The effect of work on mental health: does occupation matter?

This paper considers the effect of work choices on mental health and looks at whether this differs across occupations. This requires a model that can deal with the endogeneity in the relationship between health, occupation and work choices. We specify such a model and estimate it on a unique UK panel survey. The survey, called the National Child development Survey (NCDS), follows a cohort since their birth in 1958 until age 42. The analyses show us that early childhood health and ability have long lasting consequences for the mental health at the later ages. Females have lower levels of mental health. Mental health deteriorates with age for males and females, but the rate of deterioration is substantially lower for females. We also find that the rate of depreciation is lower when individuals work. For females we find large effects of occupation, for males we do not find this. Employment status is important for males, but not for females. For both genders we find very large effects of the onset of a long-standing illness. The probability of experiencing such an event depends on employment status, occupation and life style variables. Copyright 2004 John Wiley & Sons, Ltd.