Dr Matt Dickson, Reader in Public Policy at the Institute for Policy Research
Every year more than 350,000 students begin Higher Education (HE) degrees in England. With the majority of courses charging the maximum tuition fee of £9,250 per year, the total cost of degrees per cohort is around £17 billion, about two-thirds of which will be paid by the graduates via loan repayments, with the remainder paid by the tax-payer.
HE thus represents a large investment for students and the government and as such it is important for everyone to have accurate information about the relative earnings and employment returns to degrees taken in different subjects and at different institutions. This will allow students to make informed choices about which course to study, and will also allow the government to know which subjects and institutions are associated with the greatest returns, providing a starting point for investigating what the highest performers are getting right.
Fortunately, the newly available Longitudinal Education Outcomes (LEO) dataset developed by the Department for Education (DfE) has allowed the most accurate information to date on earnings and employment outcomes post-graduation to be analysed. The advantage of LEO is that it uses administrative data to track students through school, college, HE and into the labour market – linking up the information on individuals’ background, attainments throughout education and ultimately their employment and earnings outcomes.
This means that outcomes for graduates of different courses across England can be compared, taking into account the differing background characteristics and prior attainment of those who study different subjects at different institutions – essentially allowing comparison of returns for students with the same prior attainment and socio-economic background but studying different courses.
The DfE recently published the first comprehensive analysis of the LEO data and the findings show that the labour market returns to different degrees vary considerably even after taking account of the differences in the types of students who take each course. Both the subject studied and the institution attended have a large impact on graduate earnings but there is also considerable variation within subject across different institutions and within institution for different subjects – so neither subject nor institution is a cast-iron guarantee of high earnings. For example, the best business studies degrees have earnings returns of over 50% more than the average degree, whereas the worst business degrees give returns of less than the average degree. As we would expect, attending Oxford or Cambridge is associated with higher than average earnings for almost all subjects but there are one or two with lower than average returns, and this finding is common across the board, even amongst the elite institutions.
Given this considerable variation in earnings returns, and the fact that both subject and institution are important, the Government has set up a competition to create a free, accessible, and innovative digital tool with the aim of improving prospective students’ access to valuable data on graduate outcomes – allowing them to make better informed decisions about where and what to study.
So what should this app look like?
Accessing information derived from administrative data on the earnings of graduates from different subjects taken at different institutions, 3, 5, or 7 years post-graduation is a big improvement on the information available to students at present. However, it has to be borne in mind that not all courses will be available to all students – entry requirements will dictate the set of courses available given the predicted grades of any student. So for the app to be useful it will need to take into account the A-levels the student is taking, the grades predicted and what this means for the courses available – using information on the entry requirements for each course, already available from Unistats. Then from within the possible courses open to the student, those with the higher and lower earnings and employment prospects could be easily identified. Looking from the other direction, for students finishing GCSEs and choosing A-levels it would be useful to be able to open the app, choose a subject and see the A-levels taken and average attainment of those pursuing that degree at each institution. Seeing the earnings and employment outcomes for each potential course would also help decision-making at this even earlier stage.
But it is not only earnings and employment information that that app could provide access to. Already available for each course in the country is the information from the National Student Survey covering many aspects of each course – for example, satisfaction with course’s teaching, learning opportunities, assessment and feedback, academic support, organisation, resources and student voice. From Unistats there is information on the drop-out rates from courses, and the breakdown of final degree outcomes that could be integrated into the app. In addition, there is also information from the Destination of Leavers from Higher Education survey on the types of jobs graduates are doing six-months after graduation, whether professional/managerial or not, and the type of employer, industry and occupation.
What can’t the app do?
All of this information relating to different subjects taken at different institutions, available in one place, would be a hugely valuable tool for all prospective students. It would bring together the newly available and accurate earnings and employment data that LEO provides, with other types of information on the characteristics and outcomes of each course, providing a comprehensive picture that would be extremely useful in informing choice.
However, there are very important caveats to bear in mind with any app of this sort. First and foremost, the information on graduate outcomes derived from LEO and other sources tells us something about the pathways of those individuals who chose to study these particular courses. The choice of subject and institution is an inherently non-random one and while it would be possible for the app to take account of the entry requirements for each course, there may be preferences and other difficult to measure skills and characteristics that influence both the choice of course and later outcomes. We will never see identical people in every institution or studying every subject and the impact of a specific course may be different for different types of people, what we can only ever see is the average outcome of a course for the people who studied it. So the returns for each course can only be a guideline and may not be applicable across the board.
Similarly, the earnings and employment information in LEO captures outcomes for graduates who were studying 5 to 10 years ago and have been in the labour market in recent years. Outcomes of different graduates will reflect the demand for different skills in the labour market, as well as the impact of HE provision, and as the economy evolves and changes, skill demands will change and the financial returns to different skills will also change, impacting the returns to different degree subjects.
Even more importantly perhaps are the non-pecuniary benefits of different degree courses. While we can see the jobs that people end up in, it is also important to know how people feel about their job – do they enjoy it, are there opportunities for self-accomplishment, expression and social interaction? All of these things are factors that we know affect mental health, wellbeing and overall happiness and are equally part of the return to different degree courses. Moreover, the experience at university will differ depending on the institutions themselves – they differ in important aspects that aren’t reflected in numbers: they are in different geographical locations and provide different recreational activities, landscapes and physical environments, appealing to different tastes and preferences. The type of university – older versus newer, campus or not – and where in the country it is, will play a significant part in determining the university experience, the people you meet, what you do and where you head when you leave – all important factors and all things that an app will struggle to capture.
The overall picture
So overall the idea of an app that brings together a lot of relevant information on the labour market outcomes associated with studying different subjects at different institutions is a good one. Adding in information about the student experience of each course will allow students with different tastes and priorities to find the right course to meet their aims. But there are big caveats around what such a digital tool can and can’t do and keeping these in mind will be crucial for prospective students making big decisions about their future.
A version of this blog first appeared in Research Fortnight.
Matt Dickson is a Reader in Public Policy at the Institute for Policy Research, University of Bath, and an author of the DfE’s report ‘The relative labour market returns to different degrees’ which utilises the LEO data.