Is the effect of income on the suicide rate always negative? A test of Barnes' theory

Abstract: Durkheim’s socioeconomic hypothesis of suicide has been a popular theory to test for sociologists. However the results have been mixed, offering very little cumulative sociological knowledge. Previous theory and research have found that there are contradictory results in the direction of the income regression coefficients used to study the relationship between income and suicide rates depending on if a time-series or a cross-sectional approach has been used. It has been hypothesised that the contradictory results are caused by a specification bias, namely failing to account for lagged income, which is influencing the direction of the regression coefficients. The aim of this study is to investigate the effect of income on the Swedish municipalities’ suicide rates by replicating a study published in The American Journal of Sociology by Carl B. Barnes from 1975, in which he argued that the effect of income on the suicide rate is always negative. This cross-sectional analysis is based on municipality data on male, female and overall suicide in 2002 to 2004 from the Swedish Centre for National Prevention of Suicide and Mental Ill-Health at Karolinska Institutet among the working-age population (25-64 years). Control variables are the hypothesised lagged variable causing the specification bias, education, and three other possible contributing factors to suicide; unemployment, alcohol consumption and divorce. The results of the correlation and regression coefficients show that there is a negative effect of income on suicide for those aged 25 to 64 years when the other variables are held constant for both sexes; however the male results are not statistically significant. These results speak against the socioeconomic hypothesis of suicide, but generally confirm earlier findings of a negative relationship between median income and suicide. Low education is positively related to the suicide rate for males but there is no such relationship for female suicide. The findings also confirm alcohol consumption as an important factor in explaining the suicide rate. Unemployment and divorce show mixed results for the male and female suicide rates. The female unemployment rates are negatively related to suicide while male unemployment rates are not significant, on the other hand the divorce rates show a strong positive association with the female suicide rate and a negative association with the male suicide rate.