Are graduates overeducated and underpaid?

Though there is a wage premium attached to higher education, many graduates do not work in roles that make full use of their qualifications.

These “overeducated” employees (some 16% of UK workers, and 34.2% of those who have graduated since 2007) may earn more than their non-graduate colleagues, and may see speedier progression within the roles. But they do not earn as much as others in their graduating cohort with “matched” roles (roles that make full use of the qualifications they have).

Such is the contention of the Office for National Statistics (ONS), which has published a detailed study of the issue. In characteristic style the ONS does not draw policy conclusions from the presented data and analysis, though there are clear implications for the ongoing debate around student numbers.

A question of definitions

Wonkhe readers will recall with delight Peter Brant’s canter through the delicious vagaries of graduate job classifications. For many applications, the widely framed SOCs (Standard Occupational Classifications) are too broad an instrument for making sense of the graduateness of graduate occupations.

In identifying a graduate job the ONS has looked at the mean of the years of education experienced by those identified in the Annual Population Survey (APS) as being in the job in question. They then flag individuals one standard deviation above this mean as being “over educated”.

What’s important here is that this assumes that current job requirements are reflected in the level of education held by the current workforce. This is a big assumption to make. Consider, for example, a lab technician in a hospital. Historically this was not a graduate role, but for a number of years a degree has been expected – this would make recent entrants look overqualified even though a degree was a requirement. This gets more complex when you note moves to allow non-graduate entry to the lab technician role – apprentices will have less years of education than the graduates have just started, meaning that their cohort will look overqualified in comparison to new entrants as well.

The ONS benchmarks these statistical findings with a US vocabulary of graduate jobs, O*NET, mapped to SOC codes. There’s a serious question as to the validity of this as a UK benchmark given the difference between the two countries. And there’s a serious question about whether it was worth doing at all, given that the ONS’ own figures suggest that only a little over a half (54%) of these reflect the education of those currently working in the UK.

So what does it tell us

Though this is not a precise mapping, and therefore cannot reasonably be used to give an impression of the actual numbers of overqualified workers, the consistent way it has been applied means that we can reasonably look at trends in the data.

For instance, we can look at overqualification by broad subject of study for a range of subjects for both recent (first degree completed within the past five years) and non-recent graduates.

[Full screen]

It’s not a complete range of subjects, and the report rather unhelpfully presents them in two separate charts. But what strikes me is not the presence of arts subjects at the top (Western civilisation still has a massive problem with valuing and supporting artists) but biological sciences in second place. For many years, biological sciences have vied with business to be the most frequent subject of study in UK higher education – and where commentators are often quick to sneer at the value of business courses there is a general consensus that science is valuable.

Part of this may be explained by the recent influx of graduates into traditionally non-graduate skilled roles – nursing in particular has most of the attributes of graduate employment (requires a wide range of knowledge and skills alongside problem solving abilities and empathy) other than pay. Remember current graduate roles that have not historically required a degree will perform poorly on this metric.

The highish rates for law graduates highlight another data issue – the figures disregard those with study above first degree level. Many areas like law, medicine, and architecture require at least some post-graduate study before entry to the workforce.

A long term issue

Arguably the figures for non-recent graduates should be of more concern to us than the figures for recent graduates. It is generally accepted that the first five years after graduation may see some employment below graduate level, as young people juggle career-building with the need to earn to support their existence.

Twenty-nine point two percent of graduates are over-educated for their job role five years or more after graduation. Though we can assign some of these to personal choice – either a focus on non-work goals (for example starting a family), or a commitment to low-paid employment (for example for artists and nurses), – we have to contend with the fact that a sizeable proportion of graduates are not in graduate employment more than five years on, however loosely defined that is.

Graduates in non-graduate roles do enjoy a slight premium over their non-graduate colleagues, and are likely to see speedier progression as they remain in their roles. But this is far from the “graduate premium” so often used as a policy justification for student borrowing.

There will be some who, on reading this report, will leap to blaming the graduates themselves, or the institutions that taught them. A purely instrumentalist view of higher education would suggest that they should never have attended university in the first place. But it is equally valid to argue that our employment market is not adequately rewarding people for the skills they bring to the jobs they do – and that the notion of a “graduate job” does not cover the jobs that we all benefit from having graduates do.