NCJ Number
178181
Journal
Drug and Alcohol Review Volume: 17 Issue: 1 Dated: March 1998 Pages: 27-37
Date Published
1998
Length
11 pages
Annotation
This study of the relationship between alcohol use and suicide in Switzerland compares the results obtained with the Box-Jenkins approach (which the authors believe has serious limitations) and those obtained with equation modeling, which is more often applied in econometric time-series analysis.
Abstract
Annual data were available for 1950-90 for per capita consumption of alcohol, divorce, and unemployment rates, as well as number of male and female suicides. All rates were directly age-adjusted rates per 1,000 inhabitants, with the 1990 population as the base for comparison. The study was restricted to 1950-90 to avoid any period for which data were missing for either of the variables. Two different approaches were used in time series analysis, the so-called Box-Jenkins or ARIMA approach on first differences of the series (widely used in alcohol research), and simultaneous equation modeling, better known in the econometric literature. They yielded different results, and the authors' preferred model is the simultaneous equation model. The discussion of the findings is restricted to the latter model. In Switzerland, suicide rates and per capita alcohol consumption are among the highest in the world. This corresponds to the general finding that the higher a country's alcohol consumption the higher also its suicide rates. Nevertheless, this study found the association between these variables at the aggregated level to be weak and even negative; however, this observation fits well if the effects of alcohol consumption on suicide are compared over different cultures. In "wet" cultures, such as Switzerland, the effects tend to be lower than in "dry" cultures. A general argument is that in "wet" countries, even heavy alcohol consumption carries less stigma and is less marginalized and less deviant than in "dry" cultures; thus, suicide as a result of the social response to alcohol abuse is less likely. This also implies that the composition of alcohol abusers differs between "wet" and "dry" countries. The selection to alcohol abuse for people with such other factors as social and psychiatric problems, which increase vulnerability to suicide, may be stronger in "dry" countries; whereas, there may be a greater overlap with "normal" heavy drinkers in "wet" cultures. 2 tables, 5 figures, and appended methodological information