Education and training in a model of endogenous growth with creative wear-and-tear

How does the adoption of new technologies and the corresponding destruction of technology-specific skills influence education and training decisions as well as social welfare? We study this question by developing an endogenous growth model in which technological change is driven by R&D in a “love-for-variety” setting, while human capital accumulation occurs in a job-competition-like framework. The model combines elements from Romer (1990), Aghion and Howitt (1992) and Ramsey (1928). Productivity growth is the result of quality improvements in a context of intertemporally optimal consumer decisions and gradual obsolescence of older intermediates, as in a 'creative wear and tear' process where all varieties live forever but gradually fade away over time. To this basic set-up, we add the Ramsey intertemporal optimisation framework in order to endogenise the households’ decision about the allocation of their resources over different activities, among which consumption and self-financed education. In line with a job-competition model of the labor market, we assume that workers do not fully acquire their work-relevant skills at school, but also through on-the-job training. By creating new jobs, technological change increases the number of training slots to be filled in. General education, on the other hand, reduces the training costs associated with these new jobs. Thus, investing in education creates some explicit trade-offs. On the one hand, higher levels of education represent a signal of a higher trainability of individuals, so that firms can expect to benefit from lower training costs and workers from higher earnings profiles. Moreover, higher education fosters R&D and the production of new technologies. On the other hand, higher education implies an increase in the rate of creative destruction, forcing firms to provide more spells of specific training, a drop in total working hours available for current production and R&D, and a higher expenditure by households at the expense of private consumption. The paper finds some important results. It is able to reproduce the mixed empirical findings on the relationship between general and specific human capital accumulation. Even though they are substitutes at the microeconomic level, technological change makes them complementary in a general equilibrium setting. Hence, the effect of a rise in R&D productivity is to (re)train more people for shorter periods of time while leaving the growth maximising level of education unchanged, thus shifting their optimum time 'portfolio-mix' in favour of general education, since it provides a relatively solid basis for developing specific skills that are prone to creative destruction. However, when we endogenise education costs, we see that private households’ decisions leave growth opportunities and training cost reductions unexploited, thus ‘calling out’ for public policy intervention.