Costs and Benefits of Being #1 at Incarceration

Scott at Simple Justice is always a source of fascinating issues, and recently he wrote about the Pew report on incarceration:

The Pew report has been the subject of intense interest around the blawgosphere.  This is the report that tells us that we finally reached a 1 in 100 incarceration rate (and far higher for minorities) and the cost is blowing the doors off of budgets…

That’s a higher rate of incarceration than any other democracy in the world. It’s also a higher rate of incarceration than any totalitarian dictatorship in the world. And we put young black men in jail at a higher rate than South Africa at the height of apartheid.

(Depending what you’re trying to prove, it can be unfair to compare U.S. incarceration rates to those of unfree countries. In places like North Korea and Myanmar, the incarceration rate is arguably 100% since the entire country functions as a work camp. Put another way, would you prefer a 99% chance of living outside of prison in the U.S., or a 99.25% chance of living outside of prison in Cuba?)

Scott’s post is in response to a New York Times piece by Adam Liptak:

In juxtaposition to the report, Liptak brings in Paul Cassell, Utah lawprof, former federal judge and victim’s rights advocate (though Liptak doesn’t mention this last piece).

“We aren’t really getting the return in public safety from this level of incarceration,” said Susan Urahn, the center’s managing director. 

But Paul Cassell, a law professor at the University of Utah and a former federal judge, said the Pew report considered only half of the cost-benefit equation and overlooked the “very tangible benefits: lower crime rates.” 

Scott goes on to argue that we don’t really have good proof of the relationship between incarceration rates and crime rates. I’m sure he’s right that the relationship is weak, but I’m also pretty sure that some relationship exists.

What Cassell points out is that the Pew report only discusses the costs of operating prisons. The numbers may well be accurate, but they only tell part of the story. Cassell wants to discuss the benefits of high incarceration rates as well. In doing so, he’s trying to change the argument in two important ways.

First, he’s changing the subject of the argument from state budgets to social welfare. This makes a lot of sense. Governments supposedly exist to serve the people, so when we want to analyze a policy, the proper subject of any cost-benefit analysis is the effect on the people as a whole.

Second, Cassell wants to count non-financial costs and benefits. To do that, he would have to find a way to compare the various non-financial costs and benefits with the financial costs and benefits. In effect, he wants to assign a dollar cost for every crime committed, and then see how much is saved by reducing the number of crimes.

This sounds impossible at first, but certain kinds of economists spend their whole careers figuring out ways to do things like this. For example, you could do a factor analysis of home sales all over the country to try to tease out the effects of local crime rates on local real estate prices. This would give you an idea of how much people are willing to spend to reduce their chances of being crime victims. A similar analysis could be done on wages and on-the-job crime to figure out how much of a pay cut people are willing to take to work at a safer job.

So the question that Cassell wants us to answer is whether the $50 billion dollar annual cost of operating our nation’s prisons is more or less than the benefits we reap by reducing crime.

Scott Greenfield is essentially arguing that when analyzing our policy towards imprisonment we should only count the reduction in crime that is attributable to increased imprisonment. He’s right, but he’s missing a couple of other big gaps in Cassell’s argument.

For one thing, many victimless crimes have no net social costs. Since all the participants have consented to the “crime.” it’s a pretty safe assumption that all participants are receiving benefits that are worth the costs. So we as a society are enduring the costs of imprisoning people for victimless crimes without reaping any compensating benefits.

It gets even worse for Cassell. He argues that the Pew report is omitting the benefits to society of incarceration, but he himself is making the exact opposite mistake by omitting a major cost. It may even be the biggest cost. Cassell is omitting the cost of imprisonment to the prisoners themselves.

It may seen counter-intuitive to count the cost to prisoners, but if you’re going to make a cost-benefit argument for a social policy, you have to include all the costs and benefits to all the people involved. History is filled with horrible ideas that look good if we treat some people as less important than others. Everyone counts, and everyone counts equally.

It may seem hard to assign a cost to imprisonment, but it’s no harder than assigning a cost to being a crime victim, and Cassell is insisting we do that.

The annual cost of prison to a prisoner is the same as the amount of money he would want to be paid in order to voluntarily spend a year in prison. That’s a tricky number to estimate, but here’s one quick calculation: The average wage for high school dropouts in 2007 was $11.40 per hour, which means that it would cost $23,712 per year to hire one for 40 hours a week for a full year. Surely he would demand no less to spend the same year in prison. Multiply by the 2.3 million prisoners in the Pew report and you have a lower limit of $55 billion dollars.

That brings the total cost to $105 billion dollars per year. This seems intuitively in the ballpark: Imprisoning 1% of the people who produce a $10 trillion economy costs about 1% of $10 trillion.

(If you think I’m arguing that our policy for imprisoning rapists and murderers should take into account how unpleasant it is for them, you’re exactly right. Of course, we should also take into account the benefits to society of imprisoning rapists and murderers. I have no doubt that imprisoning rapists and murderers has more benefits than costs.)

Keep in mind, the real question is not whether we’re better off paying the $100 billion dollar cost of incarceration. Of course we are. Without our system of law and justice, we’d descend into violent anarchy—think Somalia around 1990.

No, the real question is what our policy should be on the margins: Which way should we be moving? If we make a policy change that increases our prison population by 20,000 at an annual social cost of a billion dollars, will that lead to a reduction in crime that saves us a billion dollars every year? Or if we decide to follow a policy that reduces our prison population by 50,000, will the resulting increase in crime wipe out the $2.3 billion we save?

The answers to these questions obviously depend on the specifics of what types of crimes and criminals we’re talking about. Releasing thousands of murderers is probably a bad idea. On the other hand, releasing the perpetrators of victimless crimes is an obvious win.

Similarly, we should contemplate the costs and benefits of creating new crimes or increasing the punishments for existing ones. Prison time is expensive, so we should try to avoid buying too much of it.