Das Dokument ist öffentlich zugänglich im Rahmen des deutschen Urheberrechts.

The strong relationship between various health indicators and education is widely documented. However, the studies that investigate the nature of causality between these variables became available only recently and provide evidence mostly from developed countries. We add to this literature by studying the causal effect of education on days hospitalized and days out of work for health reasons. We consider two educational reforms. One is the educational expansion of the early 1960s and the other is the 1997 increase in compulsory level of schooling from five to eight years. However, due to the possibility of weak instruments we do not further pursue this avenue. We focus on individuals in two cohorts namely, 1945-1965 which is an older cohort and 1980-1980 which is a younger cohort. We estimate Tobit models as well as Double Hurdle models. The results suggest that an increase in years of education causes to reduce the number of days hospitalized for both men and women unambiguously and the number of days out of work only for men while an increase in education increases the number of days out of work for a randomly selected women.