'LEO' refers to Longitudinal Educational Outcomes, an experimental dataset published by the Department for Education that links together education, tax and benefits records to tell us what proportion of graduates are employed and what their median earnings are at different points of time after leaving higher education.
Today's figures allow us to compare variation in employment and earnings according to the subject a graduate studied and by various demographic characteristics (including gender, ethnicity, age, home region and prior attainment). It also reports employment outcomes according to the specific institution that a graduate attended. Previously, such granular information was reported only as a snapshot six months after graduation, so we welcome this much more long-term and detailed data. This richer information could prove immensely helpful to prospective students.
It should also be very useful for universities. The ongoing UUK review of skills is building up a picture of the considerable investments universities are making to support the employability of their students. As they increase this investment in the years ahead this data will help them target their efforts to where it is needed most.
But for the data to truly be helpful, we need to recognise its limitations. For example, due to the way that tax information is collected, most of the earnings and employment figures published today (for all reported tax years apart from 2014–15) exclude graduates who are self-employed. As the publication notes, the exclusion of self-assessment data has a particularly large impact on arts graduates and, by extension, arts-focused institutions, as a larger than average proportion of their graduates are self-employed.
The figures also exclude those who are working abroad. They cannot account for whether a graduate is in full- or part-time work, nor can they account for the region in which a graduate currently works. This means those in well-paid part-time work could appear to be earning very little, and those universities operating in challenging local economic conditions could appear to produce graduates with below average employment outcomes even if their graduates' employment rate is substantially higher than the regional average.
This final point would become even more important were future iterations of LEO to publish graduate earnings by institution. Some institutions may appear to produce graduates with below-national-average salaries, even if those same graduates' earnings are relatively high for the local area.
Also, while LEO could eventually allow us to compare employment and earnings between graduates from different universities and different subjects, it is important that prospective students are able to compare the employment and earnings outcomes for different types of post-secondary education. In other words, what are the median earnings for those with the same or similar gender, ethnicity, region and entry qualifications, but different levels of education (having undertaken an apprenticeship, a degree, or a degree in one subject relative to another)?
And of course, the premium attached to a university education should be considered in a context far wider than just earnings. Some institutions specialise in fields like the arts, charity sector, or nursing and public sector professions – all of which make immense contributions to society and the economy but pay less on average. These graduates may be exceptionally satisfied with their educational choices and careers, despite having lower earnings.
So while today's figures should be welcomed as an innovative and detailed indicator of graduate earnings and employment outcomes, we should bear in mind their omissions and limitations, and avoid conflating earnings with graduate success and satisfaction.