5 Bad Reasons to Scrap Private Prisons

Slashing corporate incarceration makes people feel better. There's scant evidence that it helps inmates.

Better? Worse? The same?

Photographer: Joshua Lott/Bloomberg

If you’re in the business of running a private prison, you’ve had a bad couple of weeks. Deputy Attorney General Sally Yates announced that the U.S. Justice Department would be reducing its reliance on private prisons after a report by its inspector general suggested they weren’t doing such a great job. Now the government is looking at ending the use of private facilities to detain illegal immigrants. Vermont Senator Bernie Sanders called it “an important step in the right direction” and “exactly what I campaigned on as a candidate for president.” Investors weren’t such big fans; shares of prison corporations slid accordingly.

This is undoubtedly a victory for people who have campaigned against privately managed prisons. But is it a victory that the rest of us should celebrate? To answer that question, we need to know whether private prisons are better or worse than public ones — which is to say, we need to know what there is to dislike about private prisons. Here are the main candidates I’m aware of:

  1. “It is morally wrong for corporations to profit off the mass incarceration of millions of people in this country.” (Kamala Harris, Attorney General of California.)
  2. Private prisons cannot be run as well as public ones, because the profit motive will always cause them to cut corners and deprive the inmates.
  3. Whether or not private prisons could theoretically be run as well as public ones, empirically they aren’t.
  4. Private prisons create a powerful lobby that will influence legislators to increase our already staggering incarceration levels.
  5. Private prisons are apt to have more variance. There will be more good ones, and also, more bad ones.

The first idea plays a strong role in our national debate, even though I think that it’s logically incoherent, for reasons that Michael O’Hare has laid out. All sorts of corporations profit off of the incarceration of millions of people in this country: the makers of guns, razor wire, steel bars and armored vehicles, for example.

As O’Hare notes, “What Harris says implies that every potato on the inmates’ plates, and every brick in the building, and all the guards’ shoes, must be made by a government agency (or, I guess, donated by a nonprofit), or right there in the prison. Maybe it would not be morally wrong if all that stuff were just confiscated from farmers and manufacturers to be sure they don’t profit? Does she demand that the prison be built entirely by inmates and civil-servant hardhats?”

On the other hand, despite my basically libertarian beliefs, I was prepared to believe the second and the third arguments. There are some services for which it is hard to make a market, like police forces, and when we find one of those services, we generally have the state provide it instead. There are reasons that it might be hard to draw up a good contract that lines up the incentives of these kinds of providers with the public’s interest in, say, having a well-run prison that doesn’t abuse its inmates.

Probably the most important feature is simply that the primary “customer,” the inmate, lacks the exit rights that are so important to making markets work. If Sears cheats me, I can take my business to another store. Prisoners cannot take their business to another prison. And while contracts can be written to try to enforce quality as well as price metrics, these metrics are cruder than the kind of discipline afforded by a regular market. That’s why I tend to think that the government should buy its pencils on the open market and run its social services with in-house staff rather than outsourced private agencies; one can be bought in a normal market where customers have exit rights, and the other can’t.

But then I actually spoke to some defense attorneys, in Hawaii and the District of Columbia, and it turned out that some inmates actually prefer private prisons — even though in the case of Hawaii, that means a long plane trip to the mainland for family who might want to visit you. Why? Because the mainland private prison wasn’t as overcrowded as the one in Hawaii. They could have televisions in their cells. It had a nice area that allowed “contact visits” — where you can sit down with your family members instead of talking to them across a glass window -- with vending machines where they could buy snacks, and, as one attorney put it, “Have a little picnic.” Here in Washington, a friend who’s a public defender told me, the private prison is right next to the public prison, and there was a period of time where the public one ended visits (which are quite expensive for the prison), and the private one let them go on.

That’s not a slam-dunk case in favor of private prisons, of course, but as someone who favors incarceration reform, it seems to me that we should care what the prisoners prefer. And from what I can gather, it’s not clear that prisoners view private prisons as worse than public ones, though they’re not really overfond of either.

Nor do we have better metrics that might help us make a better decision; data on prison performance and even prison costs is hard to gather, and the best summary of the research out there seems to be equivocal. Private prisons are certainly awful places. But public prisons are also awful places. Prison is awful, no matter who is running it, so that doesn’t really tell us whether governments or corporations will do a better job. Until there’s better data, I’m agnostic on this point.

What about lobbying? I don’t think the evidence merits the conclusion that private prisons are a unique threat. Prison-guard lobbies are powerful, arguably more so than the private companies, and they lobby for all the tough-on-crime laws that you would expect from a group with a financial interest in incarcerating as many people as possible. I’m not in favor of such lobbying, but there’s not much evidence we’ll get less of it simply by ending private prisons. Indeed, it’s possible we’d get more, because monopoly providers have more incentives to invest in lobbying than players in fragmented markets. A monopoly provider reaps all of the financial benefit of its lobbying, rather than seeing some of its efforts help competitors, and therefore each dollar invested in lobbying represents more potential profit.

Argument No. 5 is the most interesting to me, and it was suggested by my friend the public defender. Bureau of Prisons facilities, she said, are pretty much all the same. But in her estimation, you could get very good private prisons, and also very bad ones — and the freedom that enabled the good ones to outshine the public prisons was probably the same thing that allowed the bad ones to sink below those standards.

I am of course glad if private prisons allow some facilities to do a better job than a government prison could. However, there’s a minimum standard of decency that we owe to people we’ve incarcerated, and it’s questionable whether we really want to “buy” better prisons for some inmates at the cost of sticking others in especially bad ones. If this really is the case, then we should probably end private prisons. But we don’t have solid empirical data to make this case.

