Evidence should replace dogma in the debate about the STEM shortage

A guest blog kindly contributed by Dean Machin, Strategic Policy Adviser at the University of Portsmouth

David Hume, it is said, awoke Immanuel Kant from his dogmatic slumbers. We might all aspire to have similar effects but most of us will fail. However, Professor Emma Smith’s and Dr Patrick White’s Nuffield Foundation study on the STEM shortage might just be an exception.

There is a widespread view that this country does not produce enough STEM graduates, particularly engineers. It is also believed (by some at least) that we produce too many of the wrong type of graduates, some of whom should be ‘diverted’ into technical education.

While others disagree, there is a consensus that the debate lacks robust evidence. Enter Smith’s and White’s study. The study utilises five datasets: UCAS, DLHE, the Annual Population Survey, the 1958 National Child Development Study and the 1970 British Birth Cohort Study.

Smith and White found that there is no overall shortage of STEM graduates and that there is as much variation in career prospects within STEM subjects as between STEM and non-STEM subjects. They also found that the majority of STEM graduates never work in highly-skilled STEM jobs and that most professional STEM workers do not have degrees. They conclude: “encouraging students to study STEM degrees on the basis of better labour market outcomes is ethically questionable”.

Heady stuff but how do Smith and White reconcile their analysis with survey data about a shortage? There are enough STEM candidates out there but employers “may have to modify their views on exactly what constitutes a valuable or desirable employee and to what extent it is their responsibility to train their workers.”

In other words, there may be enough STEM graduates available, just not with the particular qualifications employers in shortage areas demand for their immediate needs. Like Kant, pre-Hume, some employers might be dogmatic about their requirements.

Of course, simply telling people they are wrong is not the best way to change their minds (Prime Minister, are you listening?) so one strategy would be to produce more people who do fit the employers’ mould. However, as David Willetts argues in A University Education, by age 16 only 4% of young people are studying the right combination of A levels to study engineering at university.

Indeed, Smith and White note that while undergraduate recruitment more than tripled from 1986 to 2012, numbers recruited to electronic and electrical engineering degrees have increased by less than 8%. Reforming pre-18 education so that more people have the fundamental knowledge to be able to choose STEM subjects at university, particular engineering, would be a start.

In the short-term this won’t happen, and universities have a role, so another option is for universities to open their doors wider. The University of Portsmouth runs a successful foundation year in Engineering. Our successes include Timothy Lee-Lewis who arrived as a mature student with O Levels only. After getting a first in MEng Civil Engineering he is now studying for a Ph.D at the University.

Another current MEng Civil Engineering student arrived with A Levels at C, D and E none of which were relevant to engineering. He is expected to get a distinction and impressed his employer on his sandwich placement year.

The placement year is key. Without what is effectively a year-long interview, it is too easy for employers to look at poor A level results, as well as university type and reject candidates. (Smith and White found that employment prospects depend significantly on the kind of university from which you graduate.)

Neither of these suggestions challenge employers’ recruitment practices so here is one that is particularly pertinent to engineering: employers could recruit other STEM graduates and use the Apprenticeship Levy to retrain them. Of course, employers need not recruit only STEM graduates but this is where they should start. As Smith and White note, it ‘is difficult to believe that maths or physics graduates … do not have knowledge and skills that are transferable to engineering’.

Other issues remain. Why, for instance, do highly-skilled STEM professions lose people over time and how do we address engineering’s women problem? (Smith’s and White’s study did not look at gender, although they are currently working on a study that does.) However, I will conclude with a broader question. What if Smith and White are wrong? Despite Kant’s praise for Hume’s empiricism, he did reject it.

Smith and White’s study is far from the final word on this topic. But it is more comprehensive than most other studies and infinitely more sophisticated than most public commentary and debate. Soon the Augar Review will report. It will almost certainly include recommendations on the STEM ‘shortage’. The very least we can expect of it is some evidence that this impressive Nuffield study has been read and its implications considered.**

** = All quotes are taken from ‘Where Do All the STEM Graduates Go? Higher Education, the Labour Market and Career Trajectories in the UK’ Journal of Science Education and Technology (February 2019) 28:26–40 However, a publicly accessible version of the report entitled ‘The employment trajectories of Science Technology Engineering and Mathematics graduates’ (February 2019) is available here.