The Effect of Labour Contracts on the Health of the Workers
This research regards the effect of the type of employment contracts on the health of the workers and the risk of workplace accident. Given the higher diffusion of temporary jobs in many EU countries, it is important to understand if this can affect the health of the European citizens. Several reasons might determine a higher risk for the workers such as the worse working conditions due to the lower bargaining power, the smaller incentive for employers to provide safety training (Pouliakas and Theodossiou, 2013) and the greater effort of the worker to ensure confirmation at the expiration of the contract (Guadalupe, 2003). Descriptive statistics indicate that workers in temporary contracts tend to suffer more from workplace injuries. However, the empirical literature on the effects of temporary contracts is scarce because few administrative datasets contain information on the occupational history and the workers’ health. To assess the presence of this causal link, together with Prof. Dr. Jan Van Ours (Tilburg University) and Dr. Matteo Picchio (Università Politecnica delle Marche, we access the Italian Whip-Health dataset, a very rich and new dataset merging information from the Italian Ministry of Health and the Italian Social Security Institution.
An accident at work can be defined as an occurrence in the course of the work, or during the journey from home, which leads to physical or mental harm to the worker. According to Eurostat (2015), in 2013 the number of serious non-fatal accidents amounted to 3.08 million in EU-28. Furthermore, we have to add the most serious cases of occupational fatality: 3,652 in EU-28 in the same year. Besides representing a threat to the well-being of the individuals, health risks are also an economic cost to society. Statistics show that in Europe two-thirds of the fatal and serious accidents take place in specific sectors such as construction, manufacturing, transportation and warehousing, agriculture, fisheries and forestry (Eurostat, 2015).
Thanks to the rich data we can control for several factors determining health risks for the individual such as job characteristics (e.g. manual), firm size, sector, experience (e.g. in the same job), gender, age and regional unemployment rate. The covariates that can be in turn influenced by the type of contract (e.g. experience) will be referred to the period before the start of the contract to avoid capturing part of the treatment effect ("endogeneity"). As the data are in a panel format we can also control for the remaining time-invariant unobservable characteristics. This regards the relative propensity of the individual to follow dangerous behavior at work and/ or have health problems ("individual fixed effects"). Similarly, we can control for specific firms’ characteristics in causing health problems or accidents to its employees ("firm fixed-effects").
The literature suggests other factors are correlated with a higher likelihood of workplace injuries such as the firm size, age, gender, level of education and job characteristics (e.g. low level of independence, overtime and monotonous work - Pouliakas and Theodossiou, 2013). Furthermore, descriptive evidence shows that temporary workers are more likely to be affected by a workplace accident (e.g. Hernanz and Toharia, 2006) while another study estimates a causal link between a shorter job tenure and a higher probability of workplace accident (Bena et al., 2013). It is not easy, however, to distinguish between the causal effect of contracts and the intrinsic characteristics of the workers or the firms using these contracts.