The Effects of Vocational Education on Adult Skills and Wages. What can we learn from PIAAC? Giorgio Brunello and Lorenzo Rocco University of Padova This research is commissioned by the OECD. The views expressed here are those of the authors and cannot be attributed to the OECD or its member countries.
Motivation Vocational education is a key lever to meet the EU2020 goals of reducing the share of drop-out to less than 10 percent increasing the share of individuals with tertiary education or equivalent to 40 percent Vocational education is reputed to facilitate the transition from school to work and offer ready opportunities of employment
Motivation However on the one hand Hanushek, Woessmann and Zhang, 2015, argue that there is a trade-off in vocational education at early ages vocational education improves employability at later ages vocational education hampers employability, because of skill depreciation studies that exploit exogenous variation in the educational systems do not find evidence of beneficial effects of vocational education compared to academic education (Malamud and Pop-Eleches, Pischke and von Wachter, Hall, Oosterbeek and Webbink)
What we do We exploit PIAAC to evaluate the effect of vocational education on test scores labour market outcomes training health, social capital, job satisfaction
What we do Advantages: PIAAC covers many countries provides standardized test score and measures of LMO it is possible, though not always, to derive an indicator of VET Disadvantages: difficult to exploit exogenous variations reforms of vocational education exist only in a few countries and each single country sample size is relatively small only the highest degree of education is available
What we do we estimate the effect of vocational/academic education resting on the assumption of "selection on observables" certainly not optimal, but there is a reasonably long battery of controls (parents education, books at home while 16, country economic conditions at the time of the relevant school choice ) is this enough to resolve the problem of selection? difficult to tell
What we do However, thanks to the CIA we can estimate the differential effect of vocational education by length of education by country by gender by age groups on a wide range of outcomes
Vocational/academic education At ISCED 2 PIAAC does not allow to distinguish between V and A we limit our analysis only to the individuals with at least ISCED 3 At ISCED 3 and 4 we have compared UNESCO data with the data provided by PIAAC country managers (not always available or detailed enough) and retained only those countries where the two sources agreed. At ISCED 5 we have assumed 5A to be academic and 5B to be vocational
The sample We have been able to define vocational/academic education for 17 countries Australia, Austria, Canada, Czech Republic, Estonia, Finland, France, United Kingdom, Germany, Ireland, Japan, Korea, Netherlands, Norway, Poland, Spain, United States. focus on individuals aged between 25 and 59 (or 64 for some outcomes) when education is pre-determined with at least ISCED 3 (84% of PIAAC sample)
Method Four (five) possible alternative treatments vocational education at ISCED 3 or 4 vocational education at ISCED 5 academic education at ISCED 3 or 4 academic education at ISCED 5 (ISCED 2) Multi-valued treatment effect model: IPW-RA (inverse probability weighting + regression adjustment) common support CIA: conditionally to controls, each treatment is assigned as good as randomly
(Some) Results
Effectiveness of Vocational Education. Numeracy
Numeracy
Numeracy
Problem Solving in a technology rich environment
LMO: log hourly earnings
LMO: log hourly earnings
LMO: employment
LMO: employment
LMO: employment
LMO: log years at work
Training: are IVET and CVET complements?
Training
Vocational at ISCED 3-4 vs ISCED 2
Vocational at ISCED 3-4 vs ISCED 2
Conclusions Vocational education does not perform as well as academic education (of the same level) both in labor market outcomes (wages) and in the level of basic skills, such as numeracy and literacy This is especially true for higher (tertiary) education At the secondary and post-secondary level, vocational education shows relatively better performance, especially for employment, as differences with academic education are small
Conclusions Should we conclude that vocational education should be reduced and that more academic education should expand? This is what is happening in many countries! Our view is more nuanced. We have said a few things about expected benefits of different curricula, but nothing about costs. Academic education provides on average higher benefits, but for some individuals the costs may be even higher Vocational education could therefore be privately optimal for some, and may even reduce early school leaving.