A New University Ranking

A New University Ranking

Ranking universities by the proportion of entrants who go on to graduate jobs or further study.

The  higher education landscape is blessed with many detailed and accurate datasets. Many of the most commonly used sources of data, however, artificially flatter the sector’s performance. For example, the sector’s preferred measure of ‘non-continuation’ only looks at the proportion of  students who don’t continue from the first to the second year – rather than, as one might expect, the proportion that complete (or drop out) of the whole course. Similarly, most earnings data only looks at those in sustained employment. There are sometimes valid statistical reasons for this; however, the overall result can disguise what a layperson might wish to know.

I believe that one of the most important questions for an applicant – or a parent, teacher or funder is this: “What proportion of entrants go on to a good graduate job?” The table below shows the answer, by combining the drop-out rate with the proportion of graduates who go on to get a graduate job or be in further study.

HE provider Drop-Out Rate % Graduates in Graduate Jobs or Further Study % Entrants in Graduate Jobs or Further Study
Goldsmiths College 20.4 55 43%
University of Wales Trinity Saint David 13.5 54 46%
Middlesex University 19.2 58 47%
London Metropolitan University 26.3 66 48%
Birkbeck College 32.2 72 49%
Leeds Beckett University 21 62 49%
The University of Bolton 24.4 65 49%
Glyndŵr University 21.6 64 50%
Buckinghamshire New University 17.4 61 50%
York St John University 7.4 56 52%
University of Suffolk 21.6 66 52%
Newman University 17.6 63 52%
University of Abertay Dundee 20.7 66 53%
Kingston University 15.3 62 53%
The University of Wolverhampton 19.9 66 53%
Cardiff Metropolitan University 15.9 63 53%
The University of Northampton 13.8 63 54%
The University of East London 11.3 61 54%
Canterbury Christ Church University 14 64 55%
The University of Greenwich 16.1 66 55%
Southampton Solent University 14.3 65 55%
University of Bedfordshire 18.3 69 56%
The University of Winchester 9.8 63 56%
Roehampton University 17.6 69 57%
The University of West London 18.5 70 57%
Bath Spa University 9.5 64 58%
St Mary’s University, Twickenham 18.2 71 58%
Anglia Ruskin University 14 68 58%
SOAS University of London 17 70 58%
The University of Sunderland 14.4 69 59%
University of Ulster 12.9 68 59%
Edinburgh Napier University 10.6 66 59%
The University of Central Lancashire 15.9 70 59%
The University of Westminster 12 68 59%
The University of the West of Scotland 16.2 72 60%
The University of Brighton 11.6 68 60%
The University of Chichester 6.5 65 60%
University of Worcester 10 67 61%
City, University of London 10.4 68 61%
Bournemouth University 12.2 70 61%
Leeds Trinity University 9.5 68 62%
University of Cumbria 10.3 69 62%
University of Chester 10.8 70 62%
University of Plymouth 11.1 70 62%
Birmingham City University 11.3 71 63%
University of Gloucestershire 8.3 69 63%
Bangor University 9 69 63%
Falmouth University 11.4 71 63%
Sheffield Hallam University 8.2 69 63%
University of Derby 13.5 74 64%
Glasgow Caledonian University 8.6 70 64%
Staffordshire University 19 80 65%
Edge Hill University 8 71 65%
The University of Hull 14.7 76 65%
University of Northumbria at Newcastle 9.9 72 65%
Queen Margaret University, Edinburgh 15 77 65%
Oxford Brookes University 9 72 66%
The University of Manchester 5.3 69 66%
Royal Holloway and Bedford New College 6.2 70 66%
University of the West of England, Bristol 14.6 77 66%
Brunel University London 8.7 73 66%
Aberystwyth University 10.3 74 67%
Liverpool John Moores University 8.1 73 67%
University of Hertfordshire 13.6 79 68%
The University of Stirling 8.9 75 68%
The Nottingham Trent University 10 76 68%
London South Bank University 16.1 82 69%
Queen Mary University of London 5.8 73 69%
The University of Huddersfield 10.9 78 69%
The University of Lincoln 8.5 76 70%
Teesside University 11.6 79 70%
The University of Bradford 12.7 81 70%
Heriot-Watt University 8.8 77 71%
The University of Leicester 7.2 76 71%
De Montfort University 11.9 81 71%
The University of Portsmouth 11.2 80 71%
The Manchester Metropolitan University 12.5 82 72%
The University of Aberdeen 7.9 78 72%
The University of Essex 10.3 80 72%
Liverpool Hope University 10.2 80 72%
Cardiff University 4.9 76 72%
The University of Edinburgh 4.1 75 72%
The Robert Gordon University 8.1 78 72%
The University of Reading 4.9 76 72%
The University of Surrey 6.2 77 73%
The University of Liverpool 7 78 73%
Coventry University 10 81 73%
Aston University 6.8 79 73%
The University of Strathclyde 7.3 80 74%
The University of East Anglia 6.6 79 74%
Keele University 6.4 80 74%
King’s College London 10.3 83 74%
The University of Sussex 6.3 80 75%
The University of Kent 8.9 83 75%
University College London 6 81 76%
The University of Southampton 5.1 80 76%
The University of Bristol 3.9 80 76%
The University of Sheffield 7 82 77%
The University of Dundee 9.2 85 77%
The University of Leeds 6.1 82 77%
The University of Glasgow 8.2 84 77%
Newcastle University 4.1 80 77%
The University of Warwick 3.8 80 77%
Swansea University 6.2 82 77%
Loughborough University 4.9 82 78%
The University of York 4.3 82 78%
The University of St Andrews 3.4 82 79%
The University of Lancaster 7.2 85 79%
University of Nottingham 4.4 83 79%
The University of Exeter 4 84 80%
University of Durham 2.3 84 82%
London School of Economics and Political Science 4 86 83%
The University of Bath 2.6 85 83%
The University of Birmingham 3.7 86 83%
The University of Oxford 1.5 85 84%
The University of Cambridge 1.4 88 87%
Imperial College of Science, Technology and Medicine 3.4 90 87%


