Collective bargaining and skill formation: evidence from mixed-methods.

Abstract

Human capital and industrial relations are important sources of economic growth. Knowledge supports competitiveness, and a significant part of the whole stock of skills is acquired from vocational or on-the-job training. Workplace training, in turn, is often a matter regulated under collective agreements. Moreover, the standard economics view argues that industrial relations affect competitiveness also directly, with a general take in favor of decentralization. In this paper we aim at contributing to this debate by studying how decentralized collective bargaining (DCB) affects workplace training in a country-Italy-that has progressively shifted the DCB process from the central/sectoral level, to the local/firm one, arriving in 2011 to introduce a law that allows DCB to opt out of the national labor legislation. DCB may affect workplace training in many ways. First, unions and workers' representatives may provide valuable information on skill gaps and mismatches (the "collective voice" face). Second, they may promote long-term employment relationships, creating incentives for long-run human capital investments. Third, the payment of wage bonuses-typically regulated under DCB-may fix suboptimal investments in human capital. Fourth, DCB may contain wage pressures, refrain from opportunistic behaviors (the "hold-up" problem) to limit workplace flexibility, ultimately freeing resources to "invest into the future". While these causal chains are usually in place, quantitative analysis is usually ill-equipped to provide an adequate account of them. The limitation is in data availability. At best, indeed, quantitatively treatable data sources are informative on whether a firm-based agreement is in place or not. While this may be enough to ascertain that a causal link between DCB and workplace training exists, it is powerless with respect to the identification of the actual mechanisms. This is where our paper mostly contributes to the literature. In the spirit of opening the "black box" of DCB, we mix together quantitative and qualitative strategies. More specifically, we apply a sequential research design: first, the application of state-of-the-art econometric techniques to a large representative sample of Italian companies surveyed over time through an extremely detailed questionnaire, suggests that the existence of a firm-based collective agreement enhances the probability that a worker receives workplace training and makes the per-worker and per-trainee cost of training grow. Second, thematic