The Scarring Effect of Graduate Underemployment: Evidence from the UK

The UK has one of the highest proportions of tertiary educated workers in Europe but also one of the highest rates of graduate underemployment. Little is known however about the extent to which there is a scarring effect of early graduate underemployment on future labour market outcomes. In this paper, we examine the effect of early underemployment using data on 67,000 graduates from undergraduate degrees in the UK in 2013. Labour market outcomes at six-months and 42-months post-graduation are linked to administrative records covering higher education, prior attainment, demographics and family background. We find that compared to being in a graduate job six-months post-graduation, early experience of underemployment increases the probability of being underemployed three years later by 0.24. Oster bounds analysis suggests that the causal effect of early underemployment on later underemployment is at least +0.18. This is a large effect relative to the base risk of underemployment at 42-months for those in a graduate job at six-months which is just 0.09. We highlight important implications of these findings, with arguments from both equity and efficiency for policies to help graduates to attain graduate level jobs.