Occupational Safety and Health: Understanding the territorial bias

The European Agency for Health and Safety at Work (2013) underlines that OSH interventions are not evaluated by rigorous evidence-based research. Moreover, most interventions concern regulation requirements and related inspections and sanctions. In this panorama, the Italian case is noteworthy. In 2008, a system of economic incentives has been introduced, providing SMEs with grants to invest in the OSH field. This represents a sort of revolution, leveraging corporates' social resposability towards their workers. Inail financed the most important intervention, both in terms of beneficiaries and amounts: since 2010, Inail "ISI calls" supplied about 2 billion euros. This represents a huge budget in a quite long pilot experimentation, but at the policy level the balance is still leaning in favour of sticks rather than carrots. Up to now, the evaluation analyses performed by Inail on ISI calls concerned implementation processes and performance monitoring, reporting, and accounting. Recently, the need for an ex-post evaluation of impacts emerged. Therefore, Inail is funding a research aimed at identifying appropriate models to assess the impact of ISI incentives and to highlight their strengths and criticalities as an economic support to SMEs. After overviewing the main theoretical and methodological aspects in counterfactual evaluation of OSH policies, the paper presents some empirical spatial analysis aimed at identifing proxies of the firms' attitudes towards safety in different Italian regions. This study is propedeutical to investigate the external validity of OSH evaluation studies. In its preventative activity, Inail identifies firm size and sector as key variables to explain the risk of occupational incidents. If these variables completely explain risk variability, the expected frequency of incidents in a territory could be rather precisely estimated by the composition of its economic activity in terms of sectors and firm size within sectors. On the other hand, a deviation of the observed frequency from the expected one is a measure of unobserved determinants related to territorial characteristics, which we could label "territorial bias": e.g., a cultural attitude towards safety and work relations, the role of "informal economy", the economic situation, the social cohesion of the territory. We will rely on very detailed data on OSH to measure this "territorial bias" and to identify its main determinants by means of econometric (spatial) models.