The black attainment gap is a long standing issue that the sector is still struggling to treat with the urgency that it deserves.
Now, in light of TEF and the grading of institutions, it appears that there is a missed opportunity to use the brutal publicity of an Olympic-style medal awarding system to punish those whose attainment gaps are starkest.
First class degrees
Our friends at HESA have kindly helped us crunch the data showing the attainment gaps between white and black UK students at different institutions last year. The data shows that there is a diverse range of institutions where an attainment gap exists between white and black graduates. Astonishingly, there are six universities where not a single home black graduate got a first in 2015-16, including the universities of Buckingham, Oxford, and Exeter.
The disparity in awards of first class honours is stark when examined along the lines of race. It is interesting to see that a wide range of types of institutions are guilty of this disparity. Perhaps more worrying is that fact that despite the shocking nature of these statistics, the majority of institutions where a black student is far less likely to achieve an outstanding degree as a white student have been rated as Gold and Silver in the TEF: over a quarter of institutions with the worst attainment gaps are rated Gold, and over half are rated Silver. This is despite TEF’s attempts to account for differential student outcomes through the use of split metrics.
Upper second degrees
The pattern persists for the awarding of 2:1 degrees as well. While there are some institutions where black students are awarded more upper second class degrees than white students, it appears that this is at the cost of a wide attainment gap at a first honours level. Imperial College London is an example, where more black students achieve a 2:1 degree than their white counterparts by a 20% difference. However, 14% fewer black students find themselves with first class degrees. Black students at the University of Oxford achieve 13% more 2:1 degrees, but receive firsts at a 34% lower rate. The numbers in both cases are small, but significant enough: Oxford had 25 black graduates in 2015-16 (compared to 2145 white graduates), whilst Imperial had 30 black graduates (compared to 555 white graduates).
While there are those who have a tendency to defend the sector using a student deficit model to explain away the data, this is nothing short of a crisis for higher education. Unfortunately, the crisis is not a new one, and we should be asking ourselves how to make an equitable university system, where your grade is not influenced by your skin colour.
And with TEF being promoted as a tool for student choice, there is a serious question to be asked about the ethics of grading institutions Silver or Gold if a black student attending said institution is unlikely to be awarded top marks. It appears that in many cases it is Gold for white students, but Bronze for black students.
Analysis of TEF retention data highlighted that there were unexpected inconsistencies between groups – with negative flags showing up for non-continuation rates for white students more frequently than for black students. In this case, there is a clear, consistent trend of white students achieving better grades than black students. It’s indisputable, and there is clearly an institutionalised problem of how institutions treat and educate black students.
If we look at the full range of ethnicity and attainment data below we can see that in some – though fewer – institutions, there are also significant attainment gaps between white and Asian students.
A couple of caveats – this data does not control for other variables such as prior attainment or subject mix, but the universal nature of this problem across the sector points to it being a profound institutional problem.
What can be done?
The new regulator could take upon itself the opportunity to penalise institutions for attainment gaps. It is time for penalties over incentives: it’s hard to think of an issue more urgent, but we all know how slowly the sector usually moves. While TEF in its current format looks at multiple metrics, including race-based split metrics, it seems that it hasn’t picked up on universities’ failure to support black and Asian students to succeed. Ultimately, TEF has failed in its aim to “take account of any significant differences in the quality of teaching and learning experienced by different student groups” (to quote the TEF year two consultation) if it has awarded universities Gold ratings when there are significant racial attainment gaps.
The challenge is well put in this presentation on BME attainment: “it is ethically dubious to promote the idea of widening participation in such groups [ethnic minorities] if they cannot be guaranteed equitable outcomes”. In this case, it is ethically dubious to hide away the fact that the university experience at many institutions produces inequitable outcomes, and fail to do anything about it.
Further analysis on the black attainment gap from HEFCE and ECU can be found here, here and here.
CLARIFICATION: This article was amended at 4pm on 11/8/17 to clarify that all the data published here is for home undergraduates only and does not include international students.
Excludes ‘other’ and ‘unknown’. This table is best viewed in full screen.
24 responses to “Universities’ shame – unpicking the black attainment gap”
How closely is final degree classification correlated to qualifications on entrance? Do these ethnic groups generally start with lower entry tariffs and their spread of degree classifications broadly follow the same pattern as for other groups?
I’m not denying there is a problem here but the difference is between not doing enough to address an already existing gap and active discrimination at University
By excluding “unknown” (I assume you mean unclassified) you are missing the number of students who qualify as medics/dentists.
Having looked at the Heidi+ data there are black Oxford graduates who have received unclassified awards in A3 clinical medicine (rounds to 0).
That’s not to say there isn’t a problem, but whole cohorts of students are missing from the analysis if unclassified medics/dentists are excluded.
HEFCE research finds that even after accounting for prior attainment (and other factors) there is still a significant gap. http://www.hefce.ac.uk/analysis/degree/ and http://www.hefce.ac.uk/pubs/year/2015/201521/
Yes, this does exclude students qualifying as medics and dentists – as the qualification as different it can’t fit the same analysis in my opinion.
You should caveat that it’s those courses that have been excluded as it is misleading just to say unknown/other (“unclassified”) have been excluded from the analysis – it is easy to conflate unclassified with “failure”.
In the case of Oxford, the “unclassified” graduates are more diverse than the cohort of 1st class grads (20.9% of the unclassified are BME, and make up 9.4% of the BME cohort).
