The Race Debate Britain’s Universities Refuse to Finish

Britain’s argument over race in higher education has produced two incompatible diagnoses: Jason Arday says a racialised academic system keeps Black researchers at the margins, while Tony Sewell argues that class, family, geography and culture explain more of the country’s unequal outcomes than institutional racism. The data support parts of both accounts, but they do not justify the fashionable shortcut that turns every disparity into proof of discrimination.

The number that can support a serious accusation

The strongest case against the academic establishment is not a slogan about “whiteness”. It is a number. Between 2016/17 and 2018/19, UK Research and Innovation councils awarded 19,868 funded studentships. Only 245 went to Black or Black Mixed students, according to the Broken Pipeline report. Of those, just 30 went to students from Black Caribbean backgrounds.¹

Thirty awards in three years is not a trivial statistical wrinkle. It is the kind of result that demands an explanation from the institutions distributing the money. A doctoral studentship is not simply another qualification. It is the entrance ticket to the research career: the period in which an aspiring academic acquires a supervisor, develops a publication record, joins a professional network and becomes known to the people who later sit on appointment panels. If access to that stage is narrow, the professoriate will remain narrow years later.

Arday made this point in a 2021 lecture hosted by Oxford’s Department of Physics. He described the doctoral pipeline as a chasm between the large and increasingly diverse undergraduate population and the much smaller group that reaches postgraduate research. He also cited figures showing that Black academics accounted for less than 1 per cent of the UK professoriate, with about 90 Black men and 35 Black women among more than 20,000 professors. In the Russell Group salary comparison he presented, the average pay was £52,000 for a white academic, £38,000 for a Black academic and £32,000 for an academic from an Arab background.

These figures are disturbing. They are also incomplete. A disparity tells us that two groups have different outcomes. It does not tell us which mechanism produced the difference. The studentship figure does not, by itself, reveal the number of applicants in each category, their subjects, their grades, the universities they attended, the quality of their proposals, the distribution of supervisors, or the relative availability of funding in different disciplines. Nor does it show whether Black Caribbean applicants were being rejected at the final stage, applying in smaller numbers, choosing subjects with fewer funded places, or being filtered out years earlier by school and undergraduate attainment.

That distinction is not a technical escape route. It is the difference between evidence and interpretation. A weak pipeline may be caused by racial discrimination, class barriers, institutional culture, family circumstances, subject choice, migration patterns, or a combination of them. The existence of a severe outcome is a reason to investigate the process. It is not permission to declare the process understood.

Arday’s lecture was at its most persuasive when it stayed close to the human consequences. He spoke of researchers who were “not thriving, they’re surviving”, of equality work assigned to minority academics without proper pay or promotion credit, and of networks that reward people who already possess access to grants and senior colleagues. Those are plausible institutional mechanisms. They deserve examination rather than mockery. But the lecture moved too quickly from testimony and disparity to a single explanation. That is where a case for reform became a doctrine about causation.

The table becomes more revealing when the categories are separated

The additional comparative table supplied for this analysis is useful because it places detailed ethnic groups beside two aggregates: Black people as a whole and the broader BAME category. Its arithmetic is simple. The Black total is the sum of Black African, Black Caribbean and Black Other. The BAME total is the sum of Asian, Black, Mixed and Other. In the population column, 9.3 per cent Asian, 4.0 per cent Black, 2.9 per cent Mixed and 2.1 per cent Other produce an 18.3 per cent non-White aggregate. The Office for National Statistics records White residents as 81.7 per cent of the population of England and Wales in the 2021 Census, with Asian residents at 9.3 per cent and the Black group at about 4 per cent.²

Adding the aggregate lines does not create new evidence. It creates a check on the way evidence is being framed. When the detailed Black categories and the Black total appear together, the reader can see that the overall Black figure is dominated by the African subgroup. The Black Caribbean line is much smaller. The same relationship appears in the supplied higher-education estimates: Black students are described as making up roughly 8 to 9 per cent of UK-domiciled undergraduates, but most of that total is attributed to Black African students, while Black Caribbean students form a substantially smaller share.

