Brandon C. Welsh
The University of Massachusetts, Lowell
Economic Analysis
Results from Cost-Benefit Analysis Studies
Implications for Offender Reentry
Conclusion
WHILE NO ONE should really be surprised by the vast numbers of offenders returning to communities from prison each year in this country (see Latessa, 2004), a number of key factors associated with the present situation of offender reentry are cause for concern. Indeed, it is these factors that underlie concerns that offender reentry may soon contribute to an increase in national crime rates. Some of these factors include: fewer treatment resources for offenders while in prison, particularly for those in greatest need; the abolition or reduction of parole, which takes away incentives for prisoners to participate in treatment; and less transitional aid (e.g., employment, housing) for prisoners (Petersilia, 2003).
This paper takes as its starting point the current offender reentry problem of diminishing treatment resources. It aims to assess if there is an economic argument for correctional treatment and explores implications for offender reentry. To do so, it updates previous reviews of cost-benefit analyses of correctional treatment programs (Welsh & Farrington, 2000a, 2000b). These reviews provide some evidence that correctional treatment is a worthwhile or economically efficient approach to reducing reoffending in the community. It is the position of this paper that if the monetary benefits of correctional treatment programs outweigh their costs, this may be a persuasive argument for increasing treatment resources for offenders. Moreover, this cost-benefit argument may go some way toward addressing some of the pressing concerns facing offender reentry.
Economic Analysis
An economic analysis (such as a cost-benefit analysis or a cost-effectiveness analysis) can be described as a policy tool that allows choices to be made between alternative uses of resources or alternative distributions of services (Knapp, 1997). Many criteria are used in economic analysis. The most common is efficiency (achieving maximum outcomes from minimum inputs), which is the focus here. The present paper's focus on economic efficiency, however, is not meant to imply that correctional treatment programs should only be continued if benefits outweigh costs. There are many important non-economic criteria on which correctional intervention programs should be judged.
Of the two main techniques of economic analysis—cost-benefit and cost-effectiveness analysis—only cost-benefit analysis allows for an assessment of both costs and benefits. A cost-effectiveness analysis can be referred to as an incomplete cost-benefit analysis, because no attempt is made to estimate the monetary value of program effects produced (benefits or disbenefits), only resources used (costs). For example, a cost-effectiveness analysis can specify how many crimes were prevented per $1,000 spent on a program. Another way to think about how cost-benefit and cost-effectiveness analysis differ is that "cost-effectiveness analysis may help one decide among competing program models, but it cannot show that the total effect was worth the cost of the program" (Weinrott et al., 1982, p. 179), unlike cost-benefit analysis.
A cost-benefit analysis is a step-by-step process that follows a standard set of procedures. The six main steps are: 1. Define the scope of the analysis; 2. Obtain estimates of program effects; 3. Estimate the monetary value of costs and benefits; 4. Calculate present value and assess profitability; 5. Describe the distribution of costs and benefits (an assessment of who gains and who loses, e.g., program participant, government/taxpayer, crime victim); and 6. Conduct sensitivity analyses (Barnett, 1993, pp. 143รถ148).
It is beyond the scope of this paper to discuss each methodological step, but interested readers should consult the excellent reviews of this methodology in the context of offender rehabilitation programs by Weimer and Friedman (1979) and substance abuse prevention programs by Plotnick (1994). For methodological features of cost-benefit analysis in general, see the text by Layard and Glaister (1994) and the volume by Welsh et al. (2001).
Two other key features of economic analysis require brief mention. First, an economic analysis is an extension of an outcome or impact evaluation, and is only as defensible as the evaluation upon which it is based. Weimer and Friedman (1979, p. 264) recommended that economic analyses be limited to programs that have been evaluated with an "experimental or strong quasi-experimental design." The most convincing method of evaluating correctional treatment and crime prevention programs in general is the randomized experiment (Farrington, 1983, 1997). Second, many perspectives can be taken in measuring program costs and benefits. Some cost-benefit analyses adopt a society-wide perspective that includes the major parties that can receive benefits or incur costs, such as government or taxpayer, crime victim, and program participant. Other analyses may take a narrower view, focusing on one or two of these parties. In reporting on the cost-benefit findings of the studies reviewed here, the middle-of-the-road approach has been taken, by reporting on, as far as possible, a combined government/taxpayer and crime victim perspective.
