Making Access Fair: Some thoughts on this year’s admissions cycle


‘Fair access’ is a term that appears in almost every conversation about the future of higher education, but has no universally agreed definition. What the ‘fair’ in ‘fair access’ might mean looks different to different people: but as an Oxford college, we urgently needed to decide what it meant to us. At Worcester, at the beginning of 2019, we set ourselves a definition of ‘fairness’ which we agreed as a College to work towards in our access and outreach work. Our work, we decided, would be shaped by the idea that access to Worcester College could be regarded as fair if the demographics of our offer-holders matched the demographics of those who achieve AAA or higher at A-Level (AAA is Oxford’s standard conditional offer for the majority of its courses).

We analysed the data with attention to four demographic groups, all of whom are under-represented at Oxford: students studying at state schools; students who live in areas where likelihood of progression to higher education is low (measured using POLAR quintiles, with quintiles 1 and 2 recording the lowest rates of progression to higher education); students who are socio-economically disadvantaged (measured using the ACORN consumer classification system, with categories 4 and 5 registering the most socio-economically disadvantaged areas); students who are not white (sometimes called ‘BAME’ or ‘Black and Minority Ethnic) applicants. Among those who achieve AAA+ at A-Level nationally, roughly 73% are educated in state schools, around 13% live in POLAR quintiles 1 and 2, around 11% are classified as ACORN categories 4 and 5, and around 21% are BAME. The first question we asked ourselves was: are the demographics of our offer-holders shaped by the demographics of our applicant pool? Is it the case that we will have to change who applies through a long-term programme of outreach before we can change who is offered a place? The answer, it turns out, is: not really. Between 2014 and 2018, the University rejected the applications of 2000 students whose applications were flagged as ‘disadvantaged’, and who went on to either meet or exceed what would have been their conditional offer, had they been offered a place. Applicants belonging to these demographic groups were making applications, but their applications were not all equally likely (in a statistical sense) to be successful. Of all UK applicants who applied to Oxford in 2018, for example, independent school students were slightly more likely to be offered places than state school students: of the 8,207 state school educated students who applied to the University in 2018, 1,789 were offered places, whereas of the 4,265 independently educated students who applied the same year, 1069 were offered places.

Certainly there remains a huge amount of long-term outreach work to be done – great numbers of talented and successful students studying for their A-Levels at state schools do not consider making applications to Oxbridge, and we are striving to reach them – but our question for admissions in 2019 was, what could be done to make access fairer now? Unless we were to content ourselves with the idea that students educated at independent schools, or from areas of high socioeconomic advantage, necessarily had greater potential to thrive in higher education than their state school educated, or less advantaged counterparts (something which does not turn out to be true in students’ actual performance once they are studying on course at University), we needed to examine the ways in which we, as admitting tutors, went about assessing potential in our applicant pool. The stated aim of the admissions process is to identify those students with the greatest academic potential: as the University’s website puts it, ‘The University of Oxford is looking for students with the highest academic potential, regardless of background’.

The problem (as the University website acknowledges in its subsequent sentence) is that potential cannot be assessed ‘regardless of background’. A huge range of factors are known to have an impact upon how students perform at school and sixth form, including (but not limited to) the type (and performance) of school that they attend; the socio-economic position in which they live (and specifically, the area of the UK in which they are resident); the amount of information and experience that those around them are able to supply about higher education; and the level of encouragement and support they receive in making their application to university. If, as admitting tutors, we think of potential as something we can measure by reference to particular kinds of school performance and the wielding of particular kinds of cultural capital and experience, we will be disadvantaging those who have not enjoyed the same advantages as others in the application cohort – even if we do not intend to do so. Disadvantage in university admissions does not operate through the wilful exclusion of certain kinds of students by admitting tutors – admitting tutors recognise better than most the huge academic and intellectual benefits to a diverse student body. It operates via the assumption that it is ‘fair’ to look for similar indicators of potential in students who have had vastly different opportunities to develop the kinds of skills we measure and evaluate during the admissions process. This was what we needed to change: we needed to ask our tutors to think critically about their expectations of applicants, and to identify ways to discern potential in those applicants who had not (yet) had the privilege of learning to present their academic ability in the ways in which we had grown to expect to see it in more advantaged applicants.

Oxford University has this year developed a new way of using contextual data in admissions decisions, and announced two new schemes that respond the fact that not all students have received the same opportunities in their secondary school and home environments. Full details of these schemes are available on the University’s website; both operate on the assumption that with slightly more preparation and teaching than their more privileged counterparts (a short pre-degree bridging course in the case of Opportunity Oxford, and a full foundation year of study in the case of Foundation Oxford), students who have experienced disadvantage will be given the necessary skills and experience necessary to flourish on their Oxford courses. It is important to note that Opportunity Oxford, launched this year, does not involve any element of contextualised offers: although they will take part in extra teaching over the summer vacation before they begin their studies, these students have won their places at Oxford in competition with the full application cohort, and must achieve same standard offer at A-Level.

