Carnero, M.A., Martínez, B., and Sánchez-Mangas, R.
International Journal of Manpower – 33.3 (2012), 322-339

Paraules clau:: Bivariate probit model, bullying, health, workplace

Resum: The purpose of the paper is to evaluate the effect of mobbing in workers’ health. Using a dataset of Spanish workers from the VI Spanish Survey on Working Conditions, we identify mobbed and not mobbed workers and use two different health indicators: the worker’s perception that work affects health, and the presence of bad health symptoms. We analyze the incidence of both indicators in mobbed and not mobbed workers then estimate an econometric model to explain the probability of suffering bad health in terms of mobbing and a set of job and workers’ characteristics. We deal with the potential endogeneity of mobbing. It is found that, regardless of the health indicator used, the probability of suffering bad health is significantly higher among mobbed workers than among those not being mobbed. Moreover, when the health indicator is based on the worker’s perception, not taking into account the endogeneity of mobbing leads to underestimation of its effect on health.