|
Article Excerpt
Incarceration for drug-law violations in the United States has grown tenfold since 1980 (Blumstein and Beck, 1999; Greenberg and West, 2001; Caulkins and Chandler, in submission), contributing substantially to the overall growth in U.S. incarceration rates (Langan, 1991; Beck, 1997; Greenberg and West, 2001), which now greatly exceed those of other Western countries (Austin and Irwin, 2001). This increase has been controversial on a number of dimensions (Human Rights Watch, 1997; King and Mauer, 2002), with considerable attention devoted to the issue of whether those who are incarcerated are mostly "kingpins" or nonviolent, low-level offenders (Sevigny and Caulkins, 2004). Here we examine the question of how many people are imprisoned in the U.S. for drug-law violations simply because they used drugs, not because they played some role in drug distribution or other offenses.
Resolving this question is crucial to understanding how U.S. drug policies differ from those in other countries (MacCoun and Reuter, 2001). There is little question that the United States is more punitive than most other countries with respect to drug-law violators (Kuhn, 2001), but there is considerable imprecision in understanding whether this primarily reflects supply-control efforts or sanctioning of drug users (ONDCP, 2005).
One might think this question has been answered by Bureau of Justice Statistics (BJS) publications based on inmate surveys that report the distribution of inmates by conviction offense. Table 1 excerpts the relevant proportions from the BJS report Correctional Populations in the United States, 1997 (BJS, 2000b, Table 4.3). Multiplying by the total number of inmates suggests that 38% of those imprisoned for drug-law violations in 1997 were convicted of simple drug possession.
However, serving time for a drug-possession conviction does not imply innocence of involvement in drug distribution or other offenses. Some possession convictions were for possession with intent to distribute, not simple possession. Even some inmates convicted of "simple possession" self-report participating in drug distribution. Still others were caught in possession of quantities greater than are typically associated with personal use. Such individuals can be excluded using more detailed questions from the same inmate surveys that underpin the BJS publications. As we show, doing so suggests that at most 15% of drug inmates are in prison for drug-law violations that did not relate to drug distribution--less than half the 38% figure obtained by naively assuming that conviction for possession is synonymous with incarceration for use only, not distribution.
Furthermore, many of the 15% had other characteristics suggesting that they might possibly, though not necessarily, have been involved in selling, including possession convictions that stemmed from a plea bargain to reduced charges, being arrested with types of drugs that they had not themselves used, and/or having a previous history of drug trafficking. Also, only about one-third of the 15% were new court commitments. Most were already on parole or probation, and nearly half (48%) of them had technical violations or concurrent convictions for things other than drug use or possession. After describing the origins of the 85%/15% split, we try to shed light on who comprises the 15% not clearly involved in drug distribution. We then explore some policy implications of the 15% figure.
To clarify, our goal is to determine how many drug users are in prison for drug possession convictions that are unrelated to drug distribution--e.g., that stem from personal consumption. Clearly many drug sellers also use drugs, but since drug distribution is a more serious offense than mere possession or use, we view such people as being in prison because of their involvement in drug distribution. Likewise, many drug users are incarcerated for other crimes, such as burglary or murder; we wish to exclude them too. So the number we seek is not the number of people who are in prison because of drug use; that number is clearly much larger. Rather, the number we seek can be better thought of as reflecting the number of people in U.S. prisons who would not be there (at least for their present charge) if criminal penalties for drug use were not enforced. Thus it is a measure of the difference between enforcement directed at the "demand side" of the market in the United States as compared with countries such as the Netherlands that eschew using prison to punish drug use (MacCoun and Reuter, 2001). Conversely, comparing this number with total drug-related incarceration is relevant to assessments of the extent to which U.S. drug enforcement is directed primarily at the "supply side" of the market. The number we estimate also gauges potential costs or savings associated with varying the imprisonment of users. It suggests an upper bound on how much could be saved from reduced imprisonment by decriminalizing simple drug possession. Conversely, it is germane to estimates of how much would have to be spent to increase the enforcement risk faced by users by a certain proportion. Finally, it is an essential component of estimates of the sanction risk faced by drug users because of their use.
An important limitation of these estimates is that they pertain only to adult inmates in prison. Incarceration in local jails and juvenile corrections facilities is not considered. A second inherent limitation is that the data are based on self-report. However, this analysis still improves our currently imperfect understanding of important aspects of punishment for drug-law violators in the United States.
Data and methods
The Survey of Inmates in State and Federal Correctional Facilities, 1997 (BJS, 2000a) is a nationally representative survey of federal and state inmates conducted regularly by the Bureau of Justice Statistics (BJS). It collects self-report data on a wide array of offender characteristics, including current offense and sentencing information, criminal history, socioeconomic status, incident characteristics, alcohol and drug use and treatment history, health status, and conditions of confinement. The 1997 survey is the most current available in a series that has been fielded every five to six years since 1974 (since 1991 for federal inmates). Between June and October 1997, 4,041 federal inmates and 14,285 state inmates were interviewed using computer-assisted personal interviewing methods. Inmates were selected according to a two-stage stratified sampling plan from a universe of 1,409 state prisons and 127 federal prisons. Region/state by gender and administrative level by gender strata were formed at the state and the federal levels, respectively. Within each stratum, sampling was performed, first, by selecting prisons (i.e., the primary sampling unit) and, second, by selecting inmates in the sampled prisons (BJS, 2000c). Total nonresponse rates were 7.8% and 9.8% at the state and the federal level, respectively.
We focus on the 4,787 sentenced offenders with...
|
