Tuesday, 28 February 2012

University admissions and "social engineering"

FT education correspondent and part-time data nerd Chris Cook published a fascinating analysis of the Oxford admission process here.  The point was to explain why pupils in state schools - especially, of course, those in poor areas - are far less likely than independent school pupils to gain admission to Oxford. It did this by breaking down the disparity in actual admission probabilities into three stages of the process:
  • first, the relative probabilities that students from different types of school got "very strong GCSEs"; ranging from 3.4% for a student from the poorest tenth of schools, to 23.4% for those from independent schools;
  • second, the probability that a student from each type of school who got "very strong GCSEs" did in fact apply to Oxford at all, ranging from 14.1% to 24.6% (even higher for pupils from the richest tenth of state schools;
  • third, the probability that such a student who did apply got admitted - over half for pupils from independent schools, but only about 15% for students from the poorest schools.  

So there were disparities at each stage of the process.  Students from state schools in poor areas were less likely to get very good GCSEs, less likely to apply, and less likely to be accepted.  Not surprising; but what surprised me was the interpretation of the results.  Chris concluded 
"The FT’s analysis supports the view held by elite universities that the problem lies largely not with universities but in the school system."
This is undoubtedly correct, so far as it goes. But Wendy Piatt, Director General of the Russell Group, seems to think that Oxford is entirely off the hook:
“Even when poorer students get the right grades they are much less likely than their wealthier peers to apply to top universities. We cannot offer places to those who do not apply.”
But that's just not what the data show.  What it shows is that a student from one of the poorest schools who gets the right grades is a little more than half as likely as an independent school student to apply to Oxford. That's bad. But a student from one of those who does actually apply is less than a third as likely to be accepted. That is worse.  And what is even worse is that Ms. Piatt and the Russell Group don't seem to accept that there is a problem.  Eliminating the disparity in acceptance rates would more than triple the number of well-qualified applicants from the poorest tenth of schools attending Oxford. 

The Russell Group's refusal to engage with the issue isn't just a matter of an off-the-cuff quote to a newspaper either; their submission to the Commons Select Committee takes the same line, arguing - accurately enough -  that 
The most important reason why too few poorer students even apply to leading universities is that they are not achieving the required grades at school
but simply ignoring the admissions disparity; even though that is the one thing that is actually under their direct control. 

Of course these data alone don't tell the whole story (as Chris was careful to emphasise); perhaps,even among those getting "very strong GCSEs", independent school pupils are better qualified, or applying for easier courses. And there might be something odd about the applicants from the poorest tenth of schools - although even if you look at the bottom three deciles of schools, applicants are still only half as likely as those from independent schools to be admitted. 

On the other hand, there is surely a strong argument that the innate ability of a pupil who, coming from one of the poorest state schools, still manages to get very strong GCSEs and apply to Oxford, is likely to be rather  greater than a similarly qualified independent school pupil.  And indeed there is statistical evidence that the second effect appears to dominate; as this study, coincidentally also about Oxford, found:
Private school students perform less well in final examinations relative to their GCSE results when compared with state school students
This is represented very clearly on this chart: 

So in other words, students from state schools who apply to Oxford are much less likely to get in than those from independent schools with the same grades; but if they do get in, they'll probably get better degrees.  This is not news - it was true when I attended Oxford a quarter century ago, so they've had plenty of time to deal with it. 

So it is in the light of the data above that we should read articles like this in the Telegraph, that claim that efforts to reduce this disparity are "social engineering"; or reports like this, from the "Fair Access to University Group", which simply assert without looking at admissions data that "it is clear that the problem lies not in universities’ admissions policies".  

In fact, what the data and evidence suggest is that Oxford is already, and has been for some time, engaging in "social engineering" in favour of private school pupils. It has an admissions process that favours them and admits them even though they are less likely to get good degrees. Reversing that might or might not constitute further "social engineering" but the evidence suggests it would improve not only make Oxford less socially exclusive, but more rather than less meritocratic.   (To be fair, I should note that Chris, who is much better informed on this than me, thinks that Oxford is now genuinely focusing on some of these issues). 