That matters because a broad category can make an apparently uniform story look more convincing than it is. “Black students” may be an appropriate administrative heading for some purposes, but it is a poor explanation of educational outcomes if African and Caribbean pupils experience sharply different results. “BAME” is even less satisfactory. It joins groups with different migration histories, family structures, religions, locations, income profiles, school experiences and subject choices. An institution can advertise a high BAME percentage while leaving the specific group under discussion almost invisible.

There is a second problem in the table, and it is more fundamental. It compares population shares from England and Wales with higher-education figures for UK-domiciled students. The geographic denominators are not identical. Nor are the age structures. The population share includes children, working-age adults and pensioners; the undergraduate population is concentrated within a much narrower age range. Some ethnic groups have younger age profiles than others. A group with many young people can be expected to form a larger share of university students than of the total population even before attainment, selection or discrimination enters the picture.

The table is therefore best read as a map of stage-specific representation, not as a courtroom verdict. It raises questions about where groups are concentrated and where they disappear. It does not settle whether the cause is institutional racism, class, family life, subject choice, age structure or admissions practice. Aggregate rows are valuable when they expose the contribution of each component. They become misleading when they are treated as self-explanatory.

The African result that unsettles a simple racism narrative

The sharpest challenge to a single discrimination story comes from the contrast between Black African and Black Caribbean outcomes. The two groups share the broad racial classification used in national tables. They do not share the same educational profile.

The Commission on Race and Ethnic Disparities reported that children from many ethnic communities performed as well as or better than White pupils in compulsory education, while identifying Black Caribbean pupils as a major exception. Later parliamentary analysis has also noted that Black African pupils generally fare better than Black Caribbean pupils, and that Black pupils eligible for free school meals can outperform the national average for pupils receiving that support on some GCSE measures.³

The figures in the supplied table point in the same direction. Black African students are presented as the force driving the overall Black science participation rate, at about 51 per cent, while Black Caribbean participation is lower in some science and technology fields. The table also describes Black African students as the bulk of Black postgraduate-research students and Black Caribbean students as under-represented within the Black postgraduate total.

None of this proves that Black Caribbean students do not encounter discrimination. It proves something narrower and more important: skin colour alone cannot carry the full explanatory burden. If the same racial classification produces markedly different results, a serious account must identify the additional variables. It must ask about migration selection, family expectations, parental education, neighbourhoods, school composition, cultural attitudes towards academic achievement, religious communities, gender patterns and the timing of settlement in Britain.

Some of those questions make people uncomfortable because they sound like an invitation to blame families or communities. They are not. Explanation and blame are different activities. A policy that identifies family structure or school culture as relevant is not declaring a child responsible for the circumstances into which he was born. It is trying to find an intervention that might work.

The West African comparison is especially awkward for fashionable accounts of institutional power. A framework built around the claim that “whiteness” is the organising centre of excellence has to explain why Black African pupils can outperform White pupils, including disadvantaged White pupils, in important measures while Black Caribbean pupils face a much steeper educational struggle. It may respond that racism operates differently across groups, that stereotypes vary, or that institutions reward one form of minority success while penalising another. Those propositions are possible. But they require evidence. They cannot be smuggled in as automatic explanations after the headline theory has already been chosen.

Sewell’s approach has an advantage here. It begins with heterogeneity. The commission rejected the idea that ethnic minorities form a single social bloc and recommended moving away from the catch-all term BAME. That is not a complete theory of educational inequality, but it is a better starting point than an aggregate category whose internal differences have been erased.

Sewell’s axioms are unfashionable because they are testable

Tony Sewell’s report became a political and academic lightning rod because it challenged the proposition that institutional racism is the principal explanation for racial disparities in Britain. Its central sentence was blunt: “We found that most of the disparities we examined, which some attribute to racial discrimination, often do not have their origins in racism.”⁴

The report did not say that racism had vanished, that discrimination was imaginary or that every institution was fair. It said that disparity and discrimination are different things. That distinction should be uncontroversial. It has become controversial because public debate often treats the gap between two groups as though it contains its own explanation.

