The role of personal networks for the labour insertion of weak jobseekers

The enduring economic crisis, persistently disrupting job positions overall, and the consequent exacerbation of competitiveness of the labour markets claim for a robust strengthening of individual professional skills, particularly among the weak targets (immigrants, women, youth). Most of all, an high unemployment level represents a dangerous social threat, mostly weakening the socio-economic condition of already marginalized people, determining missed economic growth, increased social inequalities and crime, hence rising public expenditure. This paper investigates immigrants' socio-economic integration in Piedmont, Italy, analyzing the labour market outcomes of a representative sample of individuals participating to regional vocational training programmes. In fact, recent studies (Ragazzi et al., 2012, 2013) show that vocational training policies significantly rise trainees' employability, but the transition from vocational training to labour market remains nevertheless difficult for all disadvantaged individuals, including foreigners. Clearly, human capital investments cannot solely achieve labour integration: quanti-qualitative analyses suggest that previous working careers and the individual's social context can either amplify or overrule the effect of training (Ragazzi and Sella, 2013). In this perspective, we propose a counterfactual analysis to describe and assess the effect of individual social networks (family, friends, other acquaintance) on the employment outcome of trainees, particularly focusing on immigrants. In fact, both recruitment strategies and job-search processes are governed by spoken and unspoken rules, that are mostly determined by personal networks. Hence, the characteristics of personal networks (actors, extension, intensity) are analyzed relying on the self-assessment of individuals in the sample. Furthermore, the influence of different network configurations on the probability of getting employed after the training is explored by parametric models, controlling for individual characteristics and other factors, including previous education and working careers.