Income, employment and suicidal Behavior

Background: Little is known about the labor market outcomes of people who have attempted suicide or thought about suicide. Methods: We used micro-level data from the first wave of the National Epidemiologic Survey on Alcohol and Related Conditions 2001-2002 to examine the effects of suicidal behavior on income and employment. The data provide a representative sample of the U.S. population, with its primary purpose to provide information on alcohol use disorders for the civilian non-institutionalized population aged 18 and over. The data include employment, income, and other socioeconomic and demographic information on respondents. Since the survey included 43,093 people, the data include a large number of respondents who attempted suicide or thought about committing suicide. We estimated earnings regressions and logit and ordered logit employment regressions. We used methods of IV estimation as well as two stage linear probability models to address potential endogeneity of suicidal behavior while estimating regressions separately by sex, since there are significant differences in suicide rates, suicide attempts, and suicidal ideation between men and women. Results: We find that suicide attempts and suicidal ideation are negatively related to personal income and the probability of employment. The effects differ by sex. Men and women who attempted suicide had mean earnings lower by 16 and 13 percent, respectively. This amount reflects the combined effect of suicidal behavior and mental illness. With instrumental variable regression, the magnitude of the effects becomes larger-for example, as much as 50 percent decrease in the income of males who attempted suicide. Thoughts of suicide negatively affect income but to a smaller extent. Logit and ordered logit regressions indicate that attempted suicide reduces the probability of fulltime employment by over 20 percentage points for men and approximately 17 percentage points for women. Implications: People who engaged in suicidal behavior reported significantly lower employment and earnings. Although data were insufficient to directly address the issue, it appears the effect may persist over a long period of time. This is particularly troubling since health insurance is closely tied to employment and without health insurance, treatment options may be limited.