Results from Cost-Benefit Analysis Studies
Fourteen studies were identified that evaluated the impact of correctional treatment on reoffending in the community and carried out a cost-benefit analysis. Studies that did not perform a cost-benefit analysis were included if they presented sufficient cost and benefit data to enable an assessment of economic efficiency. Table 1 summarizes key features of the 14 correctional treatment studies and reports on program effects on reoffending in the community and cost-benefit findings. Twelve of the studies were carried out in the U.S. and the other two in England. All but one of the studies (Pearson, 1988) carried out its own cost-benefit analysis. This study instead performed a cost analysis, but published data that allowed for the calculation of financial benefits and hence a benefit-cost ratio. The studies are listed in chronological order.
All of the studies with the exception of one (Farrington et al., 2002, Colchester site) yielded a favorable benefit-cost ratio, meaning that program benefits outweighed program costs. It is important to note that the study by Gray and Olson (1989) calculated benefit-cost ratios for each of the three treatments being compared (probation, prison, and jail), but for the purposes of the present discussion, only the analysis of probation will be considered, because probation more closely fits our concern with correctional treatment than does prison or jail. For the 13 studies with favorable benefit-cost ratios, ratios ranged from a low of 1.13:1 to a high of 270:1, meaning that for each monetary unit (one U.S. dollar or one British pound) spent on the programs, society received in return 1.13 units to 270 units in various savings.
Of the 13 studies that carried out their own cost-benefit analyses, the majority were considered high quality, following to some degree the above-noted steps of the recommended cost-benefit analysis methodology. McDougall et al.'s (2003) systematic review of the costs and benefits of sentencing, which included seven of the 14 studies reviewed here, reached the same conclusion. The authors developed an innovative cost-benefit validity scale that ranks the comprehensiveness of cost-benefit analyses from lowest (level 1: cost analysis studies in which benefits are not monetized) to highest (level 5: complete cost-benefit analysis). The cost-benefit validity scale ranking of these seven studies averaged 3.6.
As shown in Table 1, half of the studies (Holahan, 1974; Friedman, 1977; Mallar & Thornton, 1978; Pearson, 1988; Roberts & Camasso, 1991, both sites; Gerstein et al., 1994) assessed and quantified in monetary terms outcomes other than recidivism. Education, employment, health, social service use, and illicit substance use were the different kinds of outcomes monetized in these studies. In five of the studies (Friedman, 1977; Mallar & Thornton, 1978; Pearson, 1988; Roberts & Camasso, 1991, both sites), benefits from improvements in these outcomes exceeded benefits from reduced recidivism.
Four of the studies (Pearson, 1988; Gray & Olson, 1989; Courtright et al., 1997; Robertson et al., 2001) limited their measurement of crime-related benefits to the criminal justice system, whereas the remaining ten assessed both the criminal justice system and crime victim expenses. Financial costs to victims of crime can be divided into two main categories: direct or out-of-pocket (e.g., lost wages, medical expenses) and indirect (e.g., pain, suffering, lost quality of life, fear of future victimization), which may also include the risk of death. Among the ten studies that measured crime victim costs, these costs were mostly limited to direct expenses; only two of these studies also measured indirect crime victim costs (Farrington et al., 2002, both sites). To be fair to the authors, the majority acknowledged the difficulties involved in assessing and quantifying in monetary terms intangible costs to crime victims. These difficulties include the lack of existing estimates of the intangible costs to victims of crime, which first appeared in the published literature in Cohen (1988; for more recent reports, see Cohen, 1998; Cohen et al., 2004), and the doubts of many researchers about the validity of these costs and the underlying theory used in their calculation (Zimring & Hawkins, 1995, p. 138).
The importance of assessing and quantifying intangible costs to crime victims in costbenefit analyses was illustrated in Cohen's (1988) reanalysis of Austin's (1986; see Table 1) cost-benefit calculations. For example, Cohen estimated that the average rape cost $51,058 (in 1985 dollars), made up of three main components: direct losses, $4,617; pain, suffering, and fear of injury, $43,561; and risk of death, $2,880. Adding the pain and suffering cost component to Austin's (1986) estimates of the direct losses incurred by crime victims, while maintaining the other costs, increased the total costs of the program to approximately $110 million (Cohen, 1988, p. 550), a six-fold increase. This resulted in a reversal of the cost-benefit findings, from producing a dividend on public expenditure (a benefit-cost ratio of 2.82:1) to a loss or an undesirable benefit-cost ratio of 0.45:1 ($49 million divided by $110 million).
