Anyone interested in the topic should take a look at _When Brute Force Fails_, by Mark Kleiman[1].

Briefly, a drug policy guy makes the case that "swift and certain" punishment with pretty usually not very severe outcomes does a lot better as deterrence than what we do now.

There's a lot to digest, and I don't think anyone is going to agree with all of it - he gores a lot of oxen. But it is meticulously researched and reasoned, and hard for this non-policy professional to argue with. Really worth at least borrowing at the library for folks who like grappling with reality at a policy level.

Mark was also a prolific blogger who died recently. I didn't know him, but did read his blog.

[1] https://www.amazon.com/When-Brute-Force-Fails-Punishment/dp/...

I'm pretty far removed from this world of crime and only relay what I read coupled with my biased experiences but I suspect crimes build on top of one another. If you get away with a small crime earlier you'll end up breaking the seal and moving on to higher risk crimes. If you keep on getting away with it you end up taking on more and more risk. At the same time there are not enough resources to enforce a low level crime so fixing the underlying problems is a much better idea.

If you're interested in this topic take a look at When Brute Force Fails (http://www.amazon.com/When-Brute-Force-Fails-Punishment/dp/0...) - I found it a good take on crime in the US.