Efficient crime-fighting policies
Clearly, it does not look optimal to spend all available resources on crime control or to spend no resource on it. But then, how do we find the optimal allocation of resources for crime control and other policy objectives?
An economist would say the optimal allocation is where an additional dollar spent on crime control generates the same amount of social gains as an additional dollar spent on other policy objectives. (If not equal, we can always achieve greater social gains by reallocating some resource from one policy objective to another.) Spending most of the available resource on crime control would not be an optimal allocation, because we can probably generate more social gains by spending one less dollar on crime control (little social loss) and spending one more dollar on other policy issues (large social gain).
This idea of efficiency can be applied on various policy measures on crime control as well. It is likely that most of the traditional crime control measures (such as adopting more advanced crime-investigation technologies, hiring more police officers, and offering treatment and rehabilitation programs to former offenders) all “work” to some extent. Hiring more police officers will probably cause crime to go down, although it is less clear by how much crime will go down.
However, we still have a decision to make here, because we may be able to achieve greater crime control at the same cost by reallocating some resource from one policy measure with little marginal gains (i.e., low crime-reduction) to another policy measure with larger marginal gains (i.e., larger crime-reduction). After all, if we can reduce more crimes while spending the same amount of resource, why not do this?
It is very difficult, however, to actually identify and compare social gains from different crime-control policy measures.
First, the effectiveness of each measure is likely to be related to the baseline level of the resource allocation. For example, if most of criminal justice resources are currently spent on maintaining a very large police force, additional spending on new police hires is unlikely to yield a high crime-reduction. But in another country with a small number of police officers, additional police hires can have a much larger crime-reducing effect.
Second, crime is very much multi-dimensional. Suppose a 24-hour GPS monitoring system on high-risk sex offenders reduces their recidivism risks by 10 percent, and a successful drug-rehabilitation program reduces drug offenders’ recidivism risks by 20 percent. It is not clear how we should compare social gains from a 10 percent decline in repeat sex offenses with a 20 percent decline in repeat drug crime.
Lastly, the crime-reducing effect of a policy measure likely depends on the resource allocation on other policy measures. For example, suppose a lot of new police officers are hired to catch more criminals but the budget on prison guards and facilities remains the same. Prison over-crowding will soon follow, and government won’t be able to send many of the arrested criminals to prison. The crime-reducing effect of an additional police officer will decline as a result. Such complications make it difficult for us to compare additional social gains from spending more on different crime-control policy measures.
© Songman Kang, Hanyang University