Monday, July 24, 2006

How to really fight crime

When they're thinking straight, microeconomists possess virtually godlike powers. John Donohue, professor of law and economics at Yale, wrote an astounding piece on crime policy for the winter 2005 Milken Review.

Despite its presence in a journal named for a junk bond crook, Donohue's article is easily the best outline of research in the field I've ever seen. His main points, with my summary and commentary:

1. Stop the Building Boom in Prisons

The pro-prison view is a case study in the vacuity of conservative rhetoric. Politicians like to make claims that are intuitively sound but logically lacking: "taking offenders off the streets will reduce crime and help our inner cities."

Yes, increased incarceration undoubtedly reduces crime, but it costs an incredible amount of money. The question shouldn't be whether the policy has any effect, but whether the effect justifies the cost. According to Donohue, it doesn't.

2. Abolish the Death Penalty

Thanks to the legal system, death penalty cases cost far more than conventional ones. There's no consistent evidence of a deterrent value, and the additional legal expense actually outweighs the cost of incarceration. So why, primitive vengeance aside, should we have capital punishment?

3. Expand the Police Force

I completely agree. But one idle thought: ideally, shouldn't a cost/benefit analysis incorporate the dynamic costs incurred by our means of raising revenue? The effects are difficult to pinpoint, but debt saps capital markets and taxes discourage work. It's possible that a few programs may be less worthwhile than they appear.

4. Adopt Sensible Gun Control

Absolutely. Donohue highlights several particular policy steps:

• preventing police from selling confiscated guns.
• instituting one-gun-purchase-per-month laws.
• plugging secondary-market loopholes.
• tracing all guns used in crime.
• producing guns that can be fired only by their owners.
• registering all handguns.

I can't see any realistic downside to any of these measures.

5. Legalize Drugs

This is debatable territory. Arguably, the optimum policy is legalization, with high taxes to maintain a deterrent. But we'd have to see how implementation really played out—I wholeheartedly agree that we should make marijuana a "trial run."

6. Expand Successful Social Programs

Yes! If you'll pardon the descent into political cliché, we need a more entrepreneurial approach to social policy. Implement what works; cut or reform what doesn't.

7. Defend Roe v. Wade

...or so Donohue argues. This is one area where I disagree. First there's simple jurisprudence: I don't think that Roe v. Wade was a well-crafted or tenable decision. But beyond that, I don't see how crime research makes a contribution to the debate: if abortion is murder, then our country is mired in a moral crisis dwarfing the effects of street crime.

8. Reduce Teen Pregnancy

Several pilot programs promise to do just this. This is an issue where the benefits appear to dramatically outweigh the costs of well-targeted policies.

9. Expand the DNA Database

I can't see any terribly alarming civil-liberties consequences of expanded DNA coverage. Donohue completely overlooks, however, what I think to be a salient issue: wouldn't it be possible to frame someone using DNA? If criminals know that "even a single hair at the scene of a crime is likely to lead to their arrest and conviction," what's preventing them from planting someone else's hair?

And that's it. All in all, I'm very fond of this article. Although it makes a few questionable propositions, with occasionally clumsy attempts at quantification (the average social cost at rape is just $90,000?), it's a tremendous compendium of research in the area.

No comments: