The Labour Market Effects of Academic and Vocational Education over the Life Cycle: Evidence from Two British Cohorts

() - Giorgio Brunello

Journal:
Link::
DOI::
Links::
Tags:: #paper #NCDS #LabourMarket #Attainment #BCS
Cite Key:: [@brunelloLabourMarketEffects]

Abstract

Notes

“Several commentators have argued that the short-term advantage of vocational versus academic education, which is a smoother school-to-work transition, trades off with long-term disadvantages, which are lower employment and/or lower wages. Using data based on the careers of individuals born in the United Kingdom in 1958, we find evidence of a trade-off, but only for real wages and only for the group with lower vocational education” (Brunello and Rocco, 2017, p. 106)

“According to the alleged trade-off, the initial advantage provided by a vocational education decreases with age and gradually turns into a disadvantage, not only for the individual herself but also for the aggregate econom” (Brunello and Rocco, 2017, p. 107)

“lthough the United Kingdom is not generally considered a country where vocational education plays a prominent role, as it does in some Central European countries, the Wolf Report (Wolf 2011) shows that this type of education is by no means marginal in a nation where a majority of “14 –19 education covers ‘academic’ as well as ‘vocational’ subjects. ...A small minority follow entirely academic GCSEs, but most 14–16 year olds take some form of ‘vocational’ qualification, and some follow a programme in which GCSEs play quite a small part. Post-GCSE, about a third take only A levels, the conventional ‘academic’ route.” (Brunello and Rocco, 2017, p. 108)

“We distinguish between two levels (“lower” and “higher”) of both vocational and academic education, depending on whether the highest qualification attained referred to NVQ (National Vocational Qualification) level 2/3—roughly corresponding to the more popular ISCED (International Standard Classification of Education) 3—or to level 4/5, corresponding to ISCED 5 (see Schneider 2008), and conduct our comparisons within each level” (Brunello and Rocco, 2017, p. 109)

“One of the merits of vocational education that is often reported in the economic literature is that it improves the labor market opportunities of youths who lack the resources, skills, or motivation to continue with higher education.” (Brunello and Rocco, 2017, p. 111)

“Ryan (2001) summarizes cross-national data showing that vocational programs, and in particular apprenticeships, increase the chances of an early working life” (Brunello and Rocco, 2017, p. 111)

“often argued that the advantage of vocational education in providing ready-to-use skills, and therefore its ability to facilitate the transition from school to work (e.g., Wolter and Ryan 2011), comes at the price that those skills quickly become obsolete in modern economies characterized by rapid technological change (e.g., Krueger and Kumar 2004” (Brunello and Rocco, 2017, p. 111)

“In another study focusing on Swedish data, Golsteyn and Stenberg (2015) use register data collected between 1978 and 2011 to compare earnings over the life cycle for individuals with vocational and general—or academic—education. They find that, for males, vocational education is associated with an initial relative earnings advantage that turns into a disadvantage after about 10 years in the labor market” (Brunello and Rocco, 2017, p. 112)

“The studies that have addressed the endogenous selection of individuals into different curricula6 by using exogenous policy changes typically find that the labor market returns to vocational and academic education are not statistically different.” (Brunello and Rocco, 2017, p. 112)

“Oosterbeek and Webbink (2007), for example, evaluate the effect on earnings when an additional year of academic education was added to basic 3-year vocational programs, a policy that was implemented in the Netherlands in 1975. Adopting a difference-in-differences strategy, they find that the extra year had no effect on wages 20 years later.” (Brunello and Rocco, 2017, p. 112)

“Malamud and Pop-Eleches (2010) find that workers in Romania with a vocational education are significantly more likely than academic education graduates to work as manual workers and in craft-related occupations.” (Brunello and Rocco, 2017, p. 112)

“Dearden et al. (2002) report that wage premiums associated with academic qualifications are typically higher than those accruing to vocational qualifications. The gap is, however, somewhat smaller when the authors control for the time required to acquire different qualifications. This is particularly important for vocational courses, which generally take a shorter time to complete.” (Brunello and Rocco, 2017, p. 113)

“We define the educational attainment of each individual in our sample as the highest qualification attained by the age of 23. Individuals older than 16 can take a broad range of qualifications, both academic and vocational, in the United Kingdom. In accordance with Dearden et al. (2002), these qualifications can be classified into five different levels—with 1 for the lowest and 5 for the highest—using the NVQ system as reference.” (Brunello and Rocco, 2017, p. 114)

“In accordance with Dearden et al. (2002) and Dodgeon et al. (2011), we first identify each individual’s highest academic and vocational qualification—in NVQ levels—and then define his highest qualification (academic or vocational) corresponding to his highest NVQ level.” (Brunello and Rocco, 2017, p. 114)

“This information is used to allocate each individual to one of the following five education types: low education, corresponding to no education or NVQ level 1; lower vocational (V1) or academic (G1) education, corresponding to NVQ levels 2 and 3; and higher vocational (V2) or academic (G2) education for NVQ levels 4 and 5.” (Brunello and Rocco, 2017, p. 114)

“As a result of attrition, there are 4,879 males composing the NCDS sample in 1981 and only 2,674 in 2013. Blundell et al. (2000) examine whether the attrition occurring between waves 1 and 5 (1991) have biased the composition of their main sample and find that the sample is similar to the representative Labor Force Survey sample” (Brunello and Rocco, 2017, p. 116)

“Lindeboom, Lundborg, 116 Journal of Human Capita” (Brunello and Rocco, 2017, p. 116)

“and van der Klaauw (2009) quote evidence from the National Child Development Survey Use Support Group showing that sample survivors do not differ in a relevant way from the original sample in terms of social and economic status, education, and health” (Brunello and Rocco, 2017, p. 117)

“Using a probit specification, we regress the probability of nonresponse in the last two, three, four, and five waves on indicators of early-life conditions, test scores at ages 7 and 11, and four dummies, one for each education type above low education (V1, G1, V2, and G2). Our results, presented in table A2, indicate that the probability of dropping out of the sample does not vary significantly between education types.14 We interpret this finding as evidence that selective attrition in NCDS is not a problem when the labor market outcomes of the two groups are being compared.” (Brunello and Rocco, 2017, p. 117)

“Over the life cycle, this initial advantage is rapidly eroded but never turns into a disadvantage.” (Brunello and Rocco, 2017, p. 139)

“f anything, our results provide evidence that higher vocational education outperforms higher academic education in terms of employment rat” (Brunello and Rocco, 2017, p. 139)