Sewell’s framework can be reduced to several propositions. Class disadvantage is widespread and affects people of every ethnicity. Because Black Britons are disproportionately represented in poorer communities, measures that improve working-class life will help many Black citizens without using racial preferences. Family structure, educational aspiration, geography and cultural expectations influence outcomes. A person’s experience of offence or exclusion is real, but the experience alone does not identify the cause. Evidence must do that work.

These are not axioms because they are beyond question. They are useful because they can be questioned. If class-first interventions failed to narrow racial gaps among comparable groups, the framework would need revision. If two applicants with similar qualifications and social backgrounds received sharply different outcomes after their ethnicity was taken into account, the case for racial discrimination would strengthen. If a university’s selection panels repeatedly produced disparities that could not be explained by subject, prior attainment or socioeconomic factors, the institution would have a case to answer.

The commission also offered policies rather than a simple denial. Its recommendations included extended school days in disadvantaged areas, better mentoring, an Office for Health Disparities and more precise use of ethnic categories. The underlying ambition was to attack the circumstances that limit opportunity instead of promising that a change in vocabulary would transform them.

Sewell’s most politically useful insight is that the people who need attention are not always the people highlighted by the prevailing story. White working-class boys have persistently poor educational outcomes, and the House of Commons Education Committee described their underachievement as a long-standing failure of policy. The committee said White British pupils eligible for free school meals persistently underperform compared with peers from other ethnic groups, while also identifying poverty, family circumstances, prior attainment and aspirations as relevant factors.⁵

There is no moral reason to force a choice between helping White working-class boys and helping Black Caribbean pupils. Both groups can be badly served by weak schools, unstable homes, poor local labour markets and institutions that confuse public relations with reform. A serious policy would ask which intervention improves outcomes for each group. A symbolic race framework asks which group is permitted to describe itself as structurally oppressed.

Class is the missing variable in the Oxford lecture

Arday’s argument about the academic pipeline is strongest when it identifies exclusion and weakest when it refuses to show how exclusion interacts with class. In a lecture about who wins doctoral funding and reaches professorial rank, he presented racial categories in detail but did not provide a comparable account of the class backgrounds of applicants and award holders.

That omission changes the question. Suppose Black Caribbean students are under-represented among funded PhD candidates. The relevant comparison is not simply Black Caribbean versus White. It is Black Caribbean applicants against White applicants with similar prior attainment, comparable degrees, similar household resources, similar school histories and similar access to academic networks. If the Black Caribbean gap survives those controls, the racial explanation becomes stronger. If the gap shrinks sharply, class and educational opportunity have been doing more work than the headline category suggested.

Class barriers in academia are not imaginary or confined to one ethnic group. Geraldine Van Bueren, who helped establish an association for working-class academics, said it remained difficult to enter academia from a working-class background because of the cost of obtaining a doctorate. She also identified accent prejudice, limited money for conferences, derogatory comments and the lack of a financial cushion as barriers that could persist into mid-career.⁶

Doctoral funding can remove tuition costs, but it does not remove every cost of entry. A student may need to move to an expensive city, accept years of modest income, attend unpaid networking events, undertake teaching, travel to conferences or maintain a research profile after the doctorate ends. A young applicant with professional parents may treat these demands as an investment. A student supporting relatives or living far from the main research centres may see the same career as an economic gamble.

The academy also rewards familiarity. The right supervisor, the right reference, an early conference paper and an informal introduction can matter at the margins between candidates. Those advantages are not reducible to race. They may follow parental education, private schooling, elite universities, metropolitan residence or the ability to work without immediate financial pressure.

Arday’s account could have been strengthened by showing these cross-tabulations. Instead, class appears mainly as an additional dimension of oppression within an intersectional framework. That language can accommodate any result, but it does not tell policymakers which variable to attack first. A theory that names every disadvantage without ranking their causal importance may be morally expansive and practically useless.

Sewell’s class-first axiom therefore deserves more than dismissal. It is not the claim that race never matters. It claims that the class variable is too important to omit. That is a modest demand, yet some academic settings treat asking for evidence as an act of hostility.

Disparity is a signal, not a verdict

The argument over race in higher education repeatedly collapses three different statements into one. The first is descriptive: group A has a different outcome from group B. The second is experiential: members of group A report unfair treatment. The third is causal: the institution discriminated against group A. All three statements may be true at once, but none logically proves the others. This is like saying correlation is not causation!

