Mainstreaming Methodology for Estimating Costs of Crime
Much of the literature in the criminal justice field is concerned with the measurement of effectiveness. This kind of analysis does not require cost of crime data. It is covered here mostly to help the site user to see how cost effectiveness analysis, cost benefit analysis and other methods are related to it.
Imagine that a police force is considering whether to introduce DNA testing as a routine part of burglary investigations. The presumption is that such testing increases the probability of identifying offenders. From a pure ‘effectiveness’ standpoint, the objective is simply to establish whether this presumption is supported empirically. So if it can be established that a higher proportion of crime scenes analysed using DNA technology result in arrests than when DNA is not gathered, then the case is made.
Suppose that a well-designed experiment (controlling for extraneous, complicating factors) demonstrates that for every 100 crime scenes investigated using conventional methods an arrest results in 40 of the cases, whilst the corresponding number of arrests when DNA testing is added is 50. The intervention is judged to ‘work’ since it increases significantly the proportion of investigations resulting in an arrest.
The design of experiments to establish effectiveness is a field of study itself. There has been substantial discussion, for example, of the role of Randomised Controlled Trials as a means of exploring the effectiveness of interventions or treatments. We do not enter that discussion here, except to note that many statisticians are very cautious about drawing inferences from experiments that do not incorporate a substantial degree of randomisation.
