This article is more than 4 years old

Is Nick Timothy right about graduate salaries? A new look at LEO

The Prime Minister's former chief wonk has got stuck into the 'is university worth it?' debate this morning. It seemed a good reason, combined with A level results day, to present some new ways of looking at graduate salaries data from LEO.
This article is more than 4 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.

A level results day has provoked a further continuation of the university bashing of recent weeks, with former Downing Street Advisor and Conservative manifesto author Nick Timothy getting in on the game in the Telegraph.

I won’t get into evaluating all the in’s and out’s of Timothy’s argument. Suffice to say; they broadly follow the line of thinking articulated in his now largely abandoned manifesto, questioning the value of university expansion and beginning with an anecdote about his barber, a ‘football studies’ graduate from Southampton Solent. Beyond these lazy cliches, Timothy is a wonk at heart and is heavily influenced in his thinking here by Professor Alison Wolf, who has rigorously questioned the assumptions upon which market-led and uncontrolled expansion of bachelors programmes have been based.

At the heart of the debate provoked by Wolf and now Timothy is the question of whether unbridled expansion of university courses driven by universities’ willingness to supply, applicants’ willingness to apply, and the government’s willingness to provide funding no matter what, is necessarily an efficient way of allocating our available nation resources for tertiary education. Wolf has made a very sophisticated argument that it is not.

Critical to the debate is the question of graduate destinations and salaries. This is one of the reasons why Longitudinal Educational Outcomes data is so important to the future of the sector, and for many courses it provides. LEO is a big help in answering the question as to whether different courses are ‘worth it’, at least in terms of pure human capital.

How precisely one cuts and interprets LEO data is a complicated methodological problem, one best left to the super-wonks of the Institute of Fiscal Studies and their ilk. I’ve written in detail about all the many many caveats that come with using LEO data before. It is not predictive, nor is it strictly a measure of the graduate premium. LEO carries a long time lag, and the raw data does not control for critically important factors such as prior attainment, social class, and region. The aggregated data (which I use below) masks huge gender pay gaps.

Nonetheless, Timothy’s comment about the football studies graduates from Solent provoked me to go back into LEO to find out whether his anecdote stands up. The answer is: probably not. Depending on whether football studies are classified under ‘mass communications and documentation’, or ‘business and administrative studies’, the median salaries in 2014-15 for those who graduated in these subjects from Southampton Solent University in 2009 were £23,000 and £24,600 respectively. That is significantly above the median salary for all 24-29-year-olds in work in 2014-15 of £20,800. Timothy’s barber appears to have been one of the unlucky ones, or perhaps Solent’s ‘football studies’ course is simply weaker compared to its other business or media studies offerings?

Having higher than non-graduate-average salaries does not necessarily make a bachelors degree ‘worth it’, either to the taxpayer or the graduate. It is difficult to discern whether the salaries commanded by university courses are down to the genuine productivity gains from higher skills developed on these course. Higher salaries could merely be due to the ‘sorting effect’: that graduates of these courses are preferable to non-graduates in-and-of-themselves, regardless of skills developed at university.

How much do graduates of different courses and universities earn?

I don’t think many students receiving their A level results read Wonkhe, but just in case they do, here are two graphs showing how graduate earnings five years after leaving university compare for different courses at different institutions. Crucially, it also shows how they compare to the national median wage for all 24-29 year olds. With creative arts subjects excluded, the graph shows how the overwhelming majority of courses lead to graduates earning more than the national median five years after graduating.

I’ve excluded creative arts subjects from these tables entirely. The challenges for creative arts subjects and graduate salaries have been well covered by Andrew McGettigan here.

The colour scheme for each data point incorporates the prior attainment data included in the LEO release by DfE. Broadly speaking, ‘1’ means a high entry tariff, ‘2’ a middle entry tariff, and ‘3’ a low entry tariff. Blue data points are those where the prior attainment data was not captured by LEO.

