In early October, CBKB convened a roundtable of criminal justice experts to discuss ways of improving the choices public safety agencies make about technology. During the meeting, members of the group cataloged some of the factors influencing these decisions, a list that, regrettably, included lobbying, hunches, and the urgent need to spend unused grant money. All agreed that careful and methodical assessments are too often missing from the equation. Fortunately, such assessments do exist, and I’d like to draw attention to a couple of recent studies⎯one on license plate readers and another on electronic monitoring⎯that exemplify a more evidence-based approach.
Cynthia Lum and fellow researchers at George Mason University investigated the impact of license plate readers (LPRs) on crime in two Northern Virginia jurisdictions. According to a 2010 survey by Lum et al., nearly 50 percent of law enforcement agencies either have LPRs or plan to acquire them soon. LPRs are capable of scanning vast numbers of license plates for instantaneous comparison against an agency database of stolen or suspect vehicles, completely automating a process that otherwise requires constant radio communication with dispatchers (or an elephant’s memory). Now that’s a device with tremendous intuitive appeal.
To test the effect of LPRs on crime, the researchers randomly assigned units to known auto-theft hot spots for one month. Some hot spots were policed by LPR-equipped patrol cars, which would sweep the area to scan all stationary and moving vehicles, and then sit in the area for approximately 15 minutes and scan vehicles passing through. Patrol cars without LPRs policed other hot spots. No significant difference in auto-related or overall crime numbers was detected between the hot spots policed by LPR patrols and those with normal patrols, either between hot spots before, during, and after the LPR test month, or between the month of the test and the same month the previous year. Their study was published last year in the Journal of Experimental Criminology. Although Lum and her co-authors acknowledge their study’s limitations (such as a short intervention period and small sample size), their analysis suggests that if law enforcement agencies invest in LPRs for hot-spot policing, there’s a good chance they’ll be disappointed with the results.
Electronic monitoring (EM) is a widely used public safety technology, and a number of quasi-experimental studies have demonstrated its potential to reduce recidivism. EM is less expensive than incarceration, but costlier than traditional parole or probation, which it typically supplements rather than replaces. Furthermore, most of the crimes EM prevents are minor offenses, raising the question of whether the benefits of EM exceed the costs.
Enter John Roman and colleagues, who just published a cost-benefit analysis of a hypothetical expansion of EM in Washington, D.C. The authors ran a computer simulation using estimated ranges for impacts, costs, and benefits. They derived these range estimates from several data sources. First, they used a meta-analysis of the aforementioned EM impact studies to determine a probability distribution of the number of arrests the technology prevented. They then applied this distribution to D.C. statistics on offenses committed by probationers and parolees. EM equipment costs varied between $9 and $20 and over a time range of 14 to 90 days, consistent with D.C.’s Court Services and Offender Supervision Agency records.
To understand how criminal justice costs are affected by a reduction in the number of parolee and probationer arrests, the research team then calculated marginal costs and conditional probabilities for criminal justice processes (for instance, the probability of conviction once an individual is arrested) using years of data from jurisdictions across the United States. Finally, the analysts calculated the averted victimization costs, including pain and suffering, by drawing from a data set of jury awards in lawsuits stemming from criminal acts. Criminal justice system savings and averted victimization costs were combined to yield societal benefits.
The simulation indicated that with 800 probationers participating, EM had about an 80 percent chance of yielding net benefits to society, and a 25 percent chance of yielding net benefits in excess of $8,700 per participant. The benefits largely reflect averted victimization costs. Savings to the criminal justice system were more modest (typically less than $1000 per participant) and less probable (70 percent). The probability of net benefits was also dependent on the number of participants; a very small EM program (for example, one with about 10 participants) could just as easily result in net costs to society.
These two studies underscore another theme that emerged at the roundtable: investing in a public safety technology is not a simple yes-or-no matter. It involves a host of decisions about the scope and manner in which the technology will be used. Agencies are more likely to make prudent investments when decision makers can examine the evidence to see not only how the technology works, but how it doesn’t.
During the past month our blog has focused on public safety technology and the use of cost-benefit analysis (CBA) in making policy and program decisions. Please comment below or send us your questions or ideas via Twitter or Facebook, or by e-mail to email@example.com.