Arday’s evidence contains examples of the first two. The studentship figures describe a disparity. Interviews with hundreds of researchers describe experiences of exclusion, unpaid EDI labour and closed networks. Those accounts should be investigated. But the move from “this happened” to “the colour of their skin is the only narrative that matters” is not an empirical conclusion. It is a decision about which explanations may enter the room.

The pay figures raise the same problem. An average salary gap between ethnic groups may reflect discrimination, but it may also reflect academic rank, discipline, contract type, length of service, institution, migration status, part-time work, career breaks or the distribution of subjects. A properly designed analysis can separate these possibilities. A headline gap cannot.

This is why the phrase “the race card” is both useful and dangerous. It is useful when it identifies the habit of attributing every negative outcome to racism before considering alternatives. It is dangerous when it is used to ridicule genuine complaints or to shut down investigation. The answer is not to believe every accusation or to reject every accusation. The answer is to establish a standard of evidence that applies to everyone.

Arday asked why the testimony of a person of colour is “neutralised or challenged” when that person says he or she has experienced racism. There is a legitimate moral point here. Institutions can use demands for proof as a delaying tactic, forcing complainants to relive an incident while protecting the organisation from scrutiny. A system that never records patterns because each event is treated as isolated will miss discrimination that operates through accumulation.

But a right to be heard is not a right to have one’s interpretation accepted without question. The same principle must protect a White working-class pupil who reports bullying, humiliation or exclusion. His account should not be discounted because a broader political narrative assigns him the status of beneficiary. Nor should a Black student’s account be dismissed because other Black students have succeeded. Individual testimony is evidence of experience. It becomes evidence of institutional cause only when it is connected to pattern, process and comparison.

The test is impartiality. If one group’s pain is presumed to reveal a system while another group’s pain is presumed to reveal personal weakness, the language of equality has become a hierarchy of credibility. That is not justice. It is a new form of group essentialism, dressed in progressive vocabulary.

The admissions gate remains a contested part of the story

The debate cannot stop at doctoral funding. The pipeline may be breaking long before a student writes a PhD proposal. Access to selective universities, the choice of subject, degree classification and the decision to continue into research all shape who later appears in the professoriate.

Research by Vikki Boliver found that young people from Black British backgrounds were more likely to attend university than their White British peers but less likely to attend the most selective universities. Her analysis of earlier UCAS data found that ethnic differences in Russell Group offer rates remained after taking account of A-level grades, facilitating subjects and degree subject area. She also recorded the counter-argument from universities and UCAS: minority applicants were more likely to choose oversubscribed courses and may have lacked the precise qualifications required for particular programmes.⁷

That is precisely the kind of dispute that requires transparent data rather than moral certainty. If subject choice explains most of the gap, the intervention should begin with advice, preparation and a wider range of applications. If a gap remains among applicants with the same grades, subjects and contextual background, admissions panels need to explain it. If the evidence changes by university or course, a national theory of institutional racism is too crude.

The same logic applies to the transition from undergraduate study to research. Black students may be highly represented among undergraduates but concentrated in subjects with fewer doctoral opportunities. They may achieve different degree classifications even with similar entry grades. They may have less access to supervisors who encourage them to apply for research. They may face financial pressures that make a doctorate unattractive. Each mechanism requires a different remedy.

The supplied table’s stage-by-stage comparison captures this movement. Minority-ethnic students are estimated to make up roughly 28 to 30 per cent of UK-domiciled undergraduates against an 18.3 per cent share of the England-and-Wales population, while their postgraduate-research share is estimated at about 18 to 20 per cent. White students move in the opposite direction: roughly 70 to 72 per cent of undergraduates, rising to about 74 per cent of postgraduate researchers against an 81.7 per cent population share.

Those figures do not show a universal pattern of White privilege and minority exclusion. They show that representation changes by educational stage. Asian students are described as over-represented at undergraduate level and particularly concentrated in science subjects, yet closer to their population share at postgraduate-research level. Black African participation pushes the overall Black undergraduate figure upwards, while Black Caribbean representation remains weaker at later stages. Disadvantaged students from the most deprived areas fall away most sharply at research level.

