Although private prisons have been sold on economic grounds, a study this year by Arizona's own Corrections Department questions whether such facilities can even deliver in terms of cost savings, reports the Arizona Republic. The state's cost study showed that it's often more expensive to incarcerate inmates in private prisons than in state-run facilities, despite the savings that private operators typically promise.
Will Wilkinson, writing at The Economist, attributes this to the perverse economic incentives that obtain for contractors.
From an economic point of view, we should expect firms that compete for and rely on government contracts, such as weapons manufacturers and prison operators, to maximise the spread between the amount billed and the actual cost of delivering the service. If contractors can get away with providing less value for money than would the government-run alternative, they will.\n
The fact that contractors try to maximize the spread between what they bill and the cost of delivering the service certainly doesn't make private prisons any better. But there's a deeper problem. We don't know how to define the service.
With a weapons manufacturer it's easy. Did you get the weapons you paid for? Did they work? Yes? Good, then even though you may have been overcharged, the service was delivered.
But what is the service that prisons are supposed to deliver? There isn't much agreement on this question. Most people probably have a vague mix of ideas swimming in their head about what prisons should deliver. Prisons should sequester criminals to protect the public; prisons should provide a deterrent to potential offenders; prisons should rehabilitate; prisons should punish criminals by giving them an unpleasant experience that they "deserve."
How the hell do we know if prisons are delivering with a mandate like that? The aims of prison, as understood by the public and articulated by politicians, are often contradictory, or at least apparently so. Do therapeutic rehabilitation programs compromise the deterrent effect of prison, or make the punitive element too weak? Do punitive policies make it hard to rehabilitate?
Even if we removed the retributive function and decided, as a society, that prison's sole metric of success was reducing crime by maximizing deterrence and minimizing recidivism within reasonable cost restrictions, it would still be nigh impossible to really measure how well a certain facility was delivering. How do you measure recidivism? Currently, we usually look at the percent of prisoners that reoffend within three years of release. That means we don't know whether a given policy is working for three years. If you start trying to get more accurate recidivism numbers by looking at reoffending within, say, 10 years, it's even longer before you get feedback. How do you design data-driven policy with time lags like that? And that's to say nothing of the challenges you encounter when you try to control for other differences among facilities, their entering populations, and the situations inmates are released back into.
Because there's so little clarity about what we want prisons to do, we give up on any serious expectation that prisons will provide anything more than a simple custodial function: keeping prisoners sequestered from society in minimally humane conditions for the length of their sentence. And, in fact, as Khimm points out, Arizona is only now taking a second look at private prisons because they failed at this basic custodial job—three inmates escaped from a private facility last month.
This total lack of clarity about the service prisons are providing, combined with the twisted economic incentives of guards' unions and the opportunistic fearmongering of politicians, has created a system of punishment that's totally divorced from the public interest. It's a problem for public and private prisons alike.