Download this article now for $27.00.
Drinking Patterns in the American Deep South
Christine Lindquist, William C. Cockerham
Objective: To determine whether the states of the Deep South are characterized by contradictory drinking norms: high abstinence paired with high consumption among drinkers. Method: Comparison of survey results of drinking patterns (N = 65,216) in four states of the Deep South (Alabama, Georgia, Louisiana and Mississippi) to those in leading states in the East (New York), Midwest (Illinois), Rocky Mountains (Colorado) and West (California) for the years 1990-93. Respondents were selected by random-digit dialing and interviewed by telephone as part of a larger study conducted by the Behavioral Risk Factor Surveillance System (BRFSS) of the Centers for Disease Control and Prevention (CDC). Results: A higher proportion of persons drank alcohol in 1993 than in any of the previous 3 years. Persons with higher education, with higher income, of younger age, male, white and not living in the Deep South were more likely to drink. Among drinkers, the likelihood of episodic heavy drinking decreased with higher education, income and age. In addition, women, blacks and married respondents were less likely to report heavy drinking. Region of residence is not significantly associated with episodic heavy drinking. Conclusions: The Deep South has a significantly higher proportion of abstainers from alcohol use than other regions. Among drinkers, Southerners are not more likely to engage in occasions of heavy drinking. The results are discussed in relation to Southern culture and drinking norms. (J. Stud. Alcohol 60: 663-666, 1999) ">Department of Sociology, University of Alabama at Birmingham, 237 Ullman Building, 1212 University Boulevard, Birmingham, Alabama 35294-3350 CULTURAL DIFFERENCES in alcohol use are an un- doubted social fact (Bales, 1946; Room, 1976). These differences exist not only between particular social groups but between regions (Hilton, 1988). In the United States, for example, the South has the highest percentage of ab- stainers (DuFour, 1995; Herd, 1990; Midanik and Clark, 1994; Room, 1983) and the lowest rates of alcoholism (Courtwright, 1988) of any region of the country. The South has been characterized as overwhelmingly Protestant, rural and dry (Courtwright, 1988) and it has a proportion of ab- stainers larger than that of the Mountain states and double that of the Northeast, Midwest and Pacific regions (DuFour, 1995; Herd, 1990). In some states of the Deep South, over half the adult population abstains from drinking (Hilton, 1988; Midanik and Clark, 1994). Yet Room (1983) claims that the prevalence of heavy drinking among drinkers is greater in drier regions than in wetter regions; i.e., while a re- gion like the South has fewer drinkers per capita, those per- sons who do drink may drink more heavily than drinkers in regions where alcohol consumption is more widespread. This would suggest that Southern states, especially the states of the Deep South, are characterized by contradictory drinking norms: high abstinence paired with high consump- tion among drinkers. It is the purpose of this article to deter- Received: September 16, 1997. Revision: January 27, 1998. * Support for this research was provided by the Centers for Disease Con- trol and Prevention (CDC), Atlanta, Georgia. The research is based on data collected by CDC's Behavioral Risk Factor Surveillance System (BRFSS). The conclusions are those of the authors. Correspondence should be sent to William C. Cockerham at the above address. mine whether this pattern of alcohol use was typical of the Deep South in the early 1990s and if Southern drinkers do in- deed drink more heavily than their counterparts in wetter re- gions. The question of whether drinkers in dry regions drink more heavily than drinkers in wet regions has been explored but findings are inconclusive (Skolnick, 1958). Although some studies support higher levels of problem drinking in dry regions (Hilton, 1988), others find no evidence of such a pattern (Klein and Pittman, 1993) or find heavier drinking in the dry South restricted to particular groups—such as black men (Herd, 1990). Consequently, our research is intended to determine both the current pattern of drinking in the Deep South and the relationship between this drinking pattern and episodic heavy drinking. This being