind the ‘research security’ paradigm without resorting to ‘securitisation’ per se.
4. China’s STEM revolution
The broader context to the developments described here is the rise of Chinese science and the geopolitical ramifications of China’s rise in general. In recent years, a growing number of studies have indicated the dramatic increase in the quality and quantity of Chinese scientists and their work, especially at the applied edge, where STEM education feeds into invention, innovation, industry and capability.
Achievements such as those of DeepSeek or of China’s hypersonic glider programme are impossible to ignore. They manifest an important fact: Chinese science is not purely derivative and Chinese innovations are not purely ‘theft’. On the contrary, whether based in Western or Chinese institutions and companies, PRC-born scientists are increasingly dominant in a range of fields. For example, data from 2023 suggested that there are more PRC-born elite artificial intelligence (AI) scientists at American institutions than there are US-born.
These statistics are a reflection of a profound social change in China over the past forty years. Data compiled by the late German sociologist Gunnar Heinsohn – discussed at previous Ditchley Foundation meetings in the UK as parsed by physicist Steve Hsu – explores the educational roots of the explosion in China’s STEM human capital. Heinsohn uses information from the Programme for International Student Assessment (PISA), run by the Organisation for Economic Co-operation and Development (OECD).
Heinsohn’s analysis is now dated and could easily be refreshed because it is simple, combining PISA scores with population statistics. Heinsohn’s analysis of teenagers’ PISA scores suggests, as Hsu puts it, that “It is possible that by 2050 the highly able STEM workforce in PRC will be ~10x larger than in the US and comparable to or larger than the rest of the world combined.” Hsu was born in the USA and began his entrepreneurial career founding a business supported by the US Central Intelligence Agency’s (CIA) venture capital fund. The conclusion he draws from this is that the US will struggle to outcompete China on technology and science without continuing to absorb Chinese talent.
Source: https://drive.google.com/file/d/10yDtwq_aIecsjGnyS_qmUTR060ZdYrT9/view?usp=sharing N.B., Heinsohn and most other observers think the 165 value for China is inflated, hence the yellow highlighting and note at the bottom of the table.
Many excellent Chinese students and scientists go abroad to further their careers or study. Many then stay abroad and enjoy stellar careers in industry or academia in the West.
How many stay abroad? In 2010, a study by the US Department of Energy’s Oak Ridge Institute found that 92% of Chinese temporary resident doctorate recipients from 2002 were still in the US in 2007. However, more recent research by the Center on China’s Economy and Institutions at Stanford University found that the rate of Chinese academic scientists, both recent PhD recipients and more experienced staff, returning to China has increased by a factor of roughly 7 since 2010.
The Stanford study discussed various push and pull factors, including anti-China political sentiment in the US. It suggested that the biggest self-reported ‘push’ factor encouraging flight from the US was growing difficulty in recruiting international students. The biggest ‘pull’ factor tempting them to stay in the US was a “desire to contribute to U.S. leadership in science and technology”.
In China, quality of life has improved dramatically in the last few decades. In more developed areas, infrastructure and consumer products are excellent. Streets are clean and safe; indeed, one major ‘push’ factor encouraging Chinese scientists to leave the USA was reportedly feeling unsafe. Anecdotally (although we have found no solid objective studies), the streets of big (Tier 1) cities in China are in many cases freer from public theft and violent crime than are those of the US or the UK.
There will, of course, be many general ‘push’ and ‘pull’ factors for each country and they will not be discussed further here. UKCT could not identify a UK study akin to that produced by Stanford. Such a study would be very valuable.
Worth discussing further, however, is the role of United Front work and Chinese ‘talent programmes’ in the CCP’s treatment of Chinese STEM human capital. In the past twenty years, the CCP has gone to great lengths to ‘reverse the brain drain’. This includes well-known schemes such as the 1000 Talents Plan, but also a constellation of other schemes, offering lucrative incentives, resources, and access to returning scientists. Whilst often administered by a Chinese government body such as the Ministry of Science and Technology, such schemes are heavily coloured by and connected to the Western Returned Scholars Association (WRSA 欧美同学会).
The WRSA is controlled by the United Front Work Department (UFWD) of the CCP, although it has recently sought to obscure this. In fact, the importance of its work within the UFWD has grown in recent years and the current head of the UFWD is a ‘Western returned scholar’ himself. Li Ganjie 李干杰 took up his role in April 2025. He is reportedly a Tsinghua-educated nuclear engineer who later studied and then worked as a science diplomat in France.
