Conventional wisdom holds that individuals with serious mental illenesses will have more difficulty acquiring and retaining a job. Measuring the magnitude of the effect of any mental illness on employment empirically is difficult because of a dual-causality problem. People with mental illnesses may have difficulty gaining employment, but losing employment also has an adverse effect on mental health. How can we solve this problem?
A paper by Frijters, Johnston and Shields (2014) uses data from 10 waves (2002– 2011) of data from the Household, Income and Labour Dynamics in Australia (HILDA) survey. Mental health is measured using nine questions from the Short-Form General Health Survey (SF-36). The authors use an IV-FE model where death of a friend in the past three years is the instrument for mental health. This instrument is valid as long as the death of a friend affects employment opportunities only through a decrease in mental health. The authors also control for age, education, marital status, children, windfall income, friendships and the deaths of relative.
The authors conclude the following:
We find robust evidence that a worsening of mental health leads to substantially reduced employment. Moreover, the size of this effect is substantial, with a one-standard-deviation worsening of mental health leading to a 30-percentage-point reduction in the probability of being employed. This effect is large for both men and women…Further investigations suggest that this employment effect is larger for older than younger workers…
In this case, conventional wisdom is proven right: poor mental health has an adverse effect on employment.
- 2014), THE EFFECT OF MENTAL HEALTH ON EMPLOYMENT: EVIDENCE FROM AUSTRALIAN PANEL DATA, Health Econ., doi:10.1002/hec.3083 , , and (