In 1991, questions on involvement in criminal behavior and being arrested and booked for a crime were added to the National Household Survey on Drug Abuse (NHSDA) to ascertain the relationship between drug use and criminal behavior. Analysis shows that drug use is a strong correlate of being booked for a criminal offense, but age is the more important correlate of criminal involvement. There were few differences in models predicting violent as opposed to property crime, although minority status was a more important predictor of violent crime, and poverty was a more important predictor of property crime. Cocaine use was the most important covariate of being booked for a crime in large metropolitan areas that were oversampled in the 1991 NHSDA.
Assessing measurement error in surveys using latent class analysis: application to self-reported illicit drug use in data from the Iranian Mental Health Survey