The paper studies the driving factors of different firm training activities using two unique cross-sectional datasets at provincial level. Since the empirical literature on training at firm level is scarce, due to the costs and the intrinsic difficulty of collecting high-quality and extensive data, the paper value added is that it adds knowledge on the issue in providing new empirical evidence on the relationships between firm training decisions and firm characteristics at local Italian level. Data derive from two structured questionnaires administered to the management of 243 firms in the Province of Ferrara in 2003 and to the management of 166 firms in the Province of Reggio-Emilia in 2002. The applied analysis uses different econometric models to explore the linkages between firm decisions over training activities and the possible explanatory factors of training, at firm level.
The potential driving factors of training here analysed compounds structural characteristics, labour demand dynamics, human resource management practices, workforce features, and firm performances. The availability of an extended dataset on firm characteristics allows controlling for many relevant factors, which may explain training decisions, reducing the possible distortions arising in a cross-sectional environment.
The core of the empirical analysis thus revolves around the investigation of what the most significant driving forces of training coverage, variety of training activities adopted and training generality content are. Given a large percentage of firms declaring they do not adopt any training are present in our dataset, both OLS, Tobit and two-stage Heckman models are implemented and compared. The need of focussing the attention on different training proxies and different econometric models strongly emerges.
Summarising results, we observe that training activities emerges positively associated with productivity, high-performance practices, innovative labour demand features, workforce skill level, firm size, and affected by labour and plant flexibility in various directions. The high relevance of both structural variables (i.e. size, sector), labour demand factors and HRM/innovation practices (also positively correlated with structural variables and labour demand dynamics) shows that regional industrial policies must support labour policies within an integrated policy effort aimed at increasing potential firm productivity The analysis also suggests that a widening gap, between innovatively evolving and more stagnant firms, could characterise the future dynamics of those local areas. This is a key concern for the current debate on local systems in the European and Italian environment.