The WRSA is older than the CCP itself. The WRSA was formed in 1913 by scientists but also humanities scholars and statesmen. It was later absorbed by the CCP and has been one of if not the most important bodies for networking, influencing, befriending and monitoring Chinese scientists overseas since they began to go abroad systematically in the 1980s. The WRSA has branches across China and deep connections to science diplomats and UFWD cadres operating as “Overseas Chinese affairs” staff in embassies worldwide, as well as to the network of Chinese Students and Scholars Associations (CSSAs) in universities around the world. There are roughly 100 CSSAs in the UK. They are connected to the Chinese embassy. A detailed study of them was published in 2023.
The WRSA’s remit involves running events, advertising talent schemes, talent spotting, promoting exchange between Western and Chinese institutions and, perhaps above all, honing the understanding, rhetoric, techniques and incentives that can help the CCP make the most of the dynamic of intensive scientific mobility and exchange between China and the West in the context of China’s STEM revolution. This is not limited to encouraging people to move. It also includes, for example, encouraging collaborative research primarily conducted over the Internet. As a UFWD body, the WRSA also has a soft intelligence and propaganda function and is likely to provide cover for intelligence operatives of the PLA or Ministry of State Security (MSS).
The WRSA and talent schemes are important and merit further research, but it is worth keeping them in perspective. General push and pull factors are likely to remain supreme and include (but are not limited to) job opportunities, public safety, politics, family proximity, and all the other elements of quality of life, doubtless down to weather and food for some.
5. Conclusion
A country’s success in developing and empowering scientists is a major contributing factor to its broader industrial and commercial success and the wellbeing of its people.
The statistics UKCT has published today reveal the deep intertwining of Britain’s STEM education system with that of the PRC. A very high percentage of elite postgraduate STEM students in the UK are PRC nationals.
This has attracted remarkably little by way of policy response, considering the PRC’s hostility towards the UK in cyberspace, well known risks in the potential transfer of technology and IP, and the cross-party consensus that the UK ought to ‘challenge and compete’ with the PRC in many areas.
UKCT does not seek to problematise the situation per se, but to situate it in the relevant context and provoke debate about how different aspects of it may be exploited or mitigated in a way that supports the public interest. In that spirit, three concluding remarks are made:
There is more or less universal agreement as to the importance of STEM for the UK’s prosperity and security in an increasingly dangerous world. Yet funding and investment are severely lacking, public demand for STEM higher education is not being met, and UK scores on school-level STEM education suggest a growing gulf with nations in East Asia (but with no concerted attempt to learn from East Asian education systems).*
Whilst research security spending and mechanisms continue to grow, there appear to be basic failures to bridge the ‘China information gap’, which is a fundamental driver of research security issues. For example, the Chinese character names of ATAS applicants are not collected by the FCDO. By way of another example, there is no evidence that the FCDO or universities collect information on CCP membership, even though it ought to be an important risk component and perhaps ought to appear on internal registers of interests. UKCT will be working on this issue in the future.
There is little evidence of positive, constructive measures to engage with, learn from or understand the lives of Chinese STEM practitioners in the UK in particular, despite the growing number and the potential of these people. No survey of attitudes to everyday, institutional and political matters has been undertaken. This neglect stands in sharp contrast with the CCP’s United Front work. Anecdotally, Chinese STEM students often live and work in relatively isolated groups, creating a social and cultural gap between them and the rest of their institutions which can only be an advantage for the CCP’s talent and United Front programmes. This is a missed opportunity to engage, learn and share, enhancing the wellbeing of Chinese students and addressing the ‘China information gap’. Universities, civil society organisations and the government could do more.
Anyone interested in these topics should contact info@ukctransparency.org. We would like to support meaningful research into them and ensure these issues receive appropriate attention from universities, funding bodies, civil society organisations and the government.
*It is interesting to note that during the Cold War, the US responded to the ‘Sputnik moment’ by passing the 1958 National Defense Education Act, providing funding for a whole variety of educational measures. The US National Science Foundation subsequently invested more than $100,000,000 ($1 billion in today’s terms) into new curricula. Within 15 years, half a million secondary school teachers in mathematics, science and modern languages had gone through new training programmes; systematic aptitude testing was introduced, with 500,000 students tested as part of a groundbreaking psychometric study, and ‘gifted talent’ programmes instituted for <1% of the brightest students.