The Effect of Training on Productivity: Evidence from a Large Panel of Firms

Purpose – The purpose of this paper is to investigate the effects of training activity on labor productivity in a panel of Italian firms. Design/methodology/approach – The use of a large panel data of individual firms allows the author to properly account for the possible endogeneity of training activity and avoid aggregation biases typical in industry-level data. Findings – The paper finds that training has a positive and significant impact on productivity. While unobserved heterogeneity leads to overestimate the impact of training, failing to account for the endogeneity of training leads to underestimate its effects on productivity. Within occupational groups, training has large and significant effects for blue-collar workers, while the effects for executives and clerks are relatively small. Finally, using a measure of effective training intensity the paper finds that failing to account for training duration may lead to underestimate the effect of training on productivity. Originality/value – Our data set is unique in terms of size and coverage and overcomes several limitations of previous research using firm-level data. Moreover, besides estimating the overall effect of training on productivity, the paper allows to address some more specific questions. Does the effect of training depend on the type of worker being trained? What is the relevance of effective participation to training activity?