Postgraduate study is not a passport to high earnings

A new report from the IFS looks at the salary benefit from postgraduate study, and sees low pay in the education sector. David Kernohan makes sense of it all.

It’s been a while coming, but today’s IFS research on returns to postgraduate study is very much a companion to IFS’ 2018 research on early-career undergraduate earnings.

The central finding is that although postgraduate study makes sense in terms of insurance (you are less likely to experience unemployment or low pay after postgraduate study) it is not a path to higher earnings. Indeed many, especially men, earn less than members of their cohort who did not study beyond undergraduate level.

Originally in the grid for April, a lot has changed since. How much looking at historic earnings and careers can tell you about the future of those currently or about to study particular subjects at particular providers has always been in question – but with the future so much in doubt these concerns become more apparent.

If you like – this examination of the return to postgraduate study is primarily of academic interest.

The shape of the space

The number of postgraduate students has doubled in 20 years, a matter that we should bear in mind as we examine the life courses of those who started postgraduate courses around 20 years ago. There are gender-based differentials – men are more likely to do PhDs and masters degrees, women are more likely to do a PGCE (postgraduate certificate of education, used as an entry route to teaching) – and although masters courses closely reflect the diversity of undergraduate study PhDs and PGCEs skew white.

The nature of postgraduate study means that we are dealing with a higher proportion of undergraduates who secured a first or upper second class degree than in the general population. This group, and thus postgraduate students, are generally more likely to come from a privileged background – and have benefited from better school-level performance resulting in higher earlier attainment. A subsidiary IFS briefing report examines access to postgraduate study based on postgraduate background – this table looks specifically at the 2002 cohort.

%UGRussell Group UGAny PGPGCEMastersPhDMasters or PhD
SES Q118.11.93.81.22.90.22.9
SES Q2273.66.31.94.90.55
SES Q335.469.22.77.20.97.4
SES Q444.49.812.83.510.11.410.4
SES Q558.417.718.14.414.72.115.2
Private81.543.927.44.6243.124.5
Population39.710.211.22.99.11.29.3

Seventy percent of masters study, and ninety percent of PhD study, takes place within the 24 providers in the Russell Group. It skews sharply towards STEM (Science, Technology, Engineering, Maths) and LEM (Law, Economics, Management) subjects – with three-quarters of all masters and PhD study in these areas. For PhDs, a quarter are in just two subject areas – biosciences and chemistry.

Earnings from learning?

After controlling for learner characteristics, Women with a masters qualification enjoy on average a 2 per cent salary premium at age 35, with men experiencing a 2 per cent detriment. For PhD study the difference is even greater – women earn (on average) 8 per cent more, men earn (again, on average) 9 per cent less.

There are sharp differences by subject – which also explain the majority of provider level differences.

[Full screen]

[Full screen]

The research looks at subject differences between undergraduate and postgraduate study – if you have a first degree in a “low earning” subject, postgraduate study in a “high earning” subject will be your best bet for a salary premium. Conversely, those already in “high earning” subjects will do better by continuing their study in their undergraduate subject.

[Full screen]

There is a marked difference between the likelihood of a person opting for particular types of postgraduate study by undergraduate subject of study.

[Full screen]

Insurance choices

For me, what stands out is the utility of PG study for the avoidance of salary and employment detriment – an insurance model of postgraduate study. If you have a postgraduate qualification (other than if you are a man with a masters degree) you are less likely to experience unemployment and less likely to experience low pay, on average. As IFS puts it:

postgraduate degrees appear to offer insurance against bad labour market outcomes. This is particularly true for PGCE qualifications, which significantly increase the chances of having ‘good’ earnings by age 35, but simultaneously reduce the chances of achieving moderately high earnings (for example, above £50,000). This finding is also true for PhDs and, to a lesser extent, for masters degrees”

For those of us of a certain age, our understanding of this phenomenon is taken from Milton Bradely’s “The Game of Life”. If you chose the university rather than business route you could enter a highly paid profession (law, medicine) at a large salary premium, or a simple “graduate” salary at a level slightly above the starting non-graduate salary. However, those entering the non-graduate route have the chance of higher salaries later in life (if they land on the appropriate square, of course). Replace “non-graduate” with “graduate” and “graduate” with “postgraduate” and it is the same thing – rumours that the next edition of the Game of Life will include data from LEO are likely to be untrue.

What does this mean?

Fundamentally the findings are simple, depressing, and familiar. Pay for teachers is low and pay for academics is low, though in both cases pay is not below the national average But for the small number of professions that require postgraduate study, pay is high. We could conclude that further study is valued most outside of the education sector.

This is LEO data, so the usual caveats apply. Salaries for part-time work are treated as a full time salary – to the detriment of the quality of statistics on women. It is possible (though we are not told) that women with postgraduate qualifications are less likely to be working part time than their peers with a first degree as a higher qualification, which would explain part of the salary differential.

One of my major criticisms of the undergraduate data version of this report was that it included data on those who started but did not complete an undergraduate course – this postgraduate version deals with completers only. We’re looking here at people who started their undergraduate study before age 21, graduated between the mid 90s and mid 00s, and completed a postgraduate qualification by age 30. In the main, we are dealing with earnings at age 35 and the way these evolved through a persons’ late 30s. On this timeline, the 2008 financial crisis would have hit around age 30.

The current circumstances are not kind to historic data. This report does advance our understanding of the postgraduate market, in that we see two main trends – professional courses and in-demand courses linked to the finance industry have a significant salary impact (we’d see more of these if we looked at completes beyond age 30), but the bulk of courses (taken by people out of interest in the subject, seem to lead to academic and related jobs that are unlikely to generate significant pay before the age of 35.

*Bonus graph*

As above, the variation by provider for postgraduate earnings is primarily linked to the subject mix and prior attainment. On these groups, the similarity of two institutions in these terms is an important component of making a comparison. Not all providers are shown, as not all had a significant number of postgraduates in these cohorts. This isn’t really a useful graph, but I thought you might enjoy it.

[Full screen]

One response to “Postgraduate study is not a passport to high earnings

  1. A really interesting! One thing I would like to know is whether the mode of study had any impact on median earnings. I imagine someone studying an MA full-time would lose out on potential earnings by being out of the workplace for a year compared to someone studying part-time while in work. Yes but does it correlate? 🙂

Leave a Reply