Heterogeneous returns over the life-cycle? Or nothing at all? Re-examining the returns to education in the UK Franz Buscha 1 and Matt Dickson 2 1 Westminster Business School, University of Westminster, 35 Marylebone Road, London, NW1 5LS, United Kingdom 2 University of Bath, Bath, BA2 7AY, United Kingdom; CMPO, University of Bristol, United Kingdom; IZA, Bonn, Germany Abstract This paper uses data from the UK Labour Force Surveys 1986-2011, the New Earnings Survey and the Annual Survey of Hours and Earnings 1975-2011 to re-examine the returns to education from the 1972 Raising of the School Leaving Age (RoSLA) reform. Importantly, the span of our data allows us to investigate whether the labour market returns from an additional year of schooling vary over the lifecycle. Keywords: Returns to education, life-cycle, earnings, employment JEL classification: I21, I28, J24 1. Introduction A recent analysis of newly released UK household data 3 suggests that the impact of the 1972 RoSLA reform had a long-run and permanent impact on hourly wages (Buscha and Dickson, 2012) for both men and women. This evidence shows that individuals who were affected by the school leaving age reform in 1972 earned on average 5% more in their early to mid-50s than similar individuals who were not affected by the ‘additional year of schooling’. Whilst this is good news for policy makers and believers in human capital theory – especially in light of the forthcoming raising of the participation ages to 17 and up to 18 in 2013 and 2015 respectively – these results are somewhat at odds with recent evidence on the returns to education in the UK. In fact, recent re-appraisals of the causal return to education (Devereux and Hart, 2010; Grenet, 2012 forthcoming) using more sophisticated identification techniques and larger datasets seem to suggest that the effect of an additional year of schooling is less than half of what was commonly estimated in studies from the 1990s and 2000s. 4 The UK returns literature is thus in a state where ‘causal’ returns to schooling apparently vary between 3% and 20%. To rationalise this emerging disparity in the returns to education Buscha and Dickson (2012) conjecture that life-cycle effects may play an important role in reconciling such differences. In other words, the returns to education could be non-constant over the life-cycle 1 Tel: +44 (0)20 7911 5000 x66596 E-Mail address: buschaf@wmin.ac.uk 2 Tel: +44(0)1225 386736 E-Mail address: m.dickson@bath.ac.uk 3 Understanding Society, Wave 1 (2011) 4 For example, see Harmon and Walker, 1995; Harmon and Walker, 1999; Chevalier and Walker, 2002.