The skill content of technological change. Some conjectures on the role of education and job-training in reducing the timing of new technology adoption.

This positional contribution has a twofold aim: the first is to explore the recent empirical literature developed around the issue of how the adoption of new technologies within the firm has changed the skill requirements of occupations; the second is to conjecture on the relationship, and on the relative sign, between technology adoption and firm sponsored onthe- job training. The basic idea is that the time-consuming dimension of the adoption process plays a direct role both in determining the profitability of the investment in new technology and in assessing the size of the productivity slowdown the firm eventually occurs after its introduction. On the extent that the timing of adoption depends on the workers' skill composition and on the distance between the skills acquired for the job and the skills required by the job, the deep understanding of the interplay between the mechanisms of human capital accumulation can be helpful in order for the firm to set suitable and efficient job-training strategies. During the last two decades the discussion around the impact of technological change on workers' human capital has been intense: the rapid diffusion of information and communication technologies (ICT) and computer-based machines (CNC, CAD), together with the large increase in the supply of highly-educated workers and rising returns to education, favoured the argument that technological change is characterized by a skill-biased nature (SBTC), leading to substantial changes in the division of labour and shifting labor demand towards employees with higher levels of education. On this purpose, different approaches have developed in the last decades that provide different evidence to a common research question. While a lot of national and international evidence still continues to support the SBTC hypothesis by employing 'traditional' aggregate measures of technological change and indirect measures of skill upgrading, a smaller literature is emerging that considers the heterogeneity of both technologies and skills at the workplace and aims at determining the demand of skills by the tasks occupations require. Even if new and interesting results emerge, many 'black holes' still remain, the most important of which seem to be the lack of theoretical and empirical models analyzing the role that school education and on-the-job training, and their interplay, can play in reducing the timing of new technology adoption.