Education, skills and a good job: A multidimensional econometric analysis

Education, skills and labor market outcomes are tightly linked. Most empirical evidence about their interconnections is obtained using rich longitudinal datasets coming from developed countries, and often treat earnings as the sole outcome of interest. Much less is known about the same in developing countries due to lack of appropriate data. This paper is an attempt to fill this gap by operationalizing the technology of skill formation framework using a static dataset with some information on past variables. Following the theoretical underpinnings of modern development paradigms, we define our variable of interest to be a multidimensional concept of work-related well-being, going beyond wages to include employment opportunities, decent working time and safe work environment. We thus apply a suitably adapted version of the above framework, resulting in a simultaneous equation model with latent variables, to Bolivian data. We find that an above-average well-being in terms of employment opportunities and earnings is only observed in the top-most quintile of the skills distributions, whereas the top three quintiles are relatively well-off in the safe work dimension. Overwork is responsive to cognitive skills but not to non-cognitive skills, and it is highly prevalent across the entire distribution of the former. These two types of skills are also differently influenced by education. An individual with a primary schooling is already in the above-average group in terms of non-cognitive skills, a condition requiring an undergraduate college degree in the case of cognitive skills. From a policy perspective, we note that, contrary to the general findings in a developed country context, the premium for cognitive skills on the labor market is higher than that for non-cognitive skills. This can be explained by the relative scarcity of the former, which is mostly acquired through formal education, a situation often encountered in many developing countries. (C) 2019 Elsevier Ltd. All rights reserved.