Ross Gittins has an interesting piece on the Western Australian decision to abolish all jail sentences shorter than six months. (Link via Professor Bunyip who includes an interesting personal anecdote, which gives clues to the continuing puzzle of his true identity, and pretty conclusive refutes my own hypothesis on this topic).
I endorse this decision. I hope it works satisfactorily and I think it should be taken further. As I pointed out here, once you accept that prison sentences don’t have much effect in either deterring crime or rehabilitating criminals, there’s no point in short sentences. On the one hand, there’s a strong case for locking hardened criminals up until they’re too old for crime. On the other hand, given that most juvenile institutions appear to be training grounds for adult criminals, there’s a strong case for second chances and slaps on the wrist for young offenders, based on the observation that a lot of them grow out of it naturally.

Let me throw another idea into the mix. One reason we resort to imprisonment so readily is that there’s no alternative punishment that’s really serious. Fines typically max out at a couple of thousand dollars, which is trivial for an employed criminal. I’d suggest making fines proportional to income and collecting them over a period of years using the same mechanisms that are used for child support. We could then impose much heavier fines as an alternative to short stretches of prison or pointless ‘community service’.

Update As always, there’s lots of good stuff in the comments thread. In response to Ken Parish, I want to clarify one point. My argument doesn’t rely on the assumption that criminal penalties have no deterrent effect, only that there is no significant deterrent effect from marginal increases in the severity of penalties e.g. short prison terms vs fines. I agree that greater probability of detection does have a deterrent effect.