That is a complicated pipeline. Anyone who describes it with one word has chosen rhetoric over diagnosis.

What affirmative action would reward

Arday was clear about the policy that followed from his analysis. Asked whether the United Kingdom should adopt something closer to American affirmative action, he said it was the one intervention he would bring into Britain. He argued that positive action already existed in law but was used weakly, and that change might need to be forced because institutions had made only “fragile gains”.

There is a serious argument for targeted intervention. If a university’s recruitment process repeatedly produces an extreme disparity that cannot be explained by applications, attainment, discipline or background, it should not hide behind neutral language. Ring-fenced doctoral funding, paid mentoring, transparent shortlisting and properly rewarded equality work may change the pipeline. The Broken Pipeline report itself says its findings helped encourage targeted funding and contributed to wider efforts by UKRI and Research England to broaden participation.¹

The question is whether the intervention should be racial, class-based or both. A race-only preference can help the most advantaged members of a targeted group. A Nigerian-heritage applicant from a professional family, an academically selective school and a prosperous London borough may receive the benefit intended for a Black student facing severe structural disadvantage. A White applicant from a poor former industrial town may be treated as the embodiment of privilege despite lacking money, networks or educational security.

This is not an argument for ignoring race. It is an argument for refusing to confuse race with disadvantage. A serious scheme could consider household income, school context, neighbourhood deprivation, parental education, caring responsibilities and disability alongside ethnicity. It could identify where a group-specific disparity persists after those factors are accounted for. It could publish the results so that the public can see whether the policy is correcting a measured barrier or merely satisfying an institutional preference.

Affirmative action also raises a moral question that its advocates often avoid. The person displaced by a group preference is not an abstraction. He or she may have spent years preparing, achieved the required grades and done nothing wrong. If a university rejects that applicant because the institution wants a different demographic profile, it owes the public a clear account of why that collective goal outranks the individual claim.

There are cases in which a society may accept such a trade-off. But the burden of proof rises when the policy is coercive. It is not enough to say that diversity is beneficial, that institutions feel more legitimate when they look different, or that students learn from one another. Those propositions may be true in some settings and false in others. The university must show what problem it is solving and how the selected remedy addresses it.

The strongest case for targeted funding is therefore narrower than the strongest case made in the lecture. It is a case for finding the exact point at which candidates disappear and correcting that process. It is not a blank cheque for treating racial identity as a substitute for evidence.

The danger of letting aggregate success erase specific failure

Critics of the Sewell framework are right about one danger: aggregate averages can conceal severe problems inside smaller groups. If Black African success raises the overall Black average, the Black Caribbean shortfall may become harder to see. If Asian students’ strong undergraduate participation improves the BAME total, the poorer outcomes of some other Asian groups may disappear into the same line. Aggregation can soften the very disparities that policy should confront.

Arday was right to criticize BAME as a catch-all label. The broad term can allow a university to report progress while avoiding the question of who benefits. A department that has a large Asian international student population may be able to describe itself as diverse even if its Black British academic staff remain almost absent. A table that counts all non-White students together can create a flattering institutional story without revealing the distribution of opportunity.

But disaggregation cuts in more than one direction. It does not merely reveal hidden disadvantage. It also reveals hidden success. That is why the Black African result matters. A narrow focus on the lowest-performing subgroup can create the impression that every member of the broader group faces the same barriers. A fair analysis must be prepared to show both the failure and the exception.

Disaggregation also creates a practical problem: small numbers become unstable. A few additional admissions or awards can move the percentage sharply for a group with a small population. Institutions must publish enough detail to expose patterns while protecting individual privacy and avoiding misleading year-to-year comparisons. They should also distinguish between applicants, offers, enrolments, completions and academic employment. Each is a different stage of the process.

The table’s aggregate lines are most useful when they function as accounting. The Black row should equal its subgroups. The BAME row should equal Asian, Black, Mixed and Other. If the totals do not reconcile, the category is being used as a political label rather than a statistical measure. Once the totals reconcile, the real questions begin: which subgroup drives the result, at which stage does representation fall, and what denominator is appropriate?

