Divorce and Drinking: An Analysis of Russian Data

Abstract

This study uses data on 374 respondents, aged 16 and older, from the Moscow Oblast Survey to examine the relationship between alcohol consumption and divorce in Russia. A logit regression model of drinking in Moscow finds that both divorced and single persons have elevated risks of drinking relative to married persons. This relationship is independent of control variables including gender, education level, and religiosity. Hence, although Russian alcohol consumption has been largely typified as a reaction to macrosociological problems such as terrorism and chronic shortages of consumer goods, a Western model of divorce and drinking is supported.Research on the relationship between marital status and drinking has been largely restricted to American and Western samples (e.g., Cahalan & Cisin, 1976; Doherty, Su, & Needle, 1989; Gray, 1978; Magura & Shapiro, 1988; Temple et al., 1991; Wechsler, Demone, & Gottlieb, 1978). As Kohn (1987) contends, cross-national research is needed to determine whether American findings will be replicated in cultural and institutional frameworks different from the United States. It may be that a particular relationship may not replicate outside of the U.S., given its relatively uncommon set of cultural and structural conditions, including high unemployment (U.S. Bureau of the Census, 1993, p. 859), high religiosity, and ethnic heterogeneity. Comparative research has the advantage of exploring whether or not a relationship is dependent on a relatively unique configuration of socioeconomic conditions.The present study explores the relationship between divorce and drinking in Russia, a nation with a substantially different social structure than the U.S. Given such factors as political totalitarianism, shortages of consumer goods, and a relatively high level of overall alcohol consumption (e.g., Partanen, 1987; Segal, 1990), the relationship between marital status and drinking may not replicate in Russia. In the Russian context, generalized political, historical, and social circumstances might instead be the factors that overwhelmingly influence alcohol consumption.DIVORCE AND DRINKINGThe literature on marital status and drinking has been largely restricted to the U.S. It has followed a variety of methodologies. A distinction is drawn here between research using aggregated data and research that employs individual-level data. Work based on aggregated data refers to ecological analyses where the rate of alcohol consumption in a population is correlated with indicators of marital status in the same population. An example would be to compute a correlation between the divorce rate (divorces/1,000 population) and the alcohol consumption rate (liters/person/year) over time. A positive correlation would indicate that, as the divorce rate increased, alcohol consumption also rose over time. From research based on this methodology, however, it is not clear if the persons who are divorced have, in fact, a higher incidence of drinking than other persons.This ecological problem can be addressed in work based on individuals. An example is to explore the actual amount of drinking in a sample of persons both before and after a divorce. If the amount of drinking is higher after a divorce, then we have a firmer basis for postulating that divorce may be causally connected to drinking.A second distinction can be drawn between research that is longitudinal and research that is cross-sectional. Longitudinal research is based on data collected at more than one point in time. An example would be to correlate the percentage who drink with the divorce rate over a period of time, say from 1950 to 1990. Cross-sectional research is based on data gathered at only one point in time. An example would be to correlate the percentage who drink with the percentage who are divorced in the 50 states in a single year, say 1990.Given these two methodological distinctions, the available research can be divided into four methodological styles. …