The labour market effects of academic and vocational education over the life cycle: Evidence from a British cohort

Giorgio Brunello and Lorenzo Rocco from the University of Padova look at the benefits and costs of academic and vocational education over the long-term.

Education economists often point out that individuals with a vocational education face a trade-off between short term benefits and long term costs. In the short term, this type of education facilitates the transition from school to work by providing ready to use skills. In the long term, however, vocational skills may depreciate relatively fast and individuals who specialize in these skills may be less capable of adapting to technical change than individuals endowed with a more general (academically oriented) education.
 
The view that the labour market benefits of vocational education are short lived has been recently supported by Hanushek, Schwerdt, Wossmann and Zhang (2015), who compare the life cycle patterns of employment (and wages) for individuals with vocational and academic education using cross-sectional data from the International Adult Literacy Survey (IALS). Since IALS does not allow to follow individuals over time and to distinguish between age and cohort effects, their estimates rely on the assumption that age - employment and age – wage profiles do not vary across cohorts, so that today’s old people in each education category are a good proxy for today’s young people when they will become old. What this literature does is compare someone’s wage with a vocational education at age 50 with the wage of someone holding a vocational qualification aged 30 in the same year, and assumes that today’s 30 year olds will fare in 20 years’ time like today’s 50 year olds. But it is possible that today’s 50 year olds do quite differently because the labour market they entered was very different or because the composition of the older labour force (e.g. in terms of qualifications) is quite different than among the younger cohorts. This makes it difficult to infer the true effects of vocational education by comparing older and younger workers.


In a recent paper (Brunello and Rocco, 2015), we address the same question– whether the alleged benefits of vocational education at labour market entry turn into disadvantages with age and labour market experience –using longitudinal data from the National Child Development Survey, a long cohort study of Britons born in 1958, whom we follow for 32 years, from age 23 to age 55. The key advantage of using these data is that we do not need to assume that employment and wage profiles are invariant across cohorts, because we can genuinely follow individuals over time. A disadvantage is the limited generalizability of our results, that apply to a single country and a specific cohort, who acquired secondary and tertiary education in the 1970s. We cannot assume they hold for younger people.

Following Dearden, McIntosh, Myck and Vignoles, 2002, we allocate individuals to vocational or academic education using their highest qualification acquired by age 23, when most young Britons have completed their full time education. For both vocational and academic education, we distinguish between two levels, “lower” and “higher”, depending on whether the highest attained qualification belongs to National Vocational Qualification level 2/3 – roughly corresponding to the more popular ISCED 3 - or 4/5, corresponding to ISCED 5, and conduct our comparisons within each level.

Any evaluation of the economic costs and benefits of education types (vocational versus academic) needs to address the fact that individuals are not randomly assigned to type. Our empirical strategy consists of estimating separately the effects of time invariant education type on employment and wages at age 23 (or level effects) and the changes in these effects as individuals age and spend time in the labour market. For each outcome, we estimate changes in a consistent way using a fixed-effects estimator, and level effects using the AIPW (augmented inverse probability weighted) estimator, a method that relies on selection on observables, including detailed indicators of individual ability and early life conditions.

We combine the fixed effects and AIPW estimates to simulate life cycle employment and wage profiles by education type and level for a hypothetical individual with average sample characteristics – see Figures 1 and 2.

Figure 1. Simulated life cycle profiles of employment probabilities for a hypothetical individual with average sample characteristics. 

Males. NCDS. Excluding the self-employed and those in full time education. 

Figure 2. Simulated life cycle profiles of annual real net wages for a hypothetical individual with average sample characteristics. 

Males. NCDS. Excluding the self-employed and those in full time education.

We find limited support for the view that the early advantages of vocational education - relative to academic education - trade off with later disadvantages. When we consider employment probabilities, the estimated life cycle profiles associated with lower vocational and academic education broadly coincide. For higher education, the estimated life cycle profile associated with vocational qualifications lies always above the one associated with academic qualifications, although the initial gap attenuates considerably with age. While these results are consistent with the view that the early advantages of higher vocational education decrease with age, they do not indicate that a long-term trade-off exists.

Turning to wages, we find support for the trade-off hypothesis only for lower education. In this case, although annual real wages are initially higher for vocational than for academic education, this advantage turns quickly into a disadvantage, that is larger in mid-career and attenuates by age 50.The trade-off is even sharper when measured in terms of hourly wages (see Figure 3). In contrast, we cannot detect any trade-off for those with higher vocational education, as both real annual earnings and working hours do not vary significantly by education type over the life cycle.

Figure 3. Simulated life cycle profiles of weekly hours worked by a hypothetical individual with average sample characteristics. 

Males. NCDS. Excluding the self-employed and those in full time education

While some of our results are in line with the findings by Hanushek et al, 2015, there are also several differences. They find that individuals completing academic education are more likely to be employed by age 50 than individuals completing vocational education. In contrast, we find that employment at age 50 is about as high or even higher for those with vocational education. Hanushek et al. pool together individuals with upper secondary and tertiary education. Our data suggest instead that – at least for the UK cohort that we study – the life cycle patterns of employment and earnings for academic and vocational education differ across education levels. It is only at lower levels of education (for this cohort) that we see a trade-off which favours academic over vocational education.



References

Brunello, G and Rocco Lorenzo, 2015, The Labour Market Effects of Academic and Vocational Education over the Life Cycle: Evidence from a British Cohort, IZA Discussion Paper 9275.

Dearden, Lorraine, Steven McIntosh, Michal Myck, and Anna Vignoles. 2002. “The Returns to Academic and Vocational Qualifications in Britain.” Bulletin of Economic Research 54(3): 249-74.

Hanushek, Eric A., Guido Schwerdt, Ludger Woessmann, and Lei Zhang. 2015. General Education, Vocational Education and Labor Market Outcomes over the Life-Cycle. Journal of Human Resources, (forthcoming).