Addressing endogeneity in the relationship between health and employment through a simultaneous equations model

Este documento analiza la relación de doble causalidad entre salud y empleo y su comportamiento dinámico usando datos de Estados Unidos tomados del PSID (Pane Study of Income Dynamics). Este estudio usa dos variables dependientes (Estado de salud auto-reportado y Empleo), las cuales son estimadas usando un modelo probit bivariado para abordar el problema de endegeneidad presente en dicha relación. Los resultados muestran evidencia significativa de la existencia de dicha endogeneidad y del impacto positivo que tiene sobre la probabilidad de ser empleado tener un buen estado de salud y vicesersa, sin embargo, el impacto de la situación de empleo sobre el estado de salud se encuentra que no es significativa.

This document examines the two-way relationship between health and employment and their dynamics using U.S. data from the PSID (Panel study of Income Dynamics). This study uses two dependent variables (Self-assessed health and Employment) which are estimated using a bivariate probit model to address the endogeneity problem present between them. The results show that there is a significant evidence of the existence of endogeneity and suggest that good health positively affects the probability of being employed (healthy people have 2.85% more chances to join the labour force than unhealthy people) and that there is a positive impact of being employed on the probability of reporting good health (employees have 0,07% more chances of being healthy than non-employees), however, the effect of employment status on health is found not significant.