Chapter 5

COST-EFFECTIVENESS

Violence costs the United States an estimated $425 billion in direct and indirect costs each year (Illinois Center for Violence Prevention, 1998). Of these costs, approximately $90 billion is spent on the criminal justice system, $65 billion on security, $5 billion on the treatment of victims, and $170 billion on lost productivity and quality of life. The annual costs to victims are approximately $178 billion (Illinois Center for Violence Prevention, 1998). The most logical way to reduce these costs is to prevent violence altogether. Preventing a single violent crime not only averts the costs of incarceration, it also prevents the short- and long-term costs to victims, including material losses and the costs associated with physical and psychological trauma.

Despite these facts, policy in the United States continues to focus on get-tough laws and incarceration for serious violent criminals, as opposed to prevention and intervention (Greenwood, 1995). Federal spending on school-based crime, violence, and drug prevention programs is quite modest, compared to spending on crime and drug control strategies such as policing and prison construction (Gottfredson et al., in press). Not only are preventive approaches more beneficial than get-tough laws, some prevention and intervention strategies cost less over the long run than mandatory sentences and other get-tough approaches.

In an effort to determine the cost-effectiveness of California’s three-strikes-and-you’re-out law, which mandates life sentences for repeat offenders, Greenwood (1995) compared that approach to the benefits and cost-effectiveness of a number of crime prevention strategies. He estimated that each serious crime—homicide, rape, robbery, assault, or residential burglary—prevented by the three-strikes law cost the criminal justice system in California an additional $16,000 over the amount spent prior to this legislation. Using this price as the standard for cost-effectiveness, Greenwood calculated the costs per serious crime prevented of four prevention and intervention strategies: (1) early childhood intervention (perinatal home visitation continuing through the first 2 years, combined with 4 years of enriched day care programs) for high-risk families, (2) parent training for families with children who have shown aggressive behavior ("acted out") in school, (3) improved public school programs that target all youth, and (4) early interventions for very young delinquents. The costs calculated for each of these interventions included only direct program costs, not such indirect benefits as the money saved by averting incarceration or preventing victim trauma and its medical and social consequences.

Table 5–2 shows the benefits of the various prevention and intervention programs with respect to the number of serious crimes each can be expected to prevent over the course of 30 years. The major disadvantage of the prevention approach is clear—there is a time lag between implementation of programs and the appearance of effects. Because of this time lag, programs that are cost-effective in the long run do not appear so in the short run. In addition, long periods between an intervention and the high-risk period of a youth’s life offer more opportunity for decay of a program’s effects (Greenwood et al., 1998). In the case of early childhood programs, it takes approximately 15 years before significant effects on youth violence can be appreciated, given the peak ages at which young people are involved in violence. Early intervention with delinquent youths that includes day treatment and home monitoring has a shorter lag time because the intervention is introduced later in life yet early in a violent career.

Table 5-2. Cost-effectiveness of early intervention in California

Of the four approaches listed in Table 5–2, the most cost-effective in the long run is parent training, which costs only $392 to implement per serious crime averted after the program has been in effect 30 years. This is less than one-fortieth the estimated cost of preventing serious crime under the three-strikes law. Day treatment and monitoring for delinquent youths are also more cost-effective than mandatory sentencing, costing less than one-sixth as much as the three-strikes approach.

The least cost-effective of the four are prenatal and early childhood intervention and school-based programs that target all students. However, early childhood interventions that include prenatal home visitation and enhanced day care can be expected to halve the incidence of child abuse among high-risk families (that is, low-income families headed by a single mother).⁵ Moreover, early childhood intervention may improve educational achievement and reduce teen pregnancy rates. School-based programs have benefits other than prevention of violent crime, including higher educational achievement for all students. In a later analysis, Greenwood et al. (1998) found that school-based prevention programs that targeted disadvantaged youths specifically and included incentives (such as cash) for graduating from high school were almost 10 times as cost-effective as the three-strikes approach.

