The impact of unemployment and familial integration on changing suicide rates in the U. S. A., 1920-1969

This study attempted to examine the Durkheimian notion of the variation in suicide rates by introducing a time series dimension instead of the cross-sectional method employed in most previous sociological investigations. The major theoretical thrust of this paper is that social solidarity is not constant, and that socio-economic and demographic forces will affect social solidarity which, in turn, will affect suicide rates. Based on time series data for the U. S. A. for the period 1920-1969, the relationships between suicide rates controlled for sex and race, and unemployment and the ratio of divorce to marriage were examined. The regression analysis produced the following significant findings: (1) Unemployment rates tend to be the most important and stable predicator of short and long term variations in suicide rates examined across time. (2) Familial disintegration as measured by the ratio of divorce to marriage was not found to be a significant predicator of suicide rates for males or females. Furthermore, the correlation between marriage, divorce and suicide rates was found to be rather low and was not found to be consistently in the direction predicted by the Durkheimian theory of suicide. A further refinement of the analysis employed a separate regression analysis for the period 1920-1941 (pre WWII era) and 1946-1969 (post WWII era). This analysis supported the contention that there is a differentiation in sex roles with regard to suicide rates. Thus in the post WWII period whereas neither males nor females (white or non-white) were affected by changes in familial integration, males seemed to respond to economic hardships as measured by unemployment, while females did not. It was also noticed that unemployment lost some of its high predictive power in explaining variations in suicide rates after the war. It is evident that future research is necessary to identify socio-structural variables that can explain changes in female suicide rates. The operationalization of familial integration is a problem, because in different time periods marriage and divorce rates are related differently to suicide rates controlled for sex and race. This also suggests a new perspective for theorizing and data interpretation, namely that the changing socio-cultural connotations of divorce and marriage and unemployment must be considered as intervening variables in the time series analysis of suicide rates.