Maternal education and adolescent drug use: a longitudinal analysis of causation and selection over a generation.

Current evidence indicates that in the USA illegal drug use among adolescents between the 1980s and 1990s became significantly more prevalent in families with lower maternal education in comparison to families with higher maternal education. In this study, we examine whether this inter-generational change results from either (a) a changing influence of socioeconomic status on drug use, as predicted by the inter-generational social 'causation' hypothesis, or (b) a negative influence of drug use on socioeconomic status, as predicted by the inter-generational social 'selection/drift' hypothesis. The analyses are based on the US National Longitudinal Study of 1979, which includes information on drug use for both a nationally representative sample of respondents aged 19-27 in 1984, as well as drug use information for the children of these respondents, who were aged 18-27 in 1998. The results indicate that inter-generation change in cocaine and marijuana use resulted almost entirely from social causation. These findings support illegal drug use as a good candidate for analyses in the 'fundamental cause' tradition that seek to understand the social factors that concentrate poor health and health behaviors in the lower social strata over historical time.