In general, Greenwood’s findings suggest that interventions targeting problem youths—either children who act out or delinquent youths—are more cost-effective than interventions that target general populations of youths. In addition, they confirm that prevention is truly more cost-effective in the long run than incarceration.

Costs aside, prevention may not have as great an effect on rates of violence as imposing longer mandatory sentences on repeat offenders. Other analyses demonstrate that the three-strikes law can reduce serious crime by 21 percent, whereas graduation incentives only reduce it by approximately 15 percent, parent training by 7 percent, early childhood intervention by 5 percent, and delinquent supervision by less than 2 percent (Greenwood et al., 1998). However, the four prevention and intervention strategies combined cost nearly $1.2 billion per year less to implement than the three-strikes strategy alone, and together they could prevent a substantial portion of the 80 percent of serious crimes that are not averted by mandatory sentencing (Greenwood et al., 1998). Graduation incentive programs could pay for themselves with the money they save by averting the eventual incarceration of many youths, and the other prevention and intervention strategies could pay for up to 40 percent of their costs in the same manner.

Studies of two targeted early childhood intervention programs, the Perry Preschool and the Elmira, New York, Prenatal and Infancy Home Visitation by Nurses, indicate that these programs can actually save the government up to three times their cost when delinquency prevention and other benefits are considered (Karoly et al., 1998). It is noteworthy that although the cost-effectiveness data in Table 5–2 were calculated using crime and population statistics for California, they have national implications with respect to the relative costs and benefits of violence prevention and incarceration.

Researchers at the Washington State Institute for Public Policy, who conducted a similar analysis (Aos et al., 1999), point out that the most effective programs are not always the most cost-effective. They note the importance of matching the intervention to the population—a particular challenge for programmers, but one that has a critical effect on both the overall effectiveness and the cost-effectiveness of an intervention.

The results of the Washington study are summarized in Table 5–3. While this table includes only the programs and approaches discussed in this report, the Washington study actually included many more programs and strategies, including some targeting adult offenders. All cost estimates in Table Table 5–3 were calculated using the same methodology so that programs can be compared. Although most costs are calculated as direct, per-participant program costs, the costs of Multidimensional Treatment Foster Care are calculated relative to regular group home costs, and the costs of intensive supervision programs and boot camps are calculated relative to regular court probation costs. (Thus, the negative program cost of boot camps means that these programs cost less to implement than regular court probation programs.) This overall approach may not be the same one used by other researchers to calculate program costs, resulting in inconsistencies between costs in this table and those projected by individual program designers (Box 5–3).

Nevertheless, the Washington study offers some useful insights into the cost-effectiveness of youth violence prevention. Looking at the benefits to the criminal justice system alone (that is, benefits to the taxpayer), many early interventions and selected strategies come close to paying for themselves with the money they save; others actually achieve benefits that are greater than program costs. The Seattle Social Development Project, for instance, now saves $0.90 from reduced rates of crime for every tax dollar spent. Programs targeting at-risk or delinquent youths can be even more cost-effective. For example, taxpayers today can expect to save $14.07 in future criminal justice costs for every dollar spent on Multidimensional Treatment Foster Care.

The same trend holds when considering the benefits of youth crime prevention to both the criminal justice system and crime victims (personal and property losses)—the largest economic returns are achieved with interventions targeted at juvenile offenders, who are at greatest risk of future offending. The Model programs in this group return $11 to $22 for every dollar invested. However, even programs aimed at nonoffenders can achieve significant cost benefits when future savings to potential crime victims (due to a reduction in the number of victims) and the taxpayer are combined. According to the Washington study, society gains at least $0.50 over program costs for each dollar spent on the Perry Preschool Program, Prenatal and Infancy Home Visitation by Nurses, the Seattle Social Development Project, and Big Brothers Big Sisters of America.

In general, these analyses underestimate the benefits of prevention programs because they fail to consider many of the indirect benefits of preventing serious or violent offenses, such as increased work productivity, increased taxes realized, reduced welfare assistance costs, and reduced victim medical costs.

Table 5-3. Comparative costs and benefits of prevention and intervention