Addressing Temporal Order
Establishing temporal order is one of the greatest challenges facing those who study drugs and crime. Most longitudinal studies indicate that drug-using offenders initiate delinquency and more serious forms of offending before they begin use of substances, especially illegal drugs (Anglin & Speckart, 1988; Brennan et al., 1981; Elliott et al., 1989; Farabee et al., 2001; Inciardi, 1990; Van Kammen & Loeber, 1994). If delinquent and criminal activity precedes substance use, then we cannot claim that drug use causes the onset of offending. However, it is possible that substance use exacerbates or maintains levels of offending, which is still a valid form of probabilistic causality. Proper identification of these relationships requires longitudinal data and analysis.
We pulled studies that established temporal order—where measures of drug use preceded measures of offending—and the result was, in fact, a long list of studies with unique and contradictory findings, with too few measuring the same things to draw any conclusions about any specific relationships. Some studies indicated that consumption of substances significantly increases the likelihood that a person will subsequently offend at a later time (DAmico et al., 2008; Dembo et al., 1991; Kandel et al., 1986; Van Kammen & Loeber, 1994) or offend more frequently while using alcohol and/or drugs than during periods of abstention (De Li et al., 2000; Gottfredson et al., 2008; Horney et al., 1995). However, the nature of the differential relationships varied from study to study, both in terms of the criminogenic substance (alcohol and/or illicit drugs) and in terms of the kind of offending that becomes more likely to occur (violence and/or theft). There were many contradictions and no emergent pattern. At best, these studies indicate that more longitudinal, dual-dependent variable studies are needed.