How to overcome the stratification of higher education

Stephen Gorard, of Durham University, says academic selection has more downsides than justification, and we can solve widening participation at a stroke with better targeting.
This article is more than 2 years old

Widening participation to higher education is widely misunderstood and therefore often misdirected. Instead of devoting so much time and money to a participation problem we do not face, we should focus on the potential of learners at a younger age, and eliminate the pointless hierarchy of selection within HE itself.  

Widening participation (WP) among under-represented socioeconomic groups is a long-term UK policy objective. The focus is usually on 18-year-old, full-time applicants. Relevant policy initiatives include bursaries and maintenance grants, free places in some home countries, outreach work, aspiration-raising, overhauls of admissions arrangements, FE in HE and HE in FE, a huge expansion of the system, removing caps on numbers, and contextualised admissions.

However, the whole field seems very confused about which groups are actually under-represented, and what the problems are that WP is intended to solve. The key confusion is the conflation of entry to HE in general, with access to highly-selective individual universities. Therefore, it is almost impossible to evaluate properly whether any policy initiatives are effective.

Access to HE in general

In England, around 33% to 38% of 18-year-olds enter HE each year. All analyses here are conducted using the National Pupil Database from 2006 to 2015, linked to the HESA undergraduate dataset for suitable years. Fewer than expected HE students from each age cohort are white, male, have lived in care or in poverty (FSM), or have special educational needs or disabilities (SEND). But, these characteristics are not usually those addressed by WP activities. The sex imbalance on entry is largely ignored. All ethnic minorities have long been somewhat over-represented in HE, and yet many are still the target of WP. Students with profound learning challenges are, perhaps by definition, unlikely to attend HE, and are not addressed in standard WP approaches.

Of all students achieving the equivalent of at least two E grades at A-level, around 75% enter HE in the next two years. The figure for FSM students in the 2008 cohort was 80%, and for those living-in-care, it was 91%. Most other purported shortage indicators show the same. Looking at only those eligible to enter HE each year, those groups who are apparently under-represented, actually enter HE to at least the same extent as the overall figure. In fact, perhaps partly because of WP, qualified students from some disadvantaged groups are actually slightly over-represented. The entry figure was 98% for black origin students with Key Stage 5 (KS5) qualifications, but only 70% for white. Both white and male students continue to be under-represented, even in proportion to their prior attainment. And yet neither group is a specific focus for WP, while black and FSM students often are.

Where does the problem lie?

The problem lies not in admission to HE, or even in the stratification of qualifications at the prior KS5. The problem lies in the entry to KS5 itself and earlier. The education system grows increasingly selective, with different learners funnelled down different tracks. Only around 52% of the cohort have historically continued to formal KS5 education, but the rates are much lower for male, FSM, and SEND students, for example. Once in KS5, most students attain minimum qualifications, and most of these enter HE. But, those entering KS5 already have much better KS4 qualifications than those who do not continue to KS5. The stratification of entry to HE is the stratification of entry to KS5, which is the stratification of KS4 and earlier attainment at school. In logistic regression analyses, predicting whether a student from KS1 onwards will enter HE or not, their background characteristics are irrelevant if prior attainment is taken into account. We could balance the intakes further by simply dropping the requirement for minimum KS5 qualifications, but WP at this stage does not really have any further role in general access to HE.

The 48% not continuing to KS5 are the more disadvantaged and lowest attaining students in the country. Nothing done from thereon in terms of widening participation will make any difference to their chances of attending HE. In fact, some will leave formal education with below functional literacy and numeracy. It is important to remember in any concern for educational justice that nearly half of each cohort is effectively ignored by WP activities and that the focus of all that policy spending and effort is on the more-advantaged half of the cohort, already on a trajectory towards HE. Of course, it may be argued that the more successful students at KS4 deserve or merit their higher attainment, but this is not the approach used by traditional WP advocates when considering KS5 attainment for contextualised admissions. There is a huge inconsistency in policy here. If disadvantage affects attainment at KS5 then surely it equally affects attainment at KS4 and before. In which case, there is no justification for leaving the 48%, those not continuing to KS5, out of current WP arrangements.

