This article is more than 6 years old

Data-riffic! What does data tell us about UK higher education in 2017?

More data is available about universities than ever before, and new software means it can be visualised to tell new stories about our sector. David Morris looks at data on TEF, REF, widening participation, university finances, and student numbers.
This article is more than 6 years old

David Morris is the Vice Chancellor's policy adviser at the University of Greenwich and former Deputy Editor of Wonkhe. He writes in a personal capacity.

To quote Graeme Wise in his introductory post for Wonkhe’s data blog a few years ago, “in the 21st century, data makes the world go round”.

We have more data available about British universities than ever before. Indeed, thanks to the coordination of organisations such as HESA, UCAS, Jisc, and the work of governments, funding councils and (yes even) league table compilers, the capacity for data comparisons and analysis of universities’ is easier than ever. Data can used to assess performance, to instigate market competition, to identify ‘hot spots’ and ‘cold spots’, and perhaps most importantly for higher education wonks, to understand patterns and trends (to steal the title of an annual UUK publication) in how our universities are operating.

Data can reveal to us essential truths about how our higher education system performs and operates. There are many different ways of approaching this, ranging from crude league tables and rankings, to sophisticated statistical analyses of larger and larger datasets. Finding the right balance between ease of understanding and complexity of analysis is a constant challenge for those of us who use these datasets. Thankfully, help is on offer from the data visualisation software Tableau, which you may have noticed we have become quite fond of in recent weeks.

Though data is a wonderful tool, we must always acknowledge that it can never give us the whole picture. Universities may appear by multiple data metrics to be very similar, and yet be completely different. Single data points, like those presented here, can never quite capture the diversity of delivery and activity that takes place within a university.

What follows is only the tip of a very large iceberg of data possibilities. The follow visuals do not necessarily present any firm conclusions about what universities are doing or what they might do better; indeed, much of what they show may be of no surprise at all to those familiar with the sector. But they will hopefully provide an insight into the comparative state of universities, their relative strengths and weakness, and the interrelationships between the many variables upon which universities’ success (or failure) is judged in the modern world.

Research, prestige, and labour market outcomes

If ever there were proof of how a universities’ age and research performance create a hierarchy of prestige in UK universities, the below graphs show it. A combination of the two enable the oldest and most research intensive universities to recruit the best students, and typically those from the most privileged backgrounds. These in turn influence employers’ preferences in the graduate labour market, so perpetuating the cycle. This is the power of the ‘signalling effect’ of higher education, originating in two variables (research performance and age) that have nothing to do with teaching excellence, quality of student experience, or the value of qualifications to the labour market. Make of that what you will.

All this said, graduate salaries vary wildly by subject and institution. The vast majority of subjects have a median 5-year-graduate salary above the average for 25-29 year olds.

TEF

Below is a graph that you may have seen before, with added annotation to demonstrate the differing strategic balances of research and ‘teaching’ performance that universities find themselves with.

As we can see, in each broad entry requirement comparators, TEF outcomes are relatively well spread, though more Bronze institutions are found in the lower tariff group than in the highest. It will be interesting to watch whether TEF has an effect on competition within these tariff groups in the coming years. Will Gold really prove an advantage for recruitment, and will damage be done to those with Bronze?

The extent to which this happens may depend on geography. Bronze awards for universities are very much concentrated in London. This may also limit the damage that TEF might do to international recruitment, given the allure of studying in the capital for many international applicants.

Access and ‘social mobility’

Universities are still – with only a small number of exceptions – dominated by the middle-classes. Some institutions – mostly post-92 universities – can stake a claim to being ‘comprehensive’, with a genuinely diverse balance of entrants from different social backgrounds. However, many of these institutions are also likely to be smaller in size. The ‘heavy lifting’ for widening access will have to come from larger, more exclusive universities, particularly those based in large provincial cities such as Manchester, Liverpool, Leeds, Sheffield, Nottingham, and Birmingham.

Recruitment

The lifting of student number controls has upped the competition in the recruitment market, and we are beginning to better understand which universities have been most successful here. Whilst more established and prestigious universities are able to trade upon their reputation, the competition is far more fierce for modern universities, where performance in those metrics used in TEF has been far more important to ensuring a healthy intake.

Finances

Increasingly, universities’ financial situation is dependent on their student recruitment, particularly as research funding becomes ever more concentrated – both geographically and around a smaller number of institutions. Yet despite disparate fortunes in expanding student numbers, highly leveraged borrowing is to be found in a wide-range of institutions – some with large asset bases, and others with very small ones.

Salaries

Given all the furore regarding salaries in universities in recent weeks, we couldn’t resist looking at whether salaries could be in anyway connected with universities’ performance, both for vice chancellors and academic staff.

 

Notes on the data

Download a spreadsheet containing much of the data used above here

Credit must firstly go to my Wonkhe colleague David Kernohan for beginning to piece together some of these data sets, in particular his work on the LEO dataset, and also for his help and guidance in navigating Excel and UKPRNs (or lack thereof).

Most of these graphs were produced before the University of East Anglia had its TEF outcome changed from Silver to Gold on appeal.

Almost all the data here excludes Northern Irish universities, and some graphs (those using POLAR data) exclude Scottish universities. There are gaps in different datasets for some small, specialist, and private providers.

Sources:

  • TEF – HEFCE, and Wonkhe’s ‘flag’ score explained here.
  • REF – 2014 GPA score from Times Higher Education.
  • Finances – HESA
  • UCAS acceptances – UCAS
  • POLAR data – UCAS
  • LPNs data – HESA
  • University foundation – multiple sources collated in this Wikipedia article.
  • LEO data – Department for Education. ‘LEO score’ uses a mean of subject medians, weighted by graduate numbers, and converted to a ‘base 100’ metric.
  • Entry tariff data – Guardian
  • Salary data – UCU

9 responses to “Data-riffic! What does data tell us about UK higher education in 2017?

  1. Excellent–but there should be some why to find a particular university besides clicking each and every dot! (If the particular university is selected in the drop down menus, then all other universities disappear.)

  2. Thanks for the feedback. I’ve added the ‘highlight’ feature to a couple of the viz’s that didn’t already have it. Hopefully should help you find particular institutions amongst the dots.

  3. HI David – love the article. Can you help with the source of the 16-17 entry tariff data?

  4. Accessing and manipulating these kinds of datasets would be a great thing for WonkHE to offer training in, if there was ever any plan to expand the courses you run??

  5. I love this – fantastic resource. Almost don’t want to share it to LinkedIn so as not to arm the competition!

Leave a Reply