In many of the world¡¯s major higher education systems, the growth in the number of doctorates being awarded shows little sign of abating. The latest data on UK universities is no exception: the number of PhD graduates they produced increased by 19 per cent between 2013-14 and 2017-18.

But, given the well-reported challenges around getting an academic job after completing a doctorate, what is driving this trend in the UK and around the globe? Looking at the data by subject offers some clues because, while the growth in the UK in disciplines classed as non-science ¨C including social science subjects such as economics ¨C?was 14 per cent, for science subjects it was 22 per cent.

For some individual non-science subjects growth was even lower: PhDs in languages and education both had growth of less than 3 per cent, while social science (the most popular PhD subject in the non-science bracket) rose by 11 per cent.

History and philosophy PhDs were up 22 per cent, but the leaps for some pure science subject areas were often much bigger: biological science PhDs were up 25 per cent, computer science?climbed 34 per cent and maths rose 37 per cent.

³ÉÈËVRÊÓƵ

Overall, the share of UK PhDs in non-science subjects fell from 36.2 per cent in 2014-15 to 34.3 per cent in 2017-18 and, looking back further, it has fallen from a peak of 36.7 per cent in 2010-11.

Does this reflect in any way the seeming focus on employability in higher education and, in particular, the view expressed by some politicians and commentators?that science leads to a better-paid career?

³ÉÈËVRÊÓƵ

One approach to answering this question is to look at PhD trends in other nations where there has arguably been a similar narrative over the past decade.

In the US, rich data on the subject of PhD completions ¨C thanks to the comprehensive ¨C?demonstrates that the growth of science PhDs is a very long-term trend.

The proportion of doctorates in the sciences . Although this classification included social science and psychology, this is the only ¡°science¡± area to have seen a fall since 1998, from 17.3 per cent of PhDs to 16.6 per cent in 2017.

At the same time, the absolute number of PhDs classed by the?National Science Foundation as non-science ¨C split into education, humanities and arts, and other fields ¨C was actually slightly lower in 2017 than in 1977.

Robert Townsend, director of the Washington office of the American Academy of Arts and Sciences, who oversees a project gathering data on the humanities, said a big part of this ¡°is just a growing demand for science PhDs outside of academia¡±.

Humanities and social science doctorates, on the other hand, are still much more closely linked to academic jobs, he said, despite attempts to improve the prospects for such PhD graduates outside higher education.

¡°Statistically, [humanities and social science PhD graduates] do well when they get jobs outside of academia but the route is still not as clear or as well lit as it is for a lot of science and engineering PhDs,¡± he said.

The result is that the availability of jobs in academia has been seen to affect the number of PhDs in the humanities and social sciences, although there is a considerable lag?because of?the time it takes to earn a doctorate.

³ÉÈËVRÊÓƵ

This cyclical downturn is beginning to bite again now in the US as a tightening in the academic job market a few years ago starts to feed through into PhD numbers, Dr?Townsend said.

Could this explain what is going on in the UK, too?

Giulio Marini, a researcher at the UK¡¯s Centre for Global ³ÉÈËVRÊÓƵ who has that influence the salaries of PhD holders in the humanities and social sciences, said that across all disciplines there was a growing pressure on postdoctoral opportunities in academia.

³ÉÈËVRÊÓƵ

This was?because of the growth in PhDs outnumbering academic retirees and the increasing ¡°process of saturation of PhD holders among staff in post-92 institutions¡± who may have been less likely to have doctorates in the past.

But, Dr Marini added, he could not see compelling evidence that fewer people were choosing to pursue a PhD in the humanities and social sciences as a result of this or anything else. Instead, the data appeared to point to slower growth in these areas at PhD level?on account of undergraduate choices.

¡°Overall, I feel that people who want to endeavour in a PhD do not select sciences more than social sciences and humanities in comparison to the past,¡± Dr Marini said. But, at undergraduate level,?¡°my two-pence guess would be that students look for higher probabilities of employability and they select disciplines accordingly¡±.

So could the employability and pro-science line seemingly toed by some politicians be affecting PhD numbers in this way? Dr Townsend said that, in the US, this was a possibility.

¡°I am not so sure that it is having an effect on [PhD numbers], except that it is channelling people away from undergraduate degrees in the humanities and social sciences and towards STEM degrees,¡± he said. This, he pointed out, might also ¡°create fewer [academic] job openings for humanities and social science PhDs because there are fewer students studying in those disciplines¡±, potentially creating a vicious cycle.

Harriet Barnes, head of policy for higher education and skills at the British Academy, said that, in the UK, ¡°pipeline¡± issues at undergraduate level could be an explanation for slower PhD growth in disciplines?such as languages. But she also pointed to other factors at play. For example, some subjects, such as?education, were ¡°quite small areas at undergraduate level¡­so to have people doing PhDs you need to attract them in from other social sciences¡±.

The longer-term drift away from the humanities also reflected a?¡°rebalancing¡± of the shape of the knowledge base as science has grown as an area of academic research, she said.

Another major factor in the huge growth of science PhDs in English-speaking higher education systems has been the demand from international students.?A country where this may have played a big part is Australia, where there has been staggering growth in the number of research doctorates awarded in recent years.

In engineering, for example, there was an 80 per cent rise in the number of research doctorates from 2011 to 2017. Separate figures on enrolments in higher research degrees (including master¡¯s) show that overseas student numbers went from 940 in 2001 to almost 5,300 in 2017.

Gwilym Croucher, senior lecturer in the University of Melbourne¡¯s Centre for the Study of ³ÉÈËVRÊÓƵ, said the propensity for overseas students to study technical disciplines was one possible explanation for science PhD growth being so high.?

But he also pointed out that student numbers in the humanities, arts and social sciences had doubled in the past two decades in Australia. PhD completion data also show that although it had the lowest growth among subject areas, society and culture doctoral awards were still up 23 per cent from 2011 to 2017.

The sustainability of this growth may well depend on job opportunities outside academia, but Dr Croucher said Australian universities were ¡°putting considerable effort (and money) in to ensuring that their PhD programmes produce qualified researchers who can leverage their skills outside academia¡±.

¡°No one is pretending that there is not more work to be done at a programme level, but surveys of PhD graduates show pretty strongly that it is valuable in many different workplaces.¡±

³ÉÈËVRÊÓƵ

It shows that the key to sustaining continued PhD growth in any system may be the ability for higher education to demonstrate a doctorate's?worth in all disciplines, and not just science.

simon.baker@timeshighereducation.com