The class problem in Chemistry won’t get fixed via APPs

That is the big question lurking underneath Beyond access: how socioeconomic background shapes the chemistry pipeline, a cracking new report from the Royal Society of Chemistry (RSC) published to coincide with Social Mobility Day.

It combines an analysis of five years of Higher Education Statistics Agency (HESA) data with a survey of 602 undergraduate chemistry students, 6 focus groups and 15 in-depth interviews – and it’s very much a proof of concept for doing access and participation work at subject rather than institution level.

The class problem at the top

The headline numbers are not where you might expect them. At the lower end of the socioeconomic distribution, chemistry looks almost indistinguishable from the all-subjects picture – the differences are a point or two at most.

The distinctive pattern is at the top, where chemistry recruits 39 per cent of its undergraduates from higher socioeconomic backgrounds (measured by parental occupation at age 14), against 34 per cent across all subjects and a UK working population benchmark of 37 per cent.

In other words, chemistry’s class problem is over-representation of advantage, not (statistically, at least) under-representation of disadvantage. That is a subtler story than the usual framing, and it is one that institution-level data can never tell – a university can hit every access benchmark it has while individual departments quietly remain finishing schools, because the institutional aggregate launders the subject-level stratification.

Aspiration was never the problem

The report’s central argument is a deliberate inversion of deficit framing. On average 81 per cent of undergraduates intend to pursue a career in the chemical sciences, and students from lower socioeconomic backgrounds are at least as likely (sometimes more likely) to say so than their better-off peers. What differs is not ambition but the capacity to convert it.

The mechanism, as in yesterday’s Student Academic Experience Survey, is risk. Money functions here less as consumption and more as insurance – it is what allows a student to take an unpaid placement, relocate to where the internships are, attempt a PhD on an inadequate stipend, or spend a summer accumulating career capital rather than wages.

Students without that insurance make rational, risk-averse choices – industry over academia, proximity over prestige, employability over enthusiasm – that look like preference but are actually constraint. The report notes that high career intent among less advantaged students may partly reflect limited alternatives rather than vocation. Persistence as risk management, not passion.

Positive ratings of overall quality of life fall from 81 per cent among students from higher socioeconomic backgrounds to 22 per cent among those from lower ones. Only 17 per cent of lower socioeconomic background undergraduates in the sample attend Russell Group universities. 44 per cent had no parental involvement at all in their university decisions, against 6 per cent of their more advantaged peers – the hidden curriculum in a single statistic.

Nobody thinks they are the posh one

My favourite finding concerns perception. Asked to describe their own socioeconomic background, 76 per cent of students whose parental occupation places them in the highest group described themselves as intermediate or lower – and not a single measured lower-background student described themselves as higher. Perception compresses towards the middle, and privilege is invisible to those who hold it.

It’s a fascinating finding about identity. It is a more worrying one about measurement, because the report goes on to recommend that universities incorporate self-perception into their data collection – a measure it also demonstrates is systematically biased towards understating advantage.

And there is an intersectional story too. Disability carries a net negative impact score of minus 55 per cent on perceived educational and career progression – dwarfing gender at minus 9 per cent and every other characteristic measured. Among disabled students from intermediate and lower backgrounds, 68 per cent report a negative impact, against 9 per cent of disabled students from higher backgrounds.

Basically, money substantially buys off the disadvantage of disability – paying for adaptations, support and time that the system does not provide. In a report about class, the flashing red signal in the data is what class does to everything else.

Friends have benefits

An especially useful finding is about peers. The report describes peer relationships as “compensatory social capital” for students from lower socioeconomic backgrounds – the networks that substitute for the family and professional connections others arrive with. That is about as strong a justification as a Royal Society is ever going to hand you for treating academic societies and subject level student associations as academic and progression infrastructure rather than nice-to-haves, and for scrutinising every participation cost attached to them.

The evidence on term-time work points the same way. Students describe paid employment crowding out placements, extracurriculars and study itself – one respondent’s grades suffered because she worked through first year to feed herself, which then locked her out of the placement schemes that required those grades.

On the lobbying side, the report supplies three asks with a learned society’s name attached. Universities should make belonging, inclusion and wellbeing explicit key performance indicators alongside attainment. Unpaid placements and internships should be replaced with paid schemes. And free school meals eligibility should stop being treated as a sufficient measure of socioeconomic background.

In England, all three map neatly onto access and participation plan territory – the plans providers agree with the Office for Students (OfS, England’s higher education regulator) setting out how they will improve equality of opportunity. Curiously, the report itself never mentions OfS or access and participation plans, which may well tell another story for another day about the lack of connection between these two types of regulator.

Students are missing

There’s an annoying absence. The report’s central empirical claim is that students from lower socioeconomic backgrounds exercise “considerable agency” – yet the recommendations imagine no collective form for that agency at all. Every action is something done for or to students by universities, funders, regulators and the RSC itself.

Co-design gets one mention in the body text. Student associations, reps and organisations get none. Students appear as voices to be amplified and experiences to be platformed, never as organised actors with their own infrastructure for change.

There are other gaps. There is no outcomes data anywhere – belonging, confidence and intent are all self-report, so whether the belonging gap actually converts into attainment and continuation gaps is asserted via the literature rather than demonstrated. The sample only includes students still in the pipeline, so those who left – the most diagnostic group – are invisible.

The RSC’s own commitments are also ordered. Its hardest lever – accreditation, through which it shapes degree structures, placement requirements and lab-hour norms in every chemistry department in the country – gets the softest language in the document, a promise to “inform the continual evolution” of standards. Convening and amplifying get firmer verbs. An accreditation condition requiring paid placement structures would do more than every roundtable combined, and the hedging suggests the RSC knows it.

The case for subject-level access work

But the report’s real significance is methodological. The analysis could not have been produced from access and participation plan data, and no standard institution-level intervention could respond to it – and both of those facts are an indictment of the level at which HE continues to do this work.

The mechanisms the report identifies are all subject-specific. The belonging gap runs through laboratory confidence. The cost barriers are lab hours incompatible with term-time work, placement-dependent progression, the integrated masters as the only fundable postgraduate route. The cultural barriers are chemistry’s own – long-hours lab culture, informal sponsorship into research groups.

A history student and a chemistry student at the same university, from the same background, are having materially different equality of opportunity experiences, and institution-level regulation treats them as one data point. The unit of regulation determines the unit of analysis, which determines the unit of intervention – and in England at least, all three are currently set at the wrong level.

The standard objection is burden – nobody wants 139 providers producing dozens of subject-level plans each. But the the analysis does not need doing 139 times by providers. It needs doing once per subject, by bodies with the disciplinary networks, the convening power and the accreditation leverage to act on it – which is to say, by learned societies and professional bodies.

It’s not the first to do so, but here the RSC has written a prototype, and even offered its methodology to other STEM disciplines in the introduction. Regulators and funders could commission or require exactly this, subject by subject, and hold institutions accountable for responding.

For that to work, students need to be in the room as actors rather than data points, because a report this good about student agency should not have managed to forget that students organise. Nevertheless, cracking report.