Absolutely worth reading previous government 2007 report on this: http://webarchive.nationalarchives.gov.uk/20130321020039/https://www.education.gov.uk/publications/eOrderingDownload/RW92.pdf. The report controlled for a wide range of factors, including prior qualifications, but still found an unexplained gap: “Our results show that, even after controlling for the majority of factors which we would expect to have an impact on attainment, being from a minority ethnic community (except the “Other Black”, “Mixed” and “Other” groups) is still statistically significant in explaining final attainment, although the gap has been significantly reduced”.
It led to a greater focus on this important issue, including work by ECU and HEA – http://www.ecu.ac.uk/wp-content/uploads/external/ethnicity-gender-and-degree-attainment-project-final-report.pdf. But it’s still a problem – see HEFCE’s current funded projects: http://www.hefce.ac.uk/sas/barriers/ 1/2
I’ve never believed trying to dig down to the nth degree in the data is the answer here (at times some of this type of activity has felt like trying to explain the issue away through granularity), or highlighting a range of other stats instead about participation etc. But rather some answers might come from institutions acknowledging and recognising there is this gap and having a range of actions to try and address it, seeing what might work.
All too often when presented with this data the first question is ‘ what were their entry qualifications?’ Innocently asked but it’s a question which distances us from culpability.
Like for like comparison of pre entry quals between WP groups shows white WP students have better attainment rates than some BME groups, particularly black males.
There is no doubt the school system lets down underrepresented groups but we also train the school teachers- the reach of our influence on perpetuating inequality as well as our potential to tackle it shouldn’t be underestimated.
I think this is exactly right Amerjit. This is a long-standing issue that needs more action to try things – and an acceptance that some things may fail – rather than more and more analysis.
There’s plenty of screaming and shouting about the attainment gap but not enough focus on how to improve it. Research shows us that the curriculum and academics themselves are not diverse. There’s also disparities and cultural differences between students who study at their local university and those who move away. Those staying local are most likely to be from BAME backgrounds or disadvantaged circumstances.The fact that there are universities where NO black graduate has received a first is disgraceful.
70. Figure 12 shows the changes for the 2013-14 graduates by entry qualification. This shows
that, after accounting for entry qualifications, White graduates have a higher proportion who
gained a first or upper second class degree; the same relationship holds true for firsts alone. The largest difference for a first or upper second class degree is among graduates entering with a prior HE level qualification, where the proportion is 21 percentage points higher for white
graduates than BME graduates. This compares with a five percentage point difference for those
entering with four A grades at A-level. The observed differences are thus not solely due to entry
qualifications. (http://www.hefce.ac.uk/media/HEFCE,2014/Content/Pubs/2015/201521/HEFCE2015_21.pdf Page 20/21)
Unfortunately, to make a change we need this deep level data analysis (much deeper than the article shows). The first reaction of any academic you speak to about the attianment gap is: A) it’s the fault of BME as they… have BTECs instead of A-levels, they went to worse schools, they have lower grades on entry, are from lower socio-economic backgrounds* (*delete as appropriate), and B) it doesnt happen here – i’m objective and the data is for other universites, faculties, departments, schools, courses, modules* (*delete as appropriate). So they will always blame the studnets and will always push the issue away from themselves. So we need strong data. For example, Kingston Unviersity have drilled down their data all the way to module level, and compared like for like (same entry qualifications, same degree, and even same socio-economic backgorund) and still found the ethnicity attainment/awarding gap indicating that the fault must be systemic as even black students from middle class backgrounds had fewer chances of getting a good degree compared to comparable white students from lower socio-economic backgorunds. But you’re right in saying that universities need to own up to the issue. I’m afraid, though, that unviersities – who care about reputation because with reputaiton come money – will very reluctantly accept that they have an issue – unless you through it in their face. This article can do that… but it will not make the change on the ground as the unversities will then have to work with individual schools and even academcis – and they will need to be armed with data specific to that school/academic.
There are only a couple of hundred black professors in the UK and no black senior managers.
We could start to fix this at the “top”.
Part of the problem is that when you start uni you’re never given a proper induction into how to go about getting the grades or any encouragement to go for the grades. As many BME students are from state schools there is not a general culture of expectations about aiming high or competing to be the best right from the start. But those from public schools may have this ingrained in them.
Good analysis, but may be of interest to know a growing number of universities are working to address the attainment gap. For example Hefce have awarded £0.5m for Kingston University to share its value added metric, (which describes the gap after taking account of entry qualifications and subject of study) and it’s inclusive curriculum framework, with five institutions including Russell Group
At my university students from public school do worse conditional on A-levels. I think it’s general. So should we be working to eliminate this attainment gap?
I believe that about 50% of the gap of the gap can be attributed to PEQ. So 50% of impacted students can have the equivalent PEQ to a white peer but are 30% less like to obtain a good degree.
Completely valid point. If an institution doesn’t provide BME role models how can you possibly inspire students? What message does it send when there are so few black (or BME) professors and other bme staff can’t get past lower management.
And if you compare that against the images on University websites – e.g. you sign up for a diverse institution that has pictures of successful BME people teaching and studying, but when you arrive on campus its incredibly eurocentric.
The same can probably be said for the curriculum. How many positive BME examples do we have in our curriculum?
Leaving aside the many external factors that probably have an influence here and which universities cannot be expected to address, one factor that has been tried and failed where I work is to introduce anonymised submission for course work assessment; and subsequently, anonymous mark handling in exam boards. Perhaps we should consider the exact opposite route, so that we can track student progress in a properly personalised way? (The ideas of “personalised learning” and “anonymous assessment” being uneasy bedfellows.)