Access to selective universities

The distinct issue of how those within HE are clustered by their background characteristics in especially selective universities is more serious. Each university takes students for each course with the highest prior attainment that they can get away with, while still filling their places. And, because of the correlation between advantage and average attainment, this fills such universities with more privileged students. Universities do so irrespective of the difficulty of the course. For all of their faults, KS2, GCSE and A-level examinations are engineered to be the same for whoever takes them. It is not meant to be harder to obtain a maths GCSE in Liverpool than in Bristol. In the same way, it is not clear that a student taking a degree in Law actually needs different prior qualifications to study Law in Manchester than in Sunderland.

Setting a national minimum threshold for entry to any course, and so eliminating selection between universities, would solve the stratification of intakes. The HE system does not need to be internally selective, and there is no net gain for the country. The main function of selection, whether intended to not, is to hinder access to only some universities within the system, and in doing so to emphasise their differences in student background.


Playing around with the KS5 qualifications needed for entry to HE, such as dropping the equivalent of a few A-level grades for under-represented groups, will not benefit overall entry to HE and may exacerbate existing problems. Most under-represented groups, including; males, summer-born, and those with severe or multiple SEND, are not even considered as WP candidates. WP does not seem to b about overcoming injustice as such. Selection does not operate as a gatekeeper to HE itself, as much as a way of sorting students into the most selective universities. This problem can be solved simply by eliminating this kind of selection and making HE more open access. It could be non-selective just as the national school system is, but not compulsory, unlike the school system. It would be a proper national system – and every HEI could be funded and supported to be as good as every other (just as schools and colleges should be).

The research underlying these comments was funded by ESRC (ES/N012046/1, ES/N01166X/1).


13 responses to “How to overcome the stratification of higher education

  1. The thrust of this article seems to be that we should “simply eliminate selection” for entry to the most demanding courses. What it then fails to do is explain how to manage the obvious results that would entail for both students and universities. No amount of mentoring and additional support will see students who only managed an E at A level maths survive a high-quality STEM or engineering degree. It seems to me that this would then result in either (1) catastrophic failure rates for unprepared and less able students (and falling NSS, TEF and “value for money” scores) or a radical drop in the demands placed on students to perform, producing cohorts of less well prepared and less employable graduates, with a knock on effect on the UK’s reputation and attractiveness to international students. Which one is Prof Gorard hoping for ?

  2. Typically misdirected conservative response I am afraid. My point was that however demanding the course STEM or otherwise it is no harder if you learn in Oxford than in Liverpool. And we cannot defend status quo based on things like TEF! Carts, horses?

  3. So you’re saying that the law course taught at Oxford is as difficult as the law course taught at Sussex? That the Imperial STEM courses are as challenging as those at UWE? If so, I’d love to see some evidence for this claim. It seems to contradict my fair few years of undergraduate teaching.

  4. “Typically misdirected conservative response” ….. Sadly no, one informed by direct experience and numerical data. I was recently director of a STEM Masters course that had a significant overlap with our broader 4th year integrated BSc + Masters so very well benchmarked against a large (100+) student cohort. We recruited several students from “lower ranked” UK institution, with strong 1st class BSc degrees and average exam and lab marks well into the mid 70s across the board. Their level of performance against our existing cohort after following the same courses was catastrophically bad, with exam marks all down in the 10-40% range. One failed to complete, one failed and scraped a pass after retakes, one scraped a bare pass. So with that as a calibration marker, “high quality” STEM graduates from a lower ranked institution are 3-4 degree classes lower and >40% lower on average on individual exams than students from higher ranked institution. 1st class students from some institutions are borderline fails at others.

