Reports on the prevalence and incidence drug use commonly list separate estimates for various drugs by race/ethnicity. However, this can be somewhat misleading because reporting results in this way can infer that race or ethnicity by themselves are important factors in drug use. Several researchers recently have advocated that race/ethnicity be used as markers for other factors that might be related to race, such as socio-economic indicators, neighborhood characteristics, cultural variables, and impact of racism or discrimination. For this reason, this project has investigated the meaning of racial/ethnic differences in crack use using data from the National Household Survey on Drug Abuse (NHSDA). Crack is one of the only drugs for which there are differences by race/ethnicity. In the 1990 NHSDA, the lifetime prevalence of crack use for African Americans was 3.1% compared to 1.6 % for Hispanics and 1.1% for Whites. However, in an analysis that held neighborhood constant, there was no overall difference between African- and White-Americans; the odds of crack use for Hispanics was only half that for Whites. An investigation of age by race interactions indicated that, even when neighborhood was held constant, the odds of lifetime crack use for African-Americans 30 - 35 years old was 2.5 that of White-Americans in the same age group. Further, the African-American crack users were more likely to have used in the last year. An examination of the trends in crack use indicated that lifetime prevalence of crack use for African- Americans increased from 5% in 1988 to 8% in 1990, whereas there was little change for other age groups or race/ethnicities. A parallel analysis was conducted that examined only Hispanics, using data from the 1988, 1990, and 1991 NHSDA and employing a post- stratification strategy that held neighborhood constant. The respondent's choice of the Spanish version of the survey instrument was used as an indicator of degree of acculturation. This analysis indicated that the odds of crack use for Hispanics who used the Spanish version of the survey instrument was approximately one-third that of those using the English version. An analysis of the interaction between language version and Hispanic origin indicated that this finding was most pronounced among Mexican-Americans, where the odds of crack use for those taking the Spanish version was one-sixth that for the English version. These findings point to the importance of degree of acculturation in initiating crack use.
