The Quiet Clustering: Shared Origins and the Narrowing Door in Hong Kong's Doctoral Pipeline

Anyone who has spent time inside a Hong Kong research department will have noticed the pattern before they can name it. A large share of the supervisors were trained in, or come from, mainland China, and a large share of the doctoral students they take on come from the same place. This is not a rumour and it is not an accusation about any one person. It is visible in the public enrolment and staffing figures, and it quietly shapes the odds for every prospective doctoral candidate weighing an application from outside the mainland.

The point of naming it is not to assign blame. Most of the pattern is the product of forces no individual supervisor set in motion. But a student deciding where to apply, and a department thinking about the health of its research culture, are both better served by seeing the pattern clearly than by pretending it is invisible.

What the numbers show

The University Grants Committee, which funds and reports on Hong Kong's eight public universities, records academic staff in three origin categories: local, mainland China, and the rest of the world. In recent years the balance between them has shifted decisively, and the shift is sharpest at exactly the level that matters most for doctoral candidates, which is who is available to supervise them.

The composition

Who staffs and studies in the system

Figures drawn from University Grants Committee statistics on Hong Kong's eight public universities and from published research on its research postgraduate population.

41% Share of academic staff drawn from mainland China in 2024–25, the largest of the three official origin groups. Local scholars and those from the rest of the world made up roughly 29 per cent each.
6 of 8 Public universities where mainland scholars now outnumber local ones. This is a recent reversal; for most of the period since the 1997 handover, local academics were the largest group.
~84% Share of non-local research postgraduate students who came from mainland China, within a doctoral and MPhil population that was roughly four-fifths non-local. Students from beyond Asia made up a single-digit share.
Since 2020 The period over which the mainland share of international doctoral cohorts has accelerated, while the proportional share of every other region has declined.

Numbers alone do not establish intent. What they establish is a supervision market in which one large, well-connected pool sits on both sides of the table.

The pattern has a name

In the research literature on hiring and mentoring, the tendency of people to select others who resemble themselves is called homophily, and in academic hiring specifically it is sometimes called cloning bias. It is one of the most consistently observed features of how academic teams form. Search committees privilege a hard-to-define sense of fit, and fit tends to run along lines of shared background, shared training, shared networks, and shared language. None of this requires anyone to act in bad faith. It is what happens by default when a system leaves selection to informal judgement rather than structured criteria.

Doctoral supervision sits at the sharp end of this. A supervisor choosing a student, and a student choosing a supervisor, are both leaning heavily on informal signals: a recommendation from a known colleague, a shared academic lineage, a first meeting conducted comfortably in a shared first language. When most supervisors and most funded applicants are drawn from the same large pool, those informal channels route candidates toward one another with very little friction, and away from candidates outside the pool with equally little.

What it means for students from other countries

The effect on doctoral applicants from outside the mainland is rarely a closed door. It is a narrower one, and a set of quieter disadvantages that accumulate rather than announce themselves.

The downstream effect

How the door narrows

None of these is a formal barrier. They are the informal costs that a clustered pipeline imposes on candidates outside the dominant pool.

1 Fewer entry points When most supervisors and most funded applicants come from one well-networked pool, the informal channels that match students to supervisors favour candidates already inside that network.
2 Language as a sorting line Where a lab's working language drifts toward Putonghua in meetings and informal exchange, students without it are quietly disadvantaged even inside nominally English-medium programmes.
3 The internationalisation gap Official policy frames these universities as international bridges, yet the diversity of the doctoral intake is moving the other way, thinning the cross-cultural research environment the policy is meant to protect.
4 Structure misread as personal failure Applicants who struggle to secure supervision often attribute it to their own profile, when much of what they are meeting is a structural pattern in how the pipeline is organised.

A narrower door is still a door. The point is to read the odds accurately, not to be discouraged by them.

The fairer reading of the causes

It would be a mistake to treat this pattern as evidence of a deliberate policy of exclusion, and the more careful reading of the causes is also the more accurate one. A great many of the mainland-origin scholars now staffing Hong Kong's universities were educated and built their early careers overseas, and were recruited for exactly the internationally competitive profiles the universities were seeking. The shift in the staffing balance owes a good deal to an emigration wave among local residents, to a wave of retirements among an older generation of faculty, and to a period in which researchers of Chinese origin found the environment in some Western countries less welcoming and looked closer to home. On the student side, the mainland is simply the largest and nearest source of high-achieving applicants, and Hong Kong's shorter, English-medium doctoral routes are a natural draw.

In other words, the clustering is mostly an emergent outcome of demography, geopolitics, and ordinary human preference, rather than a designed one. That is precisely why it is so persistent, and why it will not be undone by good intentions alone. Systems that run on informal fit reproduce their own composition unless something deliberate interrupts them.

Naming it is the first move

For a prospective doctoral candidate from outside the mainland, the practical takeaway is not to avoid Hong Kong, which remains one of the strongest research environments in the region. It is to apply with clear eyes: to look for supervisors and groups with a track record of taking on students from varied backgrounds, to ask directly about the working language of the lab, and to read a slow or difficult search as information about the system rather than a verdict on the self. For departments, the takeaway is that a research culture genuinely committed to internationalisation has to build the structured, deliberate selection practices that informal fit will never produce on its own. The pattern is quiet, and it is mostly no one's fault. Naming it accurately is the first move toward deciding what, if anything, to do about it.

Sources & further reading Staffing and enrolment figures are drawn from University Grants Committee statistics as reported by the South China Morning Post and Times Higher Education (2024–2026), and from a 2026 document analysis of the origin distribution of academic staff across the eight public universities published in Globalisation, Societies and Education. Research postgraduate composition figures draw on Jung's study of local and non-local doctoral students in Hong Kong. The concepts of homophily and cloning bias in academic hiring are discussed in the higher-education literature on faculty selection. Readers are encouraged to consult the primary University Grants Committee statistics directly, as the balance shifts year on year.