Prevalence of frequent cocaine use in urban poverty areas

Abstract

Many authors have noted that in national survey data cocaine use is more or less equally prevalent in all socioeconomic strata (Adams and Gfroerer 1991, Anthony 1992, Flewelling et al. 1992, O'Malley et al. 1991, Parker 1995, Robins and Przybeck 1985, Ritter and Anthony 1991, Trinkoff et al. 1990, Windle and Miller-Tutzauer 1991). A broader review of the heterogeneous data bearing on the question of use prevalence indicates that cocaine use (weekly or more often) is far more prevalent in poverty than elsewhere. Our review is motivated by consistent ethnographic reports of cocaine use among the poor (Booth et al. 1993, Chirgwin et al. 1991, Edlin et al. 1994, Fullilove et al. 1990, Hamid 1992, Hunt 1991, Inciardi 1986 and 1991, Krohn and Thornberry 1993, Lewis et. al. 1992, Marx et al. 1991, Ratner 1992, Rodriguez et al. 1993, Tidwell 1992, Weppner 1977). For example, studies in Ratner (1992) portray poor addicts who spend several days and nights repeatedly exchanging sex for a hit or rock of crack (or for the price of a rock) and then sleep for a few days to begin again.' Edlin et al. (1994) evaluated an street sample of 1,137 crack smokers in which the median user used 10 times per day and 28 days out of the last 30. Thirty-nine percent of the women smokers had had more than 50 sex partners, and in New York and Miami, respectively, 29.6% and 23.0% were HIV positive. A focus on prevalence in poverty areas, as opposed to prevalence among poor individuals, is appropriate because concentrations of poverty create special dynamics. The collocation of a large group of occupationally limited adults deprives children of necessary role models and deprives adults of networks supporting access to advancement. Persons residing in such collocations face a high risk of slipping into poverty, crime and dependency, even if they are currently lawfully employed and are sustaining themselves above the poverty line (Kasarda 1992, Wilson 1987). A focus on areas, as opposed to individuals, is particularly appropriate in relation to drug use, because drug use is often transmitted from person to person within social networks (Moore 1977). The social networks of drug users are more often neighborhood-based than institution-based (Krohn and Thornberry 1993). The Census Bureau defines poverty as census tracts where more than 20% of the population was poor in 1990. There is, however, a necessary imprecision in our reference to urban poverty areas in this paper. First, poverty include varying concentrations of poor individuals, averaging 39.5% (Bureau of the Census 1993a). Second, the data on drug abuse come from a number of very different sources, and there is no single analytic frame of reference within which all the sources are commensurable. The National Household Surveys define frequent users of cocaine as weekly or more users (e.g., Substance Abuse and Mental Health Services Administration 1995b). The term frequent conflates regular weekend recreational users with thoroughly addicted users consuming more than a gram per day. This imprecision is also necessary; the data do not effectively distinguish the varieties of frequent users. It is appropriate, however, to focus on users at the end of the spectrum. These are the users sustaining the greatest personal damage and causing the greatest community damage. They also account for the bulk of the cocaine consumed.2 I. Assessment of the data There are four types of data bearing on the relative prevalence of cocaine use in poverty and other areas: smallarea event-driven data, small-area survey data, national event-driven data and national survey data. By eventdriven data we mean data generated at the time of a life event (arrest, pregnancy, admission to an emergency room, death). A. Small-area event-driven data Small-area health-care studies show widely varying local prevalence of cocaine use. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.