October 12, 2009

When Brute Force Fails try a more focused approach to fight crime

In an article by Robert H. Frank in The New York Times I read about a new book about fighting crime by Mark Kleiman, professor of public policy and the University of California. The book is called When Brute Force Fails: How to Have Less Crime and Less Punishment and its central claim is that authorities should not make punishments more severe but increase the odds that lawbreakers will be apprehended and punished quickly. The idea is not entirely new of course, but Kleinman's arguments are interesting. There is some reason to doubt about the effectiveness of increasing the severity of punishment.

Over the last decades, in the US, sentences have become longer but so have crime rates. A reason may be that most criminals are very rational but act more like "impulsive children, blinded by the temptation of immediate reward and largely untroubled by the possibility of delayed or uncertain punishment". That the odds of getting caught are relatively low is caused by a lack of strategic focus in law enforcement (typically law enforcement resources are spread more or less equally among all potential offenders).

A targeted approach would work better. By strategically prioritizing, authorities target a specific group of offenders (for instance a specific gang) until its members get the lesson and give up. This can be viewed as a focused-zero tolerance approach. Step by step, moving from one strategic target to the next, we can reach a tipping point and fast progress can be made in fighting crime.


  1. The idea to focus on one area or group at a time has it's merits. However, if this is put into practice, how do we keep the criminals who discover they are not part of the special focus from taking advantage of the fact that less focus is on them?

  2. Hi Rodney, thanks. Interesting question. I'm hesitating in two direction for an answer:

    1) you might argue that criminals are already taking advantage from the fact that there is less focus on them (because the current approach is very untargeted as it is),

    2) in this targeted approach it may be unwise to start taking too much advantage of not being targeted because that seems to be a sure way to become targeted. In this approach a keep your head low tactic seems wise (for criminals)

    Anyway, here is another post on the same topich which explains the idea a bit more extensive:
    How can crime rates be brought down? → Priorities in law enforcement

  3. Hi Coert,
    In the South African context I have experience in Community Police Forums. A precinct is devided into sectors. Hot spots are identified. Targeting hot spots, eg. for housebraking, in one sector is effective only in the sense that the criminals move to another sector. It makes sense that quick and effective apprehension of criminals targeted is the answer instead of more severe punishment. This however will only be effective if crime is targeted over a wide front. The challenge here is the cost efectiveness of an adequate police force and the proper management of resources in hand.

  4. Interesting idea Stanus. I never thought of the fact that if you target a specific type of crime in one area criminals will choose to do that type of crime in another area. So this must be done over a wide enough area to make a difference.

    If the crime fighting on untargeted crimes is reduced it may have negative effects as those criminals may take greater advantage than they already are. I could imagine a conversation in which the criminal says "yeah they are focusing on home burglary in xyz district now but it's as safe (or safer) now to do carjackings."
    One way to prevent this is to have the greater focus be with additional resources or by taking police resources off of areas that have shown little benefit.

  5. Hi Stanus and Rodney,
    This effect is sometimes called the 'waterbed effect'. With a waterbed if you press one point of the bed down another must come up. Many people think that the waterbed effect MUST apply and must apply LITERALLY to crime but I say that this is a pessimistic view which creates a self-fulfilling prophecy.

    I think the view is pessimistic because it implies that the level of crime is constant and that people don't have the choice of stopping to commit criminal acts. Literally applying the waterbed theory is fatalistic and fighting crime at all would become a waste of time and money if were to always happen. It would make it rational to leave criminals alone. I think this 'logic' has been actually applied In the Netherlands to some extent.

    I am not claiming criminals may not move at all when hunted but I would like to see evidence from research for the waterbed effect. I am skeptical. I think it is overrated and the belief in it is partly based on pessimism. I think there are examples of situations in which crime has been brought down in which there was not much movement of crime. For instance, in the crime reduction in New York in the 1990s.

    I am not sure I am right about this because this is not my area of expertise. So I'll be easily convinced by empirical evidence.

    Having said this, I do believe your point is right, Stanus, that you should be prepared to focus on a wide front and you should be prepared to be persistent. I am arguing it would be good to fight crime everywhere in such a targeted way.


Enter your email address:

Delivered by FeedBurner