This paper has purposely avoided concluding that one intervention was the most economically efficient. This was because of the small number of available studies, the varied methodological rigor of the evaluation designs (four studies employed simple before-after designs with no control group), and the varied sophistication and comprehensiveness of the cost-benefit analyses. Also, despite the wide range of interventions used by the 14 studies (e.g., intensive supervision with monitoring, pretrial diversion with counseling, supported employment), the coverage of correctional intervention modalities is far from complete (see Palmer, 1994; Lipsey & Wilson, 1998; MacKenzie, 2002).
Furthermore, two of the studies reviewed here (Austin, 1986; Gerstein et al., 1994) did not evaluate correctional treatment per se. Austin (1986) evaluated the decision to release offenders from prison prior to the expiration of the prison sentence. This may not be a correctional intervention, but it does represent an alternative to incarceration that has received some attention then and of late (Butterfield, 2002). In the study by Gerstein et al. (1994) that evaluated alcohol and drug abuse prevention services throughout California, not all of the participating subjects were under the authority of the Department of Corrections at the time of treatment. These two studies were included partly because of the paucity of cost-benefit research on correctional treatment.
Implications for Offender Reentry
If there is a cost-benefit argument to be made for correctional treatment in some (but not all) of its various modalities, as the above evidence seems to support, then the first question that needs to be asked is: Is this sufficient grounds for spending more on correctional treatment? Some will answer in the affirmative. Others will hasten to add that there are other considerations that do not neatly fit in the cost-benefit ledger. Indeed, while cost-benefit arguments may be persuasive, other matters may come to dominate, such as other government priorities of the day, other concerns of the public (as revealed in national polls), and institutional and political resistance to change. On the latter, supporters of punitive sentencing regimes and further limiting treatment resources for prisoners can marshal some rather strong evidence that shows that the three decades-long prison build-up has had a sizeable effect on national crime rates in recent years (Spelman, 2000a, 2000b; Levitt, 2004; Lynch & Sabol, 2004).
These other, non-economic considerations are by no means peculiar to the criminal justice system; they affect many other areas of government interest when it comes to decisions of whether or not to increase public expenditure. What a pro-prison position truly lacks, however, is evidence of robust costbenefit accounting that shows that prison provides value for money. (For one example of this, see Piehl and Dilulio, 1995.)
The next question that needs to be asked is: What are some of the cost-benefit arguments, aside from benefits exceeding costs, that can be made in favor of an increase in correctional treatment resources that may benefit offender reentry? Perhaps one of the most important cost-benefit arguments is that benefits from correctional treatment are not limited to a reduction in recidivism. As noted above, in some of the studies benefits from improvements in education, employment, health, social service use, and illicit substance use exceeded benefits from reduced recidivism. Although it is far from conclusive, this is an important finding because it suggests that correctional treatment programs have the potential to influence other important areas of an offender's life and produce, in some cases, substantial economic returns for publicly funded services such as health and welfare.
Another important cost-benefit argument that can be made in support of increasing treatment resources is that the benefits are realized in a relatively short period of time, typically within two years post-treatment and in some cases at the completion of treatment. This may be especially important for offenders returning to communities. The longer it takes them to find a job or housing or to get their life together in general, the greater risk for reoffending.
The short-term time frame in which correctional treatment benefits are realized may also have political significance. Obtaining funding for programs that have the potential to produce immediate benefits is far more appealing because of the short time horizons of politicians (Tonry & Farrington, 1995).
Conclusion
Offender reentry programs are crucial in an effort to reduce recidivism rates. But it may be that what comes before this end stage—in the form of correctional treatment programs —is equally, if not more, important. From the cost-benefit studies reviewed here, it would seem that a case can be made for increasing treatment resources for offenders, and this may improve offenders' chances for a successful return to the community.
Of course, the present concerns with offenders returning to the community are not limited to diminishing correctional treatment resources. So while this paper has focused solely on treatment resources, one could ask if it would also be worthwhile to provide more parole opportunities to offenders (thereby providing more incentives to participate in treatment), and more transitional aid, or some other means of assistance. One of the studies reviewed here (Mallar & Thornton, 1978) found that financial and job placement assistance for released offenders reduced theft crimes, increased employment, and reduced reliance on social services such as welfare. These improvements translated into substantial monetary benefits for society. It may turn out that a program of correctional treatment followed by transitional aid will produce multiplicative rather than additive benefits.