Details on Data used

Drop-out rates

  • Drop-out rates are taken from Projected Outcomes (Table T5) of the HESA Non-Continuation: UK Performance Indicators 2016/17 (the most recent year).
  • Only students with ‘neither award nor transfer’ – meaning they do not leave with any award (whether or not it was the one they originally intended to take), nor do they are transfer to another institution – are considered to be dropping out. For the purposes of this table, transferring to either another institution or another award in the same institution are considered positive outcomes.
  • Further details on the methodology used to calculate the figure for each university can be found on the HESA page linked above

Progression to graduate job or highly skilled employment

  • Progression data is taken from the Guardian University League Table (2019) (the most recent year).
  • The original source of the data is from the most recent Destination of Leavers in Higher Education survey, which measures the destinations of graduates six months after leaving.
  • A graduate job is considered to be any job in the Standard Occupation Codes 1-3, which is the standard definition used as a positive outcome in most league tables and in the TEF.
  • Being in further study is also considered to be a positive outcome.
  • Those who said they were unable to look for work (e.g. through disability) were excluded from the study population.
  • Further details about the methodology or DLHE can be found on the Guardian or OfS websites.

General points

  • Only providers for which data was available from both sources are included in the list.
  • The outcome data is a projection, based on the best and most recent data available, but does not describe the past performance of any particular year. This is because of the different base populations used in each of the two datasets and because of the complexities of calculating the drop-out rates.

7 thoughts on “A New University Ranking

  1. While these figures might have some relative merit they are confounded with courses offered and Oxbridge employer bias. My main objection, however, is that they don’t take account of the average number if graduate jobs applied for by the graduates themselves. Saying that less than 50% of graduates get graduate level jobs implies that 100% of graduates try (or at least ought to).

    1. I accept that more complex explanations may explain some of the differences between Oxford and Oxford Brookes, or Nottingham and Nottingham Trent, they’re less able to explain away the vast difference between, for example, Portsmouth and Middlesex. Note that Portsmouth’s outcomes are (considerably) closer to Oxbridge’s than they are to Middlesex’s!

      On your last point, as I discuss in my recent Wonkhe article, the fact that employers are increasingly asking for degrees for non-graduate jobs is for me clear evidence that we’re in a situation of substantial over-supply in which signalling predominates. The Alison Wolf quote is a very pithy summary of the problem.

  2. Does the DLHE data used here take into account (or report) the proportion of students included in the survey. In previous years, it has been possible for institutions to “game” the system by preferentially including students in typically “well performing” disciplines.

    I think this issue has been addressed in the new scheme (where the more sensible lag of 18 months after graduation is used).

    1. Yes, this is an issue with all DLHE data and I agree the shift to the new scheme will address it. However, if any gaming has occurred, the real figures would be even worse!

  3. There are all sorts of other factors – based on subject/discipline, geographical region etc. So you might not expect a university in, say, North East England or West Wales, to find it as easy to find a graduate job than a student in London. This will particularly affect institutions where they have a larger local student population. Since that might also be more likely to be the lower tariff institutions, then care should be taken to ensure they’re not being “penalised” twice in league tables.

    I’ll also add that I get the impression that the under reporting has only really kicked off in recent years, which coincides with the increase in (perceived?) importance of quantitative rankings/league tables etc. I agree that it will make the data worse, but it probably more so at the bottom of your table as presented (the better performing end)…

  4. Really interesting: other commenters have flagged the main concerns – in practical terms I’d especially worry about the fact that low tariff institutions are tagged as ‘worse’ in a way that e.g. hopsitals taking harder cases aren’t and this would make it worse.

    Other worries about the grad job thing
    1. I’m not clear if people not working because they are primary caregivers count as ‘not in grad job’? This matters as if so it will tend to mean that providers with more female students look worse (not sure how big the variation on gender ratios is and thus the impact of this)
    2. For e.g. Goldsmiths you do have the problem that many people won’t be aiming for a traditional ‘grad job’. Ideally you want to know if they could get a grad job if they wanted, but this is obviously unrealistic in terms of these sort of stats.

    On Oxbridge employer bias point it really depends what the ranking is for. If employers are irrationally pro-Oxbridge then I as a student want to go to Oxbridge….

  5. On one specific point – people outside the labour market (e.g. full time care givers) get excluded, just like on national unemployment figures.

    More broadly, to me the discussion here reflects a broader over preoccupation with fine-tuning the results in the middle rather than confronting how terrible the results at the end are. Yes, we do need more sophisticated measures to assess at the mid-to-higher end – e.g. TEF – but equally, some of the results at the bottom would be considered not just ‘needs improvement’ but ‘failing’ in the school or healthcare sector, and that needs to be faced up to.

    It’s not just an abstract debate, there’s a huge human cost. For example, increasing the proportion in graduate jobs from 50% to 65% – i.e. the approximate level of Hallam or Brookes, mid-league-table, Silver-rated modern universities (so not saying everyone must be outstanding) – would, for a typically sized university, mean an extra 500 people a year in graduate jobs. That’s huge.

Comments are closed.

Comments are closed.