None of this is to understate the supreme importance of improving outcomes for poor pupils in state schools long before they get anywhere near university; Chris, the Russell Group, and the Fair Access to University Group are all correct that this matters more than anything else. But simply denying there is any problem at all with elite universities' admissions, as Ms Piatt does, or arguing that any pressure exerted by government on those universities to do something about it is "social engineering", as the Telegraph and the Fair Access to University Group do, misses the point.  The admissions status quo is bad for both the UK's best universities and bad for bright pupils from poor backgrounds. 


  1. Interesting! Perhaps post qualification admissions might help to redress the balance if students with high grades are then vigorously encouraged to apply to Oxbridge. However (and this is contentious!) this might only work if all interviews were banned (except for subjects like Medicine) to create a proper level playing field: there is no doubt that private school students have the edge here because of the extensive coaching and preparatiion they receive, based on years of historical precedent. Oxbridge would no doubt complain that they would no longer be able to identify those who would really gain from the "Oxbridge way" but if state school students with high grades genuinely outperform their independent school counterparts, who loses?

    I'd also consider imposing a quota of state school students on Oxbridge, but that really would be social engineering!

  2. blame the suit for not bowing before your genius. Engineering Careers

  3. The generally distorted pattern of elite recruitment has been known, as Jonathan Portes writes, for a long time. One future development will be in the extent to which elite failure (banks, government, press, company directors etc) is undermining national welfare. Our measuring sticks for educational success are themselves rather fragile reeds and it could be that the Public School and Oxbridge strand in our public life is becoming a burden to us. Genius will thrive (or not) regardless of educational context - but the ranks of wise and very capable people need to be nurtured from pre-school onwards. Such people are born in all parts of society.

  4. Properly interpreting this sort of data also requires knowledge of the subjects to which applicants from the various school backgrounds apply. Acceptance rates at Oxford vary dramatically between courses, and anecdotal evidence (though perhaps hard evidence exists?) suggests that students from disadvantaged backgrounds disproportionately apply to the most competitive, oversubscribed subjects (eg economics, PPE, medicine). That doesn't by default explain their lower rate of success getting into these courses, but it's possible that at that intense a level of competition, their 'strong GCSE scores' are insufficiently impressive, ie not rounded out by the 'extra something' that perhaps the more advantaged students are able to demonstrate.

    I would be very curious to see a revised set of admissions statistics were these students to apply to a broader, less competitive range of subjects - thinking in particular the humanities and some of the less-subscribed social sciences. (Of course, it makes sense that those from less well-off backgrounds are likely to shun the non-vocational for courses that provide - at least in theory - a more evident career path). Or, barring that, statistics showing success rates across different subjects.

    If it were only the most competitive courses where these students were having disproportionately bad luck (and elite admissions is, by many measures, luck), the analysis and recommendations would have to be reconsidered. For example, Oxford (-bridge) hopefuls from more disadvantaged schools might be educated by admissions counselors or outreach staff about the competitiveness of different subjects, and relative chances of gaining admission. In fact, this isn't limited to students from poor backgrounds, but is sage advice for any student (apart from the unusually exceptional, the 'too good not to takes') wanting to get into Oxford. And of course, it would be worth launching a qualitative exploration into the reasons for the disparity in admissions rates on oversubscribed courses.

    1. It's nice blog are give about the admission. If you want to get admission then read above information.

      school admission

  5. Good writing is something I can appreciate. You have made your points in a smart way. I am impressed with how interesting you have been able to present this content. Thanks for sharing a nice information.
    Top MBA Colleges in Bangalore
    MBA Colleges in Bangalore

  6. Admission In Delhi Ncr Engineering Colleges
    GITM Gurgaon Is The Top Engineering College And Business School In Gurgaon, India. Affiliated from MD University Rohtak And AICTE Approved.Offered Admission in B-tech, M-tech, B-tech (LEET), MBA, BBA .

    GITM - Global Institute Of Technology And Manegement
    5 Km Stone, Kheda Khurampur, Farrukhnagar-Haily Mandi Road, Gurgaon, Haryana (India)

    Our Admisison Helpline : +91 - 9555100100
    Email Us At: Info@gitmgurgaon.com

  7. Improvising one's knowledge is the biggest thing can be done in a life time. So as you have said above the best curriculum at the right time would make a lot changes to our life. Thank you very much for sharing this. I feel so grateful to you.

    Overseas Education Consultants Chennai
    Overseas education consultancy in chennai

  8. This comment has been removed by the author.