Skill Formation and the Potential to Have a Good Job
Krishnakumar, Jaya (2); Nogales-Carvajal, Ricardo (1) (2016). 'Skill Formation and the Potential to Have a Good Job' Paper presented at the annual conference of the HDCA, Tokyo 2016.
abstract We analyze human development in the context of labor markets building upon a combination of Heckman's (see e.g. Heckman and Mosso, 2014) and Sen's (see e.g. Sen, 2009) body of academic work. Most modern studies on this aspect of individual wellbeing rely on classic theories developed by Becker (1993) and Mincer (1974), which tend to reduce wellbeing generated in the labor market to wages or monetary income and to consider schooling as a direct diterminant of this unidimensional notion of wellbeing. In our paper, we take a different stand on both statements. On one hand, based upon Sen’s Capability Approach, we defend the idea that having a good job is a multidimensional aspect of a person’s development as human being and as a professional. Thus its assessment calls for a theoretical framework that does not reduce this wide concept to one single aspect. We take into account multiple labor market outcomes simultanously to assess the adequacy of a job, including income/wage, stability and access to social security. In our view, resourcing to one dimensional analyses implies taking a narrow view on a sensible social subject and we build our analysis explicitly avoiding such an unnecessary oversimplification. On the other hand, years of schooling or more generally, time spent in any formal educational programs are, in fact, first-order educational outcomes that enable other personal outcomes in turn, such as abilities or skills (Fasih, 2008). Even if time dedicated to formal education give important signals to employers about a person’s abilities and skills, it is these latter aspects of human capital that may transform into productivity and for which labor markets generate returns (Heckman et al., 2006; Kautz et al., 2014; Lindqvist & Vestman, 2011). This leads us to consider important links between schooling, skills and labor market outcomes. We do this based on Heckman’s most recent contributions to human development theories; we give partircular importance to the dynamic process of skill formation, from childhood through adulthood accounting for possible endogeneity of investments in their acquisition and the unobservable nature of human abilities. We develop a robust econometric framework for the assessment of the relation between individual wellbeing generated in the labor market, skills and investments for their acquisition. This framework consists of a Simultations Equation Model (SEM) with latent variables. In this framework, we consider the potential to have a good job as a person's wellbeing, respecting her individuality and freedom to choose. As this potential or capability is unobservable, we treat it as latent variable. We take into account two types of skills: cognitive skills, defined roughly as intelligence or acquired knowledge and non-cognitive skills, related to personality traits, attitudes and behaviour. These skills are also unobservable and manifest themselves partially through some tests and psychological assessments. Thus we also treat them as latent variables. Investments for skill formation are crucial in our framework and they include schooling and parental efforts to reduce negative age gaps in school-start. Our empirical application concerns Bolivia; we use data from the Bolivian household survey of the World Bank’s Skills Towards Employability and Productivity (STEP) program, which was collected in 2012. In order to secure robustness and avoid potential bias caused by endogeneity of investments for skill formation (which is heavily stressed in Heckman-based literature on this subject), we estimate three variants of our structural model. The first variant considers investments for skill formation as exogenous; the second variant considers these investments as endogenous and introduces investment policy functions to model decisions to invest in a reduced form of the structural equations; the third variant also considers investments as endogenous and estimates the investment policy functions explicitly as a first step of a standard 2-stage IV procedure. In light of a Hausman test, we confirm endogeneity of our two considered investments: years of schooling and school-start gaps, thus we retain the third variant of our SEM as the adequate instrument for assessing the Bolivian case. Among our most salient results, we find that: both cognitive and non-cognitive skills are important determinants of the capability of having a good job. Cognitive skills have an effect that is 4.5 times greater relative to non-cognitive skills. This means that the Bolivian labor market offers greater returns to acquired knowledge as opposed to positive personality traits. One reason for this is that the first are scarcer among the population of workers. years of schooling are effective drivers of better skills (cognitive and non- cognitive). People with tertiary education tend to triple non-cognitive skills compared to people that have only finished primary school. Similarly, cognitive skills tend to almost double for people with tertiary education. Minimizing schooling-start gaps is also fruitful for skill development and consequently, for the capability of having a good job. Greater school-start gaps hinders cognitive skills most severely, as it may reduce them up to 0.35 times compared to timely start of schooling. education of both mother and father have a significant influence on skills, on investment decisions and consequently, on capability of having a good job. More privileged family background yields more investments, greater skills and thus greater capabilities for a good job. Cognitive and non cognitive skills may end up doubling if one has at least one parent with tertiary education. This result indicates important shades of social and educational immobility in Bolivia.