Ross, 1984). Since both social and legal sanctions for these crimes are relatively less severe than the penalties for opiate or cocaine offenses, the findings have limited generalizability. Deterrence research on opiate and cocaine offenses must also take account of the significant overlap between use offenses and distribution offenses. Most users are not dealers, but some become involved in dealing to support their habit.

Finally, most empirical studies on the general deterrent effects of law and social control have proceeded on a separate track from studies on the specific deterrent effects of punishment. This bifurcation of the empirical literature has led some researchers to suggest a revised, “perceptual deterrence” framework that incorporates both direct (arrests, incarceration) and indirect (friends’ and acquaintances’ experiences) punishment experiences in the conceptual model (Stafford and Warr, 1993).

As recommended by a study by the Institute of Medicine (1996), a new generation of research on the deterrence of drug use should be based on a model that integrates legal deterrence into a social control framework. This model would encompass a broad range of elements relating to the perceived costs and benefits of drug use. Such elements include economic costs (e.g., money and search time), personal costs (e.g., risks of dependence, disease, and violence); social, physiological, and psychological returns (e.g., pleasures, status, lifestyle); actual and perceived direct costs of punishment (e.g., arrest, incarceration, loss of income or drugs); social costs of punishment (e.g., job or relationship loss; see Williams and Hawkins, 1989); and motivational components (e.g., risk taking and sensation seeking).

A research agenda on deterrence should also recognize the distinctions in deterrent effects across populations of drug users and in different sectors of society. Research should also take adequate account of the balance of motivations and restraints on drug use, including both external restraints from threatened legal sanctions and internal restraints reflecting social and moral inhibitions. The threat of punishment carries different weight for different people, depending on their personal circumstances.

Differences in the effects of legal controls on illegal behaviors may reflect not only individual factors, but also the effects of contextual variables that either strengthen or neutralize the effects of legal controls—for example, by increasing the returns from drug use (or drug dealing) or by discounting the social costs of arrest and punishment. Many of these factors reflect the structure of opportunities and controls at the neighborhood or community level. In some cases, neighborhood effects powerfully reinforce legal deterrents to drug use. In other cases, neighborhood effects can delegitimize law and reinforce involvement with drugs (Tonry, 1995).