At Worcester College we have welcomed the news that the University is setting its sights on more ambitious projects to ensure fair access to Oxford, and intend to play a full part in the roll-out of these programmes across the collegiate university over the coming years, but this year we have used the new contextual data available to us in a different and more immediate way, and have not needed to rely on the offer of extra teaching in order to admit disadvantaged students in far greater numbers than we ever have before. It is unnecessary and unhelpful to assume that disadvantaged students who are able to achieve AAA+ at A-Level are in need of extra teaching in order to be realistic candidates for admission to a highly selective university. Having discussed the need to empower our tutors to recognise diverse potential, we entered the 2019 admissions cycle (for admission in October 2020) with more data on applicants’ background context than ever before, informing our decisions as admitting tutors. The results have been striking for three key groups: our pool of 2020 UK offer-holders is representative of the pool of those who achieve AAA+ at A-Level nationally in three key ways: 83% of our UK offer-holders are studying at state schools; 20% are from areas of low levels of progression to higher education (POLAR quintiles 1&2), and 22% are from socio-economically disadvantaged areas (ACORN categories 4&5). Once we had asked admitting tutors to engage critically with the idea that fairness in admissions might not mean expecting potential to manifest in similar ways in each applicant it was obvious to us that these candidates were simply fantastic candidates for admission.

Empowering our tutors to recognise diverse potential has changed the overall picture markedly. The most disadvantaged students became more – rather than less – likely to be offered places than the most advantaged. The most disadvantaged candidates (Band A according to the University’s Contextual Banding system) in our College cohort of applicants were this year more than twice as likely to be successful as those marked as the most advantaged in every category (Band D). This is the kind of statistic that draws the accusation of ‘social engineering’ from certain corners of the press, who presume that ‘fair access’ must mean the active disadvantaging of independently educated applicants by highly selective universities. It is easy to see why some people feel this way: certain kinds of schools have worked very hard to become experts in training young people to present themselves and their abilities in ways that are particularly measurable in the assessment processes of university admissions systems. But if universities begin to think critically about how to uncouple privilege from assessment of potential, young people from very privileged backgrounds and who have benefitted from very advantaged educations will continue to be offered places at highly selective universities; but they will be offered them on the basis of their potential itself, not on the basis of the ability their schools have equipped them with to excel at certain kinds of admissions metrics.

All of our offer-holders, regardless of background, are expected to achieve similar A-Level grades. It is obvious, then, why those who have achieved this against a background of disadvantage should have a higher rate of success in admissions: these are, by definition, students of huge potential. Each applicant is an individual with an educational and environmental experience that is specific to them alone, but it is clear that a disadvantaged student will have had to work more independently, and overcome more obstacles, in order to be predicted to achieve AAA at A-Level and be in a position to make a competitive application to a university like Oxford.This is not to say that there are not very privileged students in Worcester’s cohort of offer-holders for 2020, who absolutely deserve their places at Oxford to exactly the same extent that their more disadvantaged peers do (they are there, in numbers more than proportionate to the national population: about 13% of the UK school population attends independent schools). But those students have their places because their potential impressed their tutors in the context of the opportunities and advantages they have had.

We are proud of the progress we have made this year at Worcester – but our work has only just begun. We are now able to say that this year’s pool of offer-holders is ‘fair’ in that it matches the demographics of those who achieve AAA+ at A-Level for those who are educated at state schools, live in areas of low levels of progression to higher education, and are in the most socio-economically disadvantaged groups. UCAS does not release data about the ethnicities of our offer holders until later in the admissions process, so we do not yet know whether we have achieved representative admissions of the 21% of students who achieve AAA+ at A-Level who are BAME (and we will not be able to consider our admissions to be fairly representative until we have assured ourselves that applicants from certain ethnic and racial groups are not less likely to be offered places). But even if we discover later this academic year that 21% of our offer holders are BAME, and if we repeat this year’s performance in all future years, our access work will not be over. Being representative of the pool of those who achieve AAA+ at A-Level is only one way to define ‘fair access’ to higher education. The demographics of this pool do not accurately reflect the demographic breakdown of the population. While approximately 73% of those who achieve AAA+ at A-Level are educated in state schools, state school educated students account for nearly 87% of the population of school students. Similarly, socio-economically disadvantaged students (ACORN categories 4 and 5) account for 20% of the population, and only 11% of those who achieve AAA+ at A-Level. The gap is exponentially wider for those who live in areas of low levels of progression to higher education, who make up around 40% of the population, but only around 13% of those who achieve AAA+ at A-Level.

Inequality does not suddenly materialise at the moment of application to higher education: individuals’ educational careers are affected by a variety of different manifestations of inequality throughout the course of their lives. The next step for us at Worcester is to set our sights on a fairer educational landscape at all levels, to take our sector-wide responsibility as an education provider seriously by increasing our long-term and sustained engagements with children and young people in schools and sixth forms, so as to make those who are most disadvantaged just as likely to excel at school as their more advantaged peers. There is so much work to be done to combat inequality across the educational landscape and wider society, and it we will continue to do that work. But we don’t need to wait for this work to happen before access to universities can be ‘fair’. Universities can have representative admissions now, by empowering tutors to recognise diverse indicators of potential. As the way that inequality manifests in education continues to shift over the coming years, we will need to constantly renew this work, but as we do so our aim is to move towards an undergraduate intake that is representative of society as a whole. We are so excited about teaching and learning from the increasingly diverse cohort of students that doing this work will enable us to admit, and we look forward to the challenge of opposing inequality in education at all levels.

Dr Marchella Ward, Tinsley Outreach Fellow

Professor Laura Ashe, Tutor for Admissions

Professor Kate Tunstall, Interim Provost