Mandatory Minimum Penalties: Their Effects on Crime, Sentencing Disparities, and Justice System Expenditures

4. Can offenders be deterred? (continued)

4. Can offenders be deterred? (continued)

4.3 Evidence on Deterrence and Incapacitation (continued)

4.3.2 Incapacitation

At first glance, estimating the crime preventive effects of incapacitating offenders appears simple enough. The average annual number of offences per offender must be multiplied by the number of people in custody. Unfortunately, the problem is complex, owing in part to enormous variations in the estimation of individual offending rates. Estimates using arrest records range between 8 and 14 index crimes per year on the street, while estimates from self-report studies demonstrate a greater variation still. Robbers are estimated to average between 5 and 75 robberies per year and burglars are estimated to commit an average of 14 to 50 burglaries per year (Nagin, 1999). Zimring and Hawkins (1995) report that estimates of the number of crimes prevented per year in prison ranges from 3 to 187. Even the low-end estimates suggest that incapacitation prevents large numbers of crimes.

Nagin points out, however, that the averages are skewed by a small number of individuals who commit extraordinarily high numbers of offences. Visher (1986), for example, found that five percent of robbers committed 180 or more robberies per year. Thus, incapacitating the remaining 90%-95% of offenders will yield a much lower crime preventive benefit than that suggested by the mean offence rate. Then there are issues of diminishing returns with age, offender replacement in relation to offences meeting a demand (e.g., drug dealing) and those committed in groups, and the possibility that incapacitation merely delays rather than disrupts criminal careers (Vitiello, 1997).

It has been estimated that, in 1975, 32.9% of potential violent offences were prevented nationally (in the US) due to incapacitation alone (Cohen and Canela-Cacho, 1994). By 1989, the near 200% increase in prison inmates, according to these authors, had prevented just an additional 9% of violent offences, for a total of 41.9% of potential offences. The marginal benefits of imprisonment decline with age and once most of the high rate offenders have been confined. A study applying a mathematical model estimating the effects of incarceration has corroborated this point by finding that crime levels are quite insensitive to the size of the prison population (Greene, 1988).

The most cost effective method of incapacitation would involve the allocation of prison resources more selectively, through the early identification of the most active offender group--selective incapacitation. Selective incapacitation, however, has drawn fierce criticism on both ethical and pragmatic grounds. From an ethical standpoint, sentencing exclusively according to risk is said to violate fundamental legal principles by allocating punishment on the basis of expected future behaviour (Dershowitz, 1973), undermines proportionality in sentencing (Capune, 1988), and would likely be applied disproportionately to various social groups (e.g., minorities, the young) (Capune, 1988; Gabor, 1985).

Practically speaking, the prediction enterprise is essentially unable to identify high-rate career offenders prospectively (in advance) (Petersilia, 1980; Greenberg and Larkin, 1998). Also, studies attempting to predict offender risk tend to have false positive rates (i.e., over-predictions of dangerousness) of over 50% (Auerhahn, 1999). Furthemore, it is believed that many high-rate offenders are already incapacitated under existing sentencing practices, thereby limiting the potential gains under a selective incapacitation policy (Beres and Griffith, 1998; Nagin, 1998). In any event, MMS tend to adopt a collective rather than selective incapacitation approach, as they are usually triggered by specific or repeat offences, rather than offender-specific risk-related attributes.

Finite prison space has implications for both incapacitation and deterrent effects. As it is difficult to simultaneously increase both the rate of incarceration and the length of prison sentences, the respective benefits of each must be considered. One study has found that a sentencing regime involving shorter sentences (one year) for a large offender base (all convicted persons) is more efficient than one involving longer sentences (four years) applied to a smaller offender base (repeat felons) (Petersilia and Greenwood, 1978). The crime preventive yield of both strategies was estimated to be the same, although the cost of the latter strategy would be far more substantial in terms of the size of the prison population.