That discipline would improve the argument on both sides. It would prevent institutions from hiding Black Caribbean exclusion behind a high BAME percentage. It would also prevent campaigners from treating Black African success as irrelevant merely because it complicates the theory. Data should not be sorted into “useful” and “unhelpful” facts. It should be allowed to contradict the preferred narrative.

A debate about suffering cannot become a competition for immunity

The emotional force of race arguments often comes from lived experience. Arday has interviewed hundreds of researchers and describes an academic culture in which Black and minority ethnic staff carry burdens that are not recognized in promotion or pay. That testimony should change how universities investigate complaints. It should make managers ask whether the same people are repeatedly assigned equality work, whether their research is valued, whether they are excluded from grant networks and whether promotion rules reward the labour institutions publicly celebrate.

Yet lived experience is not a monopoly held by one racial category. White working-class students can be poor, mocked, isolated, assaulted, bullied and excluded. They can be told that their background makes them socially suspect or politically inconvenient. A young person who is struggling is not comforted by being informed that his group occupies a privileged position in the national imagination. That label may describe a statistical average while completely misdescribing his individual life.

The danger is greatest when a framework turns disagreement into evidence of oppression. If a White academic questions a racial explanation, the question may be treated as proof of “whiteness”. If a Black academic supports Sewell’s class-first analysis, he may be accused of serving a hostile agenda. If a student disputes a diversity policy, the objection can be recoded as fear of losing privilege. A theory that makes every counter-argument incriminating has protected itself from revision.

That is not how serious journalism, science, or public policy works. A claim must be capable of being wrong. Its advocates must say what evidence would weaken it. The burden is not satisfied by citing a person’s identity, however relevant identity may be to the experience being described.

The same standard should restrain Arday’s opponents. It is lazy to dismiss his analysis as mere ideology because he uses the language of structural racism. The studentship figures demand an institutional response. So do the accounts of unpaid equality work and the low number of Black professors. The appropriate criticism is not that he notices injustice. It is that he often treats one interpretation of injustice as the only permissible one.

Britain needs a politics that can hold two thoughts at once: racism exists, and not every disparity is caused by racism. Class matters, and class does not explain every racialized experience. Some Black groups succeed in parts of the education system, and another Black group may still face a distinct barrier. An institution can be broadly open while preserving a closed network at one particular stage. Complexity is not evasion. It is the price of accuracy.

The test Britain’s universities should now face

The most useful result of the comparison between Arday and Sewell is not a verdict on which man is morally superior. It is a test for the institutions that have spent years publishing statements about inclusion without publishing enough evidence to identify failure.

Universities should report the full progression chain: applications, offers, enrolments, degree results, doctoral applications, studentship awards, completions, postdoctoral appointments, lectureships, promotions and professorial appointments. They should break down every stage by detailed ethnicity, not only BAME, and by socioeconomic background. The figures should be separated by subject, institution, domicile and gender. A single national average cannot show where the mechanism operates.

They should also publish the denominators. The public needs to know not only how many Black Caribbean students received awards, but how many applied, in which fields, with what prior qualifications and against how many available places. It needs to know whether a pay gap remains after rank, contract, discipline, experience and working hours are taken into account. It needs to know whether a Black African undergraduate advantage survives into doctoral study, and if it does not, exactly where the loss occurs.

Socioeconomic data must be treated as a central variable, not an optional supplement. Free-school-meal eligibility is not a complete measure of class. Neither is the postcode of a student’s home. But imperfect measures can still be improved through parental education, school type, neighborhood deprivation, household income, and care responsibilities. If institutions demand sophisticated racial analysis while accepting crude class data, they are not pursuing equality. They are choosing the variable that produces the most politically useful story.

The language should become more exact. “BAME” may be convenient for a headline, but it is too broad for diagnosis. “Under-represented” must state the reference population and educational stage. “Discrimination” must identify the process in which it occurred. “Diversity” must describe a measurable objective rather than serve as an all-purpose moral seal.

