Views into Studies of Differences in Drinking Habits and Alcohol Problems between Sociodemographic Groups
Makela, Pia, Contemporary Drug Problems
In the alcohol research field, associations between alcohol consumption or problems and some basic sociodemographic variables-e.g., sex, age, socioeconomic status-are often presented. The first part of this paper discusses different uses of and views to such data on sociodemographic differences. The second part discusses different paths available to research when the aim is to get a more thorough understanding of the nature and origins of the associations observed. The conclusion is that even basic descriptive data on sociodemographic differences are useful and offer means to contribute to more general scientific discussions in the fields of alcohol policy, social policy, public health, sociology, etc. However, the more we are able to find out about why the differences exist and in what contexts they appear strongest, the more useful this information becomes and the more interesting the views offered by the analysis.
KEY WORDS AND PHRASES: Sociodemographic groups, gender, socioeconomic status, alcohol use, alcohol problems.
In most empirical studies on alcohol consumption and alcohol problems, associations between variables on alcohol use or problems and some basic sociodemographic variables, such as sex, age, socioeconomic status or marital status groups, are presented. But why is the study of these differences worthwhile, and what lies behind the associations observed? In the research field of gender and alcohol, these issues have been discussed. An excellent example of this is the book Gender and Alcohol-Individual and Social Perspectives, edited by Richard and Sharon Wilsnack (1997a). In the literature on socioeconomic factors and drinking, however, there is a wealth of research reporting associations, but the aims of studying the associations are explicitly stated only in part of this research, and there is a dearth of work that studies, or even discusses, the causes of the associations or their theoretical and practical implications.
The central sociodemographic variables that I have had in mind in this discussion are socioeconomic factors and gender, although most of the discussion below should pertain to other sociodemographic variables as well, such as age, marital status or region. The terms "socioeconomic factors," "socioeconomic group" and "socioeconomic status" are used here in a general sense, referring to different dimensions of a person's social and economic position in society, the most common measures of which are education, occupational class, and income. The term "alcohol problems" is correspondingly used as a general term referring to different indicators of problematic alcohol use and to the consequences and harm related to alcohol use.
In the first part of this paper I explore some uses and motivations for analyses of sociodemographic differences in alcohol problems, and I search for answers to questions such as What kind of relevant research questions can be illuminated by studying sociodemographic differences in drinking and drinking problems? What scientific or other discussions can the researcher then participate in? Only the imagination sets limits to productive uses for data on sociodemographic differences in drinking or drinking problems. Therefore this discussion does not aim to make an exhaustive list of different possibilities, but rather gives illuminating examples. In the second part I discuss different approaches that researchers can take when trying to understand the nature and origin of differences in drinking between sociodemographic groups and to explain why the differences exist.
In many cases, sociodemographic variables are needed only for technical reasons, such as standardization or obtaining more homogeneous subgroups. For example, men and women are almost always considered separately, even when the gender difference is not the focus of interest, simply because men and women differ so much in their drinking habits that the mean values are not a good description of either group. …