    Perhaps some more quantitative research is needed here, could you persuade your own institution to volunteer to trial entry by lottery for a couple of years into one of its STEM courses and analyse the results on degree classification, dropout rates and student satisfaction while keeping the delivery and level of examination the same ?

  5. Why on earth do you want a system based on geography of where you learn? Does not happen with other formal (and appropriately moderated) qualifications. If you think Law at Sussex, to take your example, is not a proper law course then the course should not take students. If it is, as we both know it is, there is no reason for it to be easier or more difficult than a law course anywhere else.

  6. To clarify, are you arguing that (for example) the Cambridge maths degree is no more or less demanding than the Sheffield Hallam maths degree? Or are you arguing that these two courses should be changed so that they are equally demanding?

    The former position is obviously false. The second misses the point that the two courses serve different purposes. One has an explicit goal of providing a foundation for future researchers, the other focuses on preparing students for employment outside of academia. What’s wrong with a diversity of purpose?

  7. Your first ‘choice’ is made up. The piece said “In the same way, it is not clear that a student taking a degree in Law actually needs different prior qualifications to study Law in Manchester than in Sunderland.” and “every HEI could be funded and supported to be as good as every other (just as schools and colleges should be).”. I suspect you added the first clearly false position rhetorically because the second point is so weak. Both courses could and probably do prepare students for a range of possible futures. And could do so more than they do at present. It is not true or sustainable that those currently under-represented at Cambridge (your example) are incapable of a career in research. The system is socially divisive to a great extent for no real reason, and that damages all of society in the end, It could only be justified by some huge benefit. Your ‘diversification’ is not it, and is really only the divisiveness re-spelled.

  8. The sloppy mistake you make in your reply is to confuse my “demanding” with your “good”. I am not claiming the Cambridge maths degree is better than the Sheffield Hallam maths degree, I’m claiming that it’s more demanding because it has a different aim.

    Still, I wish you luck with your attempts to persuade all 129 UK universities to adopt identical mission statements.

  9. Most of Stephen Gorard’s post presents an unassailable argument: ie the content that queries “which groups are actually under-represented, and what the problems are that WP is intended to solve.” The following is spot on: “The 48% not continuing to KS5 are the more disadvantaged and lowest attaining students in the country. Nothing done from thereon in terms of widening participation will make any difference to their chances of attending HE…There is a huge inconsistency in policy here. If disadvantage affects attainment at KS5 then surely it equally affects attainment at KS4 and before. In which case, there is no justification for leaving the 48%, those not continuing to KS5, out of current WP arrangements.”

    All the responses to date have concentrated on the last section: the bit that does not appear to be evidenced by the link to the ESRC funded research project (which is concerned with schools not HEIs). This bit is a version of ‘if my grandmother had wheels, she’d be a wagon’. If HEIs all taught the same courses in the same way and had no autonomy, then all could have the same minimum (or should that be maximum?) entry standards.

  10. I’ve had several years of HE teaching experience. I started off at Sussex, and finished off at KCL and LSE. The breadth of the material, and the rigour of the exam questions, at the latter, far surpassed anything I saw at the former. For instance, the Sussex public law course for undergraduates is nothing but an extended exercise in judicial review. It deals with the same question for the entirety of that course.

    Now, I want my proof. You’re the one claiming that difficulty is the same across all unis when both I and Roland claim it isn’t. Where’s your proof that the difficulty is the same, Professor? You’re making a claim and then resorting to appeals to common agreement like “we both know it is the same” (no, you don’t know anything about law teaching at Sussex, and I know it isn’t).

    Also, please, it has nothing to do with geography. It has everything to do with the institution and its legacy and teaching staff.

    If the rigour of the courses is different, then there is room for accepting that there should be a difference in entry requirements. Some students will cope in a more intellectually demanding environment better than others. This is not to mention the likelihood of higher drop out rates (and thus wasted potential) as students struggle and eventually drop out of courses due to their time intensive nature and high marking criteria. You can contest the degree, but the difference is there nonetheless.