If you’ve decided on a broad subject area, this graph shows graduate salaries for different universities in each subject. You can highlight different universities and discern by prior attainment. For example, for applicants with lower attainment, the highest earning courses include social studies at Hertfordshire, engineering at Harper Adams, and nursing at Greenwich.

Alternatively, you can compare subjects within institutions, and institutions with each other, with this chart. This is a bit harder to navigate and is worth doing so in full screen. Note how there are no universities where all courses find themselves below our benchmark median salary.

8 responses to “Is Nick Timothy right about graduate salaries? A new look at LEO

  1. One problem as I see it is that we have a lot of data about the supply of graduates (HESA), a lot of data about fulfilment of graduates in the economy (LEO, DLHE), but patchy data about employment demand, particularly the indicators of unmet demand. If we really do want an evidence base for HE and the work, then there really needs to be much more comprehensive employment demand data. Currently only the AGR collect the sort of data required, and not only isn’t it publicly available, it only covers a small subset of employers; for the rest you have to patch together bits of UKCES, national graduate employment data, the employer vacancy survey, and all manner of other things and add a huge dollop of assumptions.

  2. Agreed. It’s really very sad that UKCES was shut down – it was far from perfect, but it was a useful source of evidence in this area. Unmet demand appears to be a very difficult thing to determine, for the prime reason that employers are always too negative about the supply of available skilled labour…

  3. @David the AGR survey is very useful in that it records the numbers of applications received for particular graduate roles as well as whether they were filled, so you can actually quantify the demand in particular occupations and sectors rather than just go on anecdotes. A proper national public survey like that would be wonderful.

  4. So is graduate salary really the only measure of HE value? Is this the only value of education? Or is Timothy’s barber a happier, more fulfilled member of society because he has spent three years at university? There is evidence that graduates: enjoy better health outcomes, are less likely to smoke, more likely to exercise, are less likely to be obese, less prone to depression, more influential in the community, more active citizens who are more likely to vote, participate in voluntary activities, show positive attitudes towards diversity and equal opportunities, etc etc – this obsession with higher wages is ridiculous: a postdoc probably works longer hours for lower wages than most unskilled workers, yet we assume a postdoc position is high attainment!

  5. I’d also be interested to see how graduate earnings after tax, NI and Student Loan repayments compare to the national average for 25-29-year-olds. Challenge for the weekend?

  6. That’s the true graduate premium. That’s for the super wonks to work out, along with getting access to this raw data in order to control for personal circumstances: region, class, gender, prior attainment and the rest. I’m sure the IFS are on it.

    Julian – this is fine, and true, and you’re not the first person to make this argument. But the sector has never proved that the ‘happier, more fulfilled members of society’ cannot be ‘produced’ (for want of a better word) for cheaper than the £9k + maintenance that the government (for it is the government if earnings aren’t high enough) stumps up. Just because something is good does not mean it is *necessarily’* good value for our money.

  7. David, Scott

    The ESS and EPS surveys on skills demand that took place under the UKCES aegis and were extremely useful on issues such as occupation shortage are continuing under DfE. ESS 2017 is currently in the field and will report next year. It surveys over 87,000 employers.

    These surveys do much of what Scott describes but were underused. Of course, they cover the whole economy and not just graduates but contain plenty of information about professional level occupations that is of great value to the sector.

    Nevertheless, the basic point stands.

    Graduates are not solely or even primarily motivated by the desire to maximise their salaries, and nor should they be (besides anything else the unintended consequences would be horrible for whole sectors of the economy; especially health, education and science), but the policy drivers and narratives appear to be reluctant to acknowledge that issue.

  8. “I’d also be interested to see how graduate earnings after tax, NI and Student Loan repayments compare to the national average for 25-29-year-olds. Challenge for the weekend?”

    I’ve occasionally mused on how to go about comparing this for graduates from the pre 1998, 1998-2005, 2006-11 and 2012+ cohorts, using the tax and NI rates and thresholds pertaining at the time each graduated, to establish which were ‘better off’ after all deductions. There is a website somewhere with the income tax and NI time machine required, though I’m not sure how reliable it is.

Leave a Reply