The policy should follow the diagnosis. Where poverty blocks entry, fund maintenance and reduce the cost of doctoral study. Where school attainment is the barrier, improve teaching and early intervention. Where subject choice narrows opportunity, provide better information and preparation. Where panels produce unexplained ethnic gaps, audit the panels and change the process. Where equality labour is routinely assigned to minority staff without reward, pay for it and count it in promotion.

And where a race-conscious intervention is proposed, its architects should answer the question the Oxford lecture avoided: why race rather than class, or why race as well as class? They should show who benefits, who bears the cost, what evidence triggered the policy and what result would justify ending it.

The figures do not permit the comfortable conclusion that Britain’s universities are free of racism. Nor do they permit the fashionable conclusion that racial identity explains the academic hierarchy from undergraduate study to the professoriate. They show a broken and uneven pipeline, with different groups falling away at different stages. They show that Black African success complicates a uniform Black experience, that Black Caribbean under-representation warrants targeted scrutiny, that White working-class disadvantage is real, and that postgraduate research remains more socially selective than the undergraduate sector.

That is enough to demand reform. It is not enough to suspend judgment. The first duty of a university is not to repeat the correct vocabulary. It is to discover where opportunity is being lost, establish why, and repair that point without replacing one arbitrary hierarchy with another.

Sources: ONS Census 2021; HESA 2024/25 ethnicity tables (Black African ≈ 155 000, Caribbean ≈ 31 500); earlier HESA science-subject percentages; OfS/House of Commons Library IMD and progression data. BAME is the conventional sum of the non-White groups listed.White students form 81.7 % of the England-and-Wales population yet only about 70–72 % of UK-domiciled undergraduates. That is under-representation at the undergraduate stage. At postgraduate research, their share rises to roughly 74 %, still below the population proportion but substantially closer. Within the White group, students from the most-deprived IMD quintile are markedly under-represented in PGR (typically 9–11 % at high-tariff providers), confirming that socioeconomic disadvantage continues to limit progression to research degrees even for the majority ethnic category.Asian students, by contrast, are over-represented among undergraduates (12–13% versus a 9.3% population share) and show a strong concentration in science subjects (54–55% of their own group). Their PGR share, however, is close to the population proportion at around 9%. The over-representation is therefore concentrated at undergraduate level and in STEM fields; it largely disappears by the research stage.The key figure in this debate is the value for Black (all) X approx. share of UK-domiciled PGR 4% to 5%: with Population share 4%!Taken together, the figures indicate that White students as a whole are under-represented relative to population share at undergraduate level and still modestly so at PGR, while the most-deprived White (and other) students face a sharper drop-off into postgraduate research. Asian students are over-represented at undergraduate level, especially in science, yet return closer to population share among PGR students. The patterns are stage-specific rather than absolute.

References

1. The Broken Pipeline: Barriers to Black PhD Students Accessing Research Council Funding, Leading Routes. https://www.leadingroutes.org/the-broken-pipeline-report ↩ ↩
2. Office for National Statistics, Ethnic group, England and Wales: Census 2021. https://www.ons.gov.uk/peoplepopulationandcommunity/culturalidentity/ethnicity/bulletins/ethnicgroupenglandandwales/census2021 ↩
3. House of Commons Library, Educational outcomes of Black pupils and students. https://commonslibrary.parliament.uk/research-briefings/cbp-9023 ↩
4. Commission on Race and Ethnic Disparities, The report of the Commission on Race and Ethnic Disparities, GOV.UK. https://www.gov.uk/government/publications/the-report-of-the-commission-on-race-and-ethnic-disparities ↩
5. House of Commons Education Committee, The forgotten: how White working-class pupils have been let down, and how to change it. https://committees.parliament.uk/committee/203/education-committee/news/156024/forgotten-white-workingclass-pupils-let-down-by-decades-of-neglect-mps-say ↩
6. Jack Grove, “Working-class academics ‘face discrimination throughout careers’”, Times Higher Education, 25 March 2019. https://www.timeshighereducation.com/news/working-class-academics-face-discrimination-throughout-careers ↩
7. Vikki Boliver, “Hard Evidence: why aren’t there more black British students at elite universities?”, The Conversation, 9 April 2014. https://theconversation.com/hard-evidence-why-arent-there-more-black-british-students-at-elite-universities-25413 ↩