The measurement of alcohol-related social problems in Sweden.

Abstract

During the first decades following the end of World War II, registers were the predominant data source in Sweden for studying alcohol-related problems and, before the abolition of the rationing system, also for mapping drinking habits. This was possible due to the strict individual control system of alcohol that dominated in Sweden for many decades. With the gradual shift from individual control to general control in the 1960s, and 1970s, the possibility of using registry data was reduced, and for the past 20 years or so they have almost never been used to study the relationship between drinking and social problems. Instead, not only drinking levels and drinking patterns, but also social problems associated with alcohol, have slowly but to an increasing extent been measured by self-reported questions in general population surveys. This paper discusses experiences gained from working with self-reported survey data on alcohol-related social problems in Sweden for the past few years and describes discrepancies between register- and survey-based analyses. Swedish examples suggest that it is unlikely that survey data can be used successfully to estimate the prevalence of serious alcohol-related social problems in society. Survey data may be more suitable for estimating risks associated with different drinking levels and drinking patterns in the general population on an ordinal level. However, there is still much room for improvement in the Swedish alcohol surveys that include questions on alcohol problems.