an, as yet, unresolved is- sue, relevant findings can help shape intervention measures designed to reduce problems associated with alcohol use. Method Data were collected by telephone in a nationwide survey conducted by the Behavioral Risk Factor Surveillance Sys- tem (BRFSS) of the Centers for Disease Control and Pre- vention (CDC) (BRFSS, 1997). Respondent households were selected by random-digit dialing, with one adult aged 18 or over within the household randomly selected to be in- terviewed. Telephone interviews were conducted monthly in order to provide seasonally adjusted annual estimates (Rem- ington et al., 1988). In the present study, as the major re- search question involves regional differences in alcohol consumption and drinking behavior, two sets of states were included in the analysis. The first, the Deep South, comprises Alabama, Georgia, Louisiana and Mississippi. The second set comprises leading states of four other regions: New York (Northeast), Illinois (Midwest), Colorado (Mountain) and California (Pacific). The latter set of states represent four di- verse regions, culturally distinct from the Deep South, for purposes of comparison. Each contains one or more major metropolitan areas but also has a sizeable rural population and a significant economy. In order to examine temporal trends, in addition to regional trends, data for a 4-year period (1990-93) were included in the analysis. The initial sample consisted of 65,216 respondents; however, missing data, deletion of nondrinkers on alcohol use items, deletion of His- panics and other minorities on comparisons of blacks and whites, refusals or "did not know" on income, etc. reduced the total number of cases available for analysis for particular variables. The number of cases for each variable is shown in Table 1. The two dependent variables are alcohol use and episodic heavy drinking. These outcomes are dichotomous, indicating simply whether or not the respondent has engaged in the be- TABLE 1. Distribution of dependent and independent variables among the sample (N = 65,216) Dependent variables Alcohol use (drank in past month) Episodic heavy drinking (drank 5 + drinks) n 30,745 7,953 % (47.2) (26.3) Independent variables Year 1990 14,008 (21.5) 1991 15,693 (24.1) 1992 17,714 (27.2) 1993 17,801 (27.2) Education 8th grade or less 4,238 (6.5) Some high school 6,699 (10.3) High school graduate or GED 20,323 (31.2) Some college or technical school 17,103 (26.3) College graduate or postgraduate 16,735 (25.7) Income < 10,000 10,278 (17.4) 10,000-14,999 6,574 (11.1) 15,000-19,999 5,919 (10.0) 20,000-24,999 6,090 (10.3) 25,000-34,999 9,739 (16.5) 35,000-49,000 9,657 (16.3) 50,000+ 10,825 (18.3) Age 18-24 6,896 (10.6) 25-34 15,332 (23.5) 35-44 14,703 (22.5) 45-54 9,620 (14.8) 55-64 7,935 (12.2) 65+ 10,730 (16.5) Gender Men 27,478 (42.1) Women 37,738 (57.9) Marital status Single, separated, divorced, widowed 30,888 (47.5) Married 34,215 (52.6) Race White 48,179 (83.2) Black 9,714 (16.8) Region Deep South 28,418 (43.6) Other 36,798 (56.4) havior. Alcohol use indicates whether or not the respondent reported having had a drink within the past month. All re- spondents were asked: "Have you had any beer, wine, wine coolers, cocktails or liquor during the past month, that is since------." Of the entire sample, 47.2% reported alcohol use within the previous month. This measurement of alcohol use may underestimate the prevalence of drinking due to both the social desirability bias inherent in self-report mea- sures of alcohol use and the 1-month recall period to which respondents were limited. However, this is a simple question, easy to recall and able to establish whether or not a person has had a drink within the time frame. Additional information regarding alcohol use was ob- tained only from respondents who reported having had at least one drink during the past month; thus, the second out- come variable includes drinkers only. Episodic heavy drink- ing indicates whether or not the respondent reported drinking more than five drinks on one occasion within the past month. Although the wording of the question ("Considering all types of alcoholic beverages, how many times during the past month did you have five or more drinks on an occasion?") was designed to obtain interval-level responses, the highly skewed distribution necessitated dichotomizing the re- sponses into "zero" and