Law and Order: An Economics (sunk cost) Perspective

A raging debate is going on over at Colin Espiner’s blog on National’s new law and order policy. Given that people are making lots of arguments on both sides, I thought it might be worthwhile laying out an analytical framework for how an Economist might view the justice system. I’m not going to go into whether or not I think National’s policy is good or bad, that’s for you to decide.

Before we go any further we need to understand the concept of a sunk cost, as this will be crucial to my discussion. I’ll let you read the wikipedia definition, but put simply a sunk cost is one which cannot be recovered after it has been incurred. Therefore economic theory states that ex ante you should take into account the sunk portion of the cost of an action but that ex post the sunk portion should be ignored.

I’m going to use the term cost quite loosely here, it can refer to someone being impaired financially as the result of a robbery or suffering emotional harm as the result of a crime. This is not a discussion about the financial cost of running a prison system or anything like that (although I acknowledge that is an important issue).

Now, Let’s talk Justice

My friends who have studied law tell me that there are three purposes of the justice system

  1. Deterrence
  2. Punishment
  3. Rehabilitation

I would argue that each of these factors can be explained using economics and that there is trade-off in achieving all three. This trade off is why there is such heated debate on this issue. Now I’m going to wave my hands and convert these three factors into economics

  1. The cost (punishment) imposed on the offender must be sufficiently large to exceed the personal benefit he/she obtains from committing the crime so that ex ante we can avoid the cost the crime imposes on society
  2. The cost that the crime imposes on society can be partially reversed by punishing the offender and thus he should be punished (in technical terms the cost imposed on society is not entirely sunk)
  3. If the offender can contribute to society going forward and this benefit exceeds the non-sunk cost imposed on society as a result of the crime (as well as the cost of rehabilitation) then the offender should be rehabilitated

The Trade-offs

With that loose framework in place let’s discuss the trade-offs. I think we would all agree ex ante that we would like crimes not to and thus avoid the cost being imposed on society. What happens ex post when a crime is committed is generally where most people disagree.

With the framework I’ve laid out, ex post if the majority of the cost can be reversed through punishment then the offender should be punished to an extent that reverses the non-sunk component of the crime. Then you would think about rehabilitation. On the other hand if the majority of the cost is sunk, punishment achieves nothing and you should just rehabilitate the offender if he/she can contribute to society

Don’t call me a sociopath yet – I’m getting to the trade off

It has happened before on this blog when other commentators have used economics to analyze delicate social issues, so let me finish before you get too angry. An astute reader may have noticed, crimes such as murder are likely to be mostly sunk (you can’t bring someone back to life) yet I appear to be saying this is a situation where you should rehabilitate. This is where the trade off between rehabilitation and deterrence becomes particularly important. Arguably a crime such as murder has an almost infinite cost on those affected by it and thus we would really like to prevent it. However, if an offender knows that he will be rehabilitated ex post, then ex ante he knows the cost he will face for the committing the crime will be relatively low.

This is effectively what economists would call a time inconstancy problem (from societies perspective, not the criminals). Criminals know that ex post society is not able to commit to a harsh punishment and is likely to be less severe, so ex ante they are more likely to commit crimes. A strong rehabilitative justice system mitigates the deterrent effect of the justice system.

Thus there is a trade-off. If society is too lenient ex post, more crimes will be committed. On the other hand if you are less lenient, fewer crimes are committed but then those offenders who could contribute to society ex post don’t get the chance too.

Basically, you can’t have the perfect effect of all three forms of justice, if you want more of one, you genrally need to have less of another, thus the optimal justice system is likely to have components of all three.

Life is about tradeoffs