Measuring the Specificity of Occupational Training Curricula and Labor Market Flexibility

In vocational education and training (VET), workers are typically required to learn bundles of skills that are specified in official occupational training curricula. In this paper we extract detailed data on the skill bundles of occupations from such official training curricula and investigate how these skills affect the labor market mobility of the workers trained in the respective occupations. Our aim is to investigate differences in the transferability of skills of an occupational curriculum, develop an empirical measure for occupational specificity and study how differences in the measured occupational specificity affect graduates’ long-term labor market outcomes.

As theoretical framework, we use Lazear’s (2009) “skill weights” approach, arguing that all single skills are general per se and that only the combination and the respective weighting of skills determines the specificity of any given occupation. The model was originally developed for firm-specific human capital, but it also provides an ideal model to conceptualize occupation-specific human capital and its labor market outcomes. We develop a method to identify a relevant set of all single skills and the skill “weights” in any given occupation by using the training goals from the official vocational training curricula. We do this for Switzerland because its VET institutions guarantee that workers indeed acquire the prescribed skill bundles and we can investigate how this affects their later labor market outcomes.

We thus build a new data set containing a set of all single skills and their respective weights in Swiss VET occupations. We firstly calculate a measure for the skill distance between different occupations, i.e., a measure for the transferability of skill bundles between pairs of occupations. Based on Lazear’s theoretical framework, we hypothesize that the skill distances determine the wage changes after occupational changes. Using a standard OLS-regression model, we find that workers who change between occupations with very dissimilar skill bundles are faced with larger wages losses than workers who change between more similar occupations.

We secondly calculate the overall degree of specificity of occupations to see whether it determines the likelihood of occupational changes of the respective workers. We do so by combining the information on the previously mentioned skill distances between occupations with the number of jobs available in all given occupations in the labor market. We use the representative Swiss Social Protection and Labour Market (SESAM) data to derive the number of jobs in all occupations in the labor market. We expect that workers who are trained in occupations with a higher degree of occupational specificity have a smaller probability of an occupational change. Our probit estimates support the hypothesis.

Our results shed light on the policy question of how the specificity of an occupational curricula affects the long-term flexibility of VET workers. The specificity of an occupation as measured by our new indicator can thus become an important policy instrument for the development of VET curricula that are supposed to support the labor market flexibility of VET trained workers.

Reference: Lazear, E. P. (2009). Firm‐Specific Human Capital: A Skill‐Weights Approach. JPE, 117(5), 914–940.