One of the most important dilemmas young people face if they’ve chosen to pursue Higher Education is which subject to study. Should they study something they’re passionate about, or the subject with higher chance of good graduate employment and higher income?

For young people graduating into the aftermath of the recession and experiencing more precarious employment, lower income in real-terms and little chance of getting on the housing ladder, the latter may seem more important. More young people go to university now than ever before, making the graduate job market particularly competitive, which could make employers place more emphasis on factors that could differentiate graduates, including the subject they choose.

What do we know about the returns of different subjects?

Whilst most subjects offer a broad range of skills that are transferable across many sectors, many industries require graduates with specific expertise. The subjects that appear most in demand in industry tend to be Science, Technology, Maths and Engineering (STEM) subjects, with many employers and organisations worried about the lack of suitable graduates. This suggests studying these subjects would put students in an ideal position when searching for jobs.

But how much do we really know about the occupations outcomes of young people after studying at university? Evidence from the Labour Force Survey, a representative study of the occupational circumstances of people in the UK, suggests that STEM graduates do earn more than their peers who graduated in other subjects, as do students who studied Law, Economics, or Management (LEM). It is possible that returns to particular subjects have changed over the years, and returns for the most recent graduates may be very different. Some new evidence looking at early career destinations (1994-2010) six-months after graduation questions the idea that STEM graduates really do so much better than other graduates.

The release of the most recent wave of Next Steps (previously known as Longitudinal Study of Young People in England), collected when participants were 25, has enabled a renewed look at the destinations of ‘millennial’ graduates born between 1989 and 1990. These young people would have graduated into a job market that is more competitive both due to the recent recession and stagnation of wages, and because of increased university attendance overall. Next Steps is a longitudinal study that captured information about young peoples’ educational trajectories, personal and family characteristics and current occupational outcomes. This is an important age to test the emergence of differences in outcomes; young people are beginning to find their feet professionally, and most young people had not started a family or had children yet.

Students were asked if they were in work, and if so what their job and income were. Jobs were coded using the National Statistics Socio-Economic Classification (NS-SEC), a common measure of social class. They were also asked how many hours they worked each week (with 35 or above classed as full-time), whether their qualification was a formal requirement or gave them an advantage in gaining their current job (classed as graduate employment). Finally, they were asked a number of questions about their mental and physical health, one of which included the General Health Questionnaire (GHQ) measuring mental wellbeing.

Figure 1: Work and personal characteristics by subject group studied

Table one compares graduate outcomes across a range of indicators by subject group studied. STEM graduates in work at 25 are most likely to be in a professional or managerial job, and to be in a graduate job. STEM graduates, along with law, economics and management graduates, are also most likely to work full-time. Thus, overall they have relative positive occupational outcomes compared to their peers who studied social sciences or arts and humanities.

Possibly in line with their better occupational outcomes, STEM graduates also have the lowest rates of poor mental wellbeing compared to their peers. Whilst recent research suggests that around a quarter of 25-year-olds experience symptoms of poor mental health, only one in five STEM graduates do, compared to almost 30 per cent of social science and arts and humanities graduates.

Figure 2: Mean income by subject group studied

In contradiction to findings for other occupational outcomes, arts and humanities graduates appear to have the highest weekly incomes at 25. This is despite the fact that they are less likely to be working in professional graduate jobs or to be working full time.

Who you are matters more than what you do?

It is possible that the higher pay packets of arts and humanities graduates reflect the types of students who choose to study this group of subjects, rather than the labour market returns of the subjects themselves. In England students are segregated in field of study at university by their gender, family background and ethnicity, and these characteristics are all independently associated with income. Students who study Arts and Humanities are more likely to be female, to come from relatively advantaged social backgrounds, and to be white.

In the current cohort of 25-year-olds, women with a degree earned slightly more than men. Very early in their career young women in the most recent generations appear to earn more than men, however this trend is reversed for people over 30. It is impossible to know now whether this shows a fundamental shift in the gender pay gap, or whether men in this cohort will overtake women in their 30s. Students whose parents have lower education levels also earn less per week then their peers[1], reflecting research showing that people from working class backgrounds earn less even when they reach the higher rungs of the professional ladder. However, the starkest differences in income are by ethnicity. All ethnic minority graduates in the current cohort earned less than white graduates despite being more likely to study subjects that traditionally have higher returns on graduation.

Figure 3: Mean income by gender, family background and ethnicity

Regression analysis was used to help tease out the effect of differences in student characteristics on differences in income, and likelihood of entering professional occupations, by subject studied. Table 1 shows what happens to the income disparities when controlling for different family background characteristics. Arts and Humanities graduates earn around 6 per cent more than STEM graduates just after university, and these differences remain statistically significant when controlling for gender and family background. When controlling for ethnicity however the difference reduced substantially and is no longer statistically significant. Thus, the relative disadvantage experienced by ethnic minority graduates in the labour market is skewing average earnings, making it appear that arts and humanities subjects offer higher returns than they really do.

Table 1: Income differences by subject with and without personal characteristics controlled

In contrast, table 2 shows STEM graduates have a clear advantage in gaining employment in professional or managerial occupations that persists when controlling for other characteristics associated with studying STEM. Graduates from all other groups of subjects are over 10 percentage points less likely to be working in a professional or managerial occupation compared to STEM graduates

Table 2: Propensity to be working in a professional or managerial occupation by subject studied at university compared to STEM graduates, with and without personal characteristics controlled.

These findings raise a number of important questions for researchers and policy makers.

Why are income returns by subject area not in line with other occupational outcomes? Whilst STEM graduates seem to have an advantage in gaining professional graduate employment and in enhanced mental wellbeing this isn’t reflected by increased incomes. Is it possible that the advantage conferred by studying particular subjects is realised through occupational types first? Whilst (when controlling for ethnicity) incomes are similar across subjects at age 25, incomes of graduates in professional and managerial occupations may grow at a much faster rate.

The fact that ethnic minority graduates have such lower incomes compared to white graduates, despite studying subjects typically thought of as high return and beneficial in the labour market is another key concern. To what extent is this the consequence of discrimination in the labour market, and how could this be tackled? Many employers have increased awareness of the effects of both explicit and implicit biases, with name-blind applications being increasingly used. With the changing nature of employment, and the rise of the ‘gig economy’ in which workers have few employment rights, and in which the young are over-represented, how far will these well-meaning strides make a difference?

[1] Parents’ social class showed a similar pattern, and is included in regressions as a measure of ‘family background,’ but its not included in figure 3 to save space.

Note: All analysis was weighted to take account of attrition over time and over-sampling of people from ethnic minority backgrounds.



  • The post gives the views of its author, not the position of LSE Business Review or the London School of Economics.
  • Featured image credit: Graduation, by Kivensilence, under a CC0 licence
  • When you leave a comment, you’re agreeing to our Comment Policy.

Natasha Codiroli Mcmaster is a PhD student based at the Institute of Education, UCL. Her research focuses on inequalities in education, and her most recent work considered disparities in students’ subject choices. She has also worked as an analyst at the Department for Work and Pensions on projects aimed at understanding the drivers and consequences of disadvantage.