    Of course, even if all the courses were the same, this does not support your conclusion that we should be picking at random. Smart, motivated people want to be with smart, motivated people. That’s how they exchange ideas and feed off each other in learning and competition. This is beneficial and not to be laughed at because someone with abhorrent grades jumps into the discussion and says “I want in too!”

    In summary, you have provided zero proof for the most fundamental of your claims. I’m embarrassed that someone at the top of the pecking order could come up with such a baseless article. Reflects badly on the rest of us.

  11. I think what some commentators are missing here is that, if you have a course of study that is so “demanding” that it is exclusive to high achievers (invariably the most socioeconomically advantaged) you must take responsibility for your role in reproducing social inequality, instead of burying your head in the sand.

    Basing admissions on a lottery would not work in the way we do HE right now. But if we care about social justice we can use more resource for increased academic skills support, foundation courses, inclusive teaching, repeating years to more fairly distribute the kind of educational opportunities we offer.

    Why should only well off 18 year olds be trained to be maths researchers and the rest have no choice but to go for a course that aims at employment outside the sector?

    I agree with Stephen’s aim. We need more thinking out of the box on this one.

  12. I am saddened by several of these posts which focus on just one aspect of the piece – selection of students by universities and other HEIs. Whatever the merits or otherwise of Stephen’s arguments about this, the main thrust of the article should make both institutions and government sit-up and take notice. I doubt very much that either will do so in any serious way.

    However, I think there are two aspects of this debate missing from Stephen’s piece.

    First, institutional WP approaches are required to face in several directions at once and not just on access: retention, student experience, outcomes,
    employability, are all in the frame following years of increasing regulation of WP funding and, therefore, what is expected of institutional WP policy and practice. Such stretching of what is meant by WP is, in my view, a necessary development (morally and not just politically) but it can mean that we may have taken our eye off the “access ball” and simply “gone with the flow” in terms who we should be focusing our efforts towards.

    Second, the stratification of higher education (something I have commented upon for many years) is not just a product of how students are dispersed across the system – critical though that is. It is also a consequence of the ways in which the status of some universities plays out in the public view of higher education; something which has been reinforced and shaped by the HE policy discourses over many decades. Such status differentiation results from, and in turn shapes, academic staff recruitment; research grant success rates and sizes; donations from highly successful alumni; and so on. In short, the stratification of higher education reflects the patterns of power, wealth and influence across the sector and will not be overcome even by radical changes to student recruitment strategies – necessary as these are.

  13. We have non selective schools with no sixth forms in our area which top out at KS4, the focus being on safety-net C grades to meet floor targets. Surrounding grammar schools provide KS5 and on to university. The issues with this system are numerous, and so not all grammar schools perform equally. For example, the POLAR data shows a major cold spot in one of our poorer grammar school wards for participation rates in higher education compared to a neighbouring ward where incomes are much higher (20.6% vs 46.2%). I will leave to the reader’s imagination which political party governs which ward, suffice to say both sides will tackle any issue with the exception of selective education. Additional issues are rurality and transport links: passing the 11+ comes with a free bus pass from designated transport areas, whereas, failing limits choices for the poor to whatever concessions/schools are available. The geographical scope of the POLAR map also shows grammar schools located in poorer isolated areas to be in slightly lower quintiles for participation in Higher Education compared to those located along better transport links. However, this doesn’t tell the full story as some grammar schools located in poorer areas have their own minibus fleets. The expansion of grammar school places appears to be more about capturing that top 25% so-called ability group funding comes with the child than about giving all children a chance to live in cohesive communities. Finally, the generalisation that all young people in grammar schools are wealthy is incorrect. The largely unsupported transition between KS4 secondary modern schools to KS5 grammar schools in poor areas can be very difficult for young people. Many have been rejected at 11+ but still manage to achieve against the odds in places that value grammar schools beyond human sensitivity.,of,young,participation,areas/

Leave a Reply