Recent sentencing proposals for the selective incapacitation of criminal offenders have generated a great deal of enthusiasm and controversy. The concept has been greeted enthusiastically because it promises simultaneously to decrease the crime rate and to reduce crowding in the nation's prisons. The controversy stems from two sources: concerns of science and of ethics. This article describes the selective incapacitation proposal and the scientific and ethical controversies it has generated. Finally, an alternative strategy for using risk predictions is presented. It is thought to meliorate the ethical concerns discussed and to hold promise for reducing prison crowding without endangering the public.
Get full access to this article
View all access options for this article.
References
1.
1. Peter W. Greenwood with Allan Abrahamse, Selective Incapacitation (Santa Monica, CA: Rand, 1982).
2.
“Key to Criminals' Futures: Their Pasts,”U.S. News and World Report, Oct. 1982.
3.
Andrew von Hirsch and Don M. Gottfredson, “Selective Incapacitation: Some Queries about Research Design and Equity,”New York University Review of Law and Social Change, 12(1): 11-51 (1983-84).
4.
Gerhard O. W. Mueller , Sentencing: Process and Purpose (Springfield, IL: Charles C Thomas, 1977); B. A. Grossman, ed., New Directions in Sentencing (Toronto: Butterworth, 1980).
5.
Steven Garber , Steven Klepper, and Daniel Nagin, “The Role of Extralegal Factors in Determining Criminal Case Dispositions,” in Research on Sentencing, ed. National Research Council Panel on Sentencing Research and Blumstein; Steven Klepper, Daniel Nagin, and Luke-Jon Tierney, “Discrimination in the Criminal Justice System: A Critical Appraisal of the Literature,” in ibid.
6.
6. Other less commonly cited goals such as retribution or retaliation also do not require prediction.
7.
7. In this exercise, the judges were asked to distribute 100 points among several commonly cited purposes—or to assign this value to any single purpose. The only constraint imposed was that the total points assigned summed to 100. We are currently preparing a more complete report of this study.
8.
8. Stephen D. Gottfredson and Ralph B. Taylor, The Crisis in Corrections: Prison Populations and Public Policy (Washington, DC: Department of Justice, National Institute of Justice, 1983).
9.
9. Margaret M. Hale, “The Influence of Sentencing Goals on Judicial Decision-Making” (Ph.D. diss., Johns Hopkins University, 1984).
10.
10. Cohen, “Incapacitation as a Strategy for Crime Control”; idem, “Incapacitating Criminals.”
11.
11. Cohen, “Incapacitation as a Strategy for Crime Control,” p. 12.
12.
12. Ibid., pp. 23, 30.
13.
Cohen , “Incapacitation as a Strategy for Crime Control”; Greenwood with Abrahamse, Selective Incapacitation.
14.
14. See, for example, Cohen, “Incapacitation as a Strategy for Crime Control”; von Hirsch and Gottfredson, “Selective Incapacitation.”
15.
15. It must be noted, however, that although the nature of the prediction problem is somewhat different in the aggregate approach, prediction is involved nonetheless. See Cohen, “Incapacitation as a Strategy for Crime Control,” pp. 73 ff.
16.
16. Our discussion of this research—Greenwood with Abrahamse, Selective Incapacitation— is intentionally nontechnical. Detailed technical critical reviews of this research are available in Cohen, “Incapacitation as a Strategy for Crime Control”; and in von Hirsch and Gottfredson, “Selective Incapacitation.”
17.
17. Greenwood with Abrahamse, Selective Incapacitation.
18.
18. In a prediction study, that which is to be predicted is known as the criterion or dependent variable; items of information used to make the prediction are known as predictor or independent variables.
19.
19. Cohen, “Incapacitation as a Strategy for Crime Control”; von Hirsch and Gottfredson, “Selective Incapacitation.”
20.
20. Cohen, “Incapacitation as a Strategy for Crime Control,” pp. 48-49.
21.
