Unemployment among computer science graduates

The landscape

During the late 1990s undergraduate entrants studying full-time on computer science courses at English higher education institutions increased rapidly peaking at around 29,000 in 2002-03. Over the subsequent four years entrant numbers quickly declined to around 17,500 by 2006-07 before returning to more modest rates of growth. The impact of the 2012-13 higher education fees and funding regime changes seems to have been particularly pronounced for computer science. Numbers recovered somewhat in 2013-14, but remain 9 per cent lower than in 2010-11.

In each of the last six years, more students have begun computer science courses than physics, chemistry and maths combined. So there was obvious concern when it was noted that of undergraduates who qualify across all higher education subjects, computer science has consistently had the highest rate of unemployed graduates (see for example HESA’s statistical first release or HEFCE’s interactive data tool).


Proportion of full-time first degree qualifiers assumed to be unemployed


Computer science is taught at the large majority of universities – 90 of 130 English higher education institutions report students studying computer science. Most other STEM subjects are concentrated in a smaller number of institutions. Institutions teaching computer science display substantial differences in the proportion of their graduates who appear to be unemployed (see the Unistats course assistant to view the employment and accreditation data for computer science courses).


Graph 2


* Full-time first degree qualifiers who provide a valid response to the Destination of Leavers from Higher Education (DLHE) survey. Based on qualifiers in academic years 2011-12, 2012-13 and 2013-14. Institutional grouping is defined by the average tariff score of an institution’s young (under 21) UK-domiciled undergraduate entrants holding level 3 qualifications which are subject to the UCAS Tariff.


And all of this combines with wider concerns. There are concerns about the proportion of undergraduate computer science students who progress into low-paid or non-graduate level employment. There are concerns over the reliance our computer science departments have on international recruitment to fill their labs and postgraduate courses. And there are concerns from industry about the skills, agility and work-readiness of those computer science graduates flowing into the workforce.

Time to reflect and review

In this landscape, it seems right to reflect and review. The Government has asked me to look at these issues through the work of an independent review. Our review is also tasked with generating recommendations about how to change accreditation so that it keeps pace with the needs of the profession. It is due to report towards the end of November.**

In this blog I wanted to present some of the basic data the review will draw on and solicit help to make sense of the data and even extend it. The review will need to take a closer look at the evidence, and the various ways to interpret it. Some of this will draw on publicly available data, but we welcome responses to the review in the light of other data and evidence.

The existing base of evidence is already considerable. HEFCE’s recent publication on the destination of full-time students on first degrees provides more granular data than ever before. The most recent in Summer 2015 includes an examination of employment rates of graduates 6 and 40 months after they left higher education.

HEFCE’s interactive data tools allow anyone to delve into a number of largely unexplored characteristics of graduates’ employment. Here I am looking specifically at those qualifying from computer science. We explore one or two of the findings that the review team thought were particularly interesting. This is not a comprehensive analysis. Others may well identify other and different traits that could be of interest. But it is food for thought.

A question of background

Many will know already that computer science has a distinctive profile of students. Notably it has more men, students from black and minority ethnic backgrounds, and students with lower previous levels of attainment.

The statistics show that unemployment among black and minority ethnic graduates from full-time, first degrees is six percentage points higher than among white graduates. But interestingly this difference is smaller than it is for graduates from comparable subjects like electronic and electrical engineering or mathematical sciences.


Graph 3

* Full-time first degree qualifiers whose ethnic background in known and who provide a valid response to the Destination of Leavers from Higher Education (DLHE) survey. Based on qualifiers in academic years 2011-12, 2012-13 and 2013-14.


The data also allows us to look at ‘disadvantaged’ students – defined here as students who come from neighbourhoods with low levels of participation in higher education.

Computer science seems to be doing at least as well as other STEM subjects. The level of unemployment among full-time graduates from computer science courses who come from disadvantaged backgrounds is only one percentage point higher than for the majority of STEM subjects.

And from a different angle – gender – the differences disappear. The unemployment rate for men and women graduating from computer science is the same.

Salary and sector

What about those students who are employed? Does this show us anything distinctive about the profile of computer science graduates?

Around 65 per cent of computer science qualifiers who enter full-time paid employment within six months of leaving HE tell us that they earn less than £25,000. This compares with 55 per cent of equivalent qualifiers from electronic and electrical engineering courses, and with 60 per cent of those qualifying from mathematical sciences.


Graph 4

* Full-time first degree qualifiers who provide a valid response to the Destination of Leavers from Higher Education (DLHE) survey; report that they enter full-time paid employment; and provide salary information. Based on qualifiers in academic years 2011-12 and 2012-13.


When we look at the sectors of industry in which graduates are employed, a high proportion (38 per cent) work in the information and communication industry. The rest are dispersed across a wide range of industries; this is unusual among many STEM disciplines.


Graph 5

* Full-time first degree qualifiers who provide a valid response to the Destination of Leavers from Higher Education (DLHE) survey and report that they enter employment. Based on qualifiers in academic years 2011-12 and 2012-13.


Computer science graduates show some differences from their peers in other STEM subjects in terms of their likelihood of securing employment in a professional or managerial role.

Among those who gain employment in a micro, small- or medium-sized business, around 86 per cent occupy a professional or managerial role within that business. For those in the largest firms, 68 per cent were in these types of roles. This means that a computer science graduate gaining employment in a larger firm is more likely to work in an administrative, trade, service, sales, plant operative or elementary role than their counterpart who is working in a smaller firm.

The story is similar for those qualifying from electronic and electrical engineering, though the differences between smaller and larger firms are less.

But for maths graduates those employed in large and medium firms were the ones most likely to be in the professional and managerial roles, with their counterparts in smaller firms being least likely to be in these types of roles.


Graph 6

* Full-time first degree qualifiers who provide a valid response to the Destination of Leavers from Higher Education (DLHE) survey and report that they enter employment. Based on qualifiers in academic years 2011-12, 2012-13 and 2013-14.


Now over to you

Now over to you. We’re keen to hear your thoughts around computer science graduate employability. Whether accreditation plays any role or could assume a more significant one in the future. Do provide your own comments on the review in the survey. In doing this we hope that readers will take the opportunity to explore our quantitative data themselves, as well as drawing on their own evidence bases.

** The review is now due to be published in March 2016.