Alcohol Availability and Domestic Violence
According to the distribution of consumption model, increases in per capita alcohol consumption predict increases in the rate of heavy use and alcohol-related health and social consequences (1). Alcohol availability, in mm, is considered a key influence on alcohol sales and consumption (2, 3). Ecological studies of the relationship among alcohol availability, consumption, and consequences have been conducted at various levels of analysis, ranging from city blocks, to municipalities and counties, to states. Few researchers argue that availability alone explains the variation in rates of alcohol-related problems among given geographic units; rather, the issue is whether it adds substantially to the amount of variance in such rates accounted for by sociodemographic variables such as age and gender structure, ethnicity, and socioeconomic status. On the one hand, a number of studies have shown that the physical availability of alcohol, usually measured in terms of outlet density or laws and ordinances governing sales, explains a significant amount of additional variance in alcohol-related problems such as motor vehicle crashes, liver disease, and violence (4-7). In contrast, others have found that alcohol outlet density adds little to the explanatory power of models comprised of sociodemographic variables (8-12).
The social and health costs of alcohol use and abuse in the United States are considerable (13). Interest in access and availability, and the development of strategies to affect these, have arisen in recent years as individual-level programs and intervention tend to have little influence on alcohol use (14, 15). The present study focuses on a presumed consequence of alcohol use not considered in previous ecological studies of alcohol availability–namely, domestic violence. Individual-level analyses, based on clinical and survey research, show that heavy alcohol use is associated with increased spousal abuse (16-19). However, as Leonard observes (18), there exists a general paucity of methodologically sound and conceptually driven research that examines the role of alcohol in domestic violence, and we know of no other macrolevel analysis that examines this relationship.
In this study, we examine the relationship among sociodemographic factors, alcohol availability (measured in terms of outlet density), and domestic violence. Specifically, we test whether alcohol outlet density adds significantly to a sociodemographic model to explain variation in rates of domestic violence across the 223 largest municipalities in the state of New Jersey. The sociodemographic variables included in the model have been shown in previous research to be associated with higher rates of domestic and other forms of family violence (20-22). These variables include age, family, ethnic and gender composition, poverty level, and population movement. Conceptually, they are derived from social disorganization theory, with the hypothesis that communities that are more socially disorganized will exhibit higher rates of domestic violence, just as they display higher rates of other social problems such as violent crime and child neglect (20, 23, 24).
METHODS
Data were collected for the 223 municipalities in the state of New Jersey with populations greater than 10,000 in the year 1990. The largest municipality in the sample had a population of 275,221 and the smallest a population of 10,074. The mean population size across the 223 municipalities was 27,931 (SD 29,057). Means and standard deviations of study variables are shown in Table 1 and are described below.
Table 1. Means and Standard Deviations for Sociodemographic Variables, Alcohol Availability, and Domestic Violence Rates for the 223 New Jersey Municipalities
Variable Mean SD
Age, family, and gender composition
< 11 years (%) 15.3 2.8
12-17 years (%) 7.3 1.5
18-30 years (%) 18.2 4.8
> 50 years (%) 28.2 7.3
Children per 100 adults 21.1 5.0
Males per 100 females aged 21-64 95.7 5.3
Female-headed households (%) 4.5 3.2
Educational attainment (%)
< High school education 14.1 6.7
University graduates 18.5 9.6
Ethnic composition (%)
Black 9.0 14.6
Latino 6.8 10.6
Poverty
Unemployed (%) 5.1 2.4
Public assistance (%) 4.3 3.6
Median household income ($) 45,646 13,383
Population movement (%)
Current household < 10 years 21.1 4.5
Moved between 1985 and 1990 15.7 4.1
Moved between 1989 and 1990 5.5 2.2
Urbanicity
Population density (per square mile) 4,515 5,270
Alcohol availability
Total outlets per 10,000 population 12.3 8.1
Domestic violence rate (per 10,000 population)
1990 63.5 52.8
1991 70.7 57.9
1992 64.3 54.0
1993 84.7 68.5
1994 92.0 74.7
1995 112.2 91.7
Mean 1990-1995 81.2 63.7