21. Ibid., p. 50.
22.
22. This is because the apparent power of a prediction device derives from two sources: the detection and estimation of underlying relations likely to be observed in any similar sample of subjects, and the peculiar or individual properties of the specific sample on which the device has been constructed. Within the original sample alone, there is no adequate way to distinguish how much of the observed relation is due to characteristics and underlying associations that will be shared by new samples and how much is due to characteristics of the first sample only. When applied to new samples, apparent predictive accuracy usually is reduced relative to that observed in the original sample.
23.
23. Stephen D. Gottfredson and Don M. Gottfredson, Screening for Risk: A Comparison of Methods (Washington, DC: Department of Justice, National Institute of Corrections, 1979).
24.
and in “Selective Incapacitation: Two Views of a Compelling Concept.”
25.
25. We consider utilitarian sentencing purposes to include those that have a crime-reductive intent. As noted earlier, it is these that require prediction.
26.
26. Stephen D. Gottfredson and Don M. Gottfredson, “The Accuracy of Prediction Models” (Paper prepared for the meeting of the National Academy of Sciences' Panel on Criminal Careers, Woods Hole, MA, July 1984).
27.
27. Ibid.
28.
28. These too will be difficult to achieve, even if desired. Many complex issues of measurement remain before the goal of ensuring desert could be met adequately.
29.
Robyn M. Dawes , “The Robust Beauty of Improper Linear Models in Decision Making,”American Psychologist, 34(7):571-582 (1979).
30.
30. Hillel J. Einhorn and Stephen Schacht, “Decisions Based on Fallible Clinical Judgment,” in Human Judgment, ed. Kaplan and Schwartz, pp. 126-144.
31.
31. Joan Petersilia, “Criminal Career Research: A Review of Recent Evidence,” in Crime and Justice: An Annual Review of Research, ed. Norval Morris and Michael Tonry (Chicago: University of Chicago Press, 1980), 2:322.
32.
32. Cohen, “Incapacitation as a Strategy for Crime Control,” p. 49.
33.
33. As analogies: income may be measured in dollars and infinite fractions thereof; age may be measured in days, months, or years, and infinite fractions thereof.
34.
34. In the figure drawn, both measures are represented in standardized form. This simply equates the arithmetic means and the standard deviations—a standard deviation being a common measure of variability—of the two variables.
35.
35. The base rate for any given event is defined as the relative frequency of occurrence of that event in the population of interest. One serious consequence for prediction is that prediction is made greatly more difficult as the base-rate differs, in either direction, from 0.50.
36.
36. We assume that the criterion cutting score—Yc—represents a standard, perhaps identified by societal consensus, judicial policy, or some other indicator. The standard could, of course, be changed; this too could have consequences for the ratio of false positives to false negatives.
37.
37. Other criteria could of course be used. For example, we may choose to release those least deserving of punishment.
38.
38. Robert A. Gordon, “A Critique of the Evaluation of the Patuxent Institution, with Particular Attention to the Issues of Dangerousness and Recidivism,”Bulletin of the American Academy of Psychiatry and the Law, 5(2):210-255 (1977).
39.
39. We do not claim that all original purposes necessarily are appropriate. We simply point out that the scheme appears to be atheoretical with respect to them.
40.
40. See von Hirsch and Gottfredson, “Selective Incapacitation.”
41.
41. Rolf Loeber and T. J. Dishion, “Early Predictors of Male Delinquency: A Review,”Psychological Bulletin, 94(1):68-99 (1983).
42.
42. We do not argue that this position necessarily has been advanced by the proponents of the strategy.
43.
43. Stephen D. Gottfredson, “Institutional Responses to Prison Crowding,”New York University Review of Law and Social Change, 12(1): 259-273 (1981-84).
44.
44. Marvin D. Dunnette, Personnel Selection and Placement (Belmont, CA: Brooks/Cole, 1966), pp. 173-183.
