A Longitudinal Study of Alcohol and Drug Use in the Workplace

Abstract

Alcohol abuse and illicit drug use in the United States are major concerns of American households, as well as of the White House. This dissertation research evaluates alcohol abuse and controlled drug use by American workers in the context of various individual, organizational, and occupational settings. It tests the importation and organizational stress perspectives, the occupation subculture perspective, and the lifecycle wage compensation theory. The analyses are developed utilizing (1) logistic regression, (2) generalized linear modeling, including Poisson regression and negative binomial regression, (3) weighted modeling estimation, taking the clustering effects of complex survey design into account, and (4) the hierarchical growth curve modeling of intra- and inter-individual differences. Using data from the National Longitudinal Survey of Youth 1979â 1993, the 1997 National Household Survey on Drug Abuse, and the 1998 National Occupational Information Network (O*NET 98), I find that employees' drinking and controlled drug use behavior are predicted by a number of individual background characteristics, as well as workplace-environment variables. I also find that occupational characteristics influence alcohol and drug using behaviors of workers, although in more complex ways than suggested by much of the organizational stress and occupational subculture literature. It appears that occupations with higher levels of steady employment prospects exert the most significant negative effect on employees' alcohol use, marijuana use, and any illicit drug use, regardless of an employee's age, gender, race, education, and income. It also appears that the etiology of cocaine use is different from that of either alcohol use or other drugs, such as marijuana. Finally, I find that when education and years employed are held constant, employees' current marijuana use is negatively associated with their earnings. No evidence has been found that current alcohol use, current marijuana use, or lifetime cocaine use predicts future growth rates on earnings. Having examined the factors of occupational, organizational, and individual social/demographic characteristics as they influence patterns of alcohol abuse and controlled drug use in multiple large representative samples of the labor force, discussions on the research findings, the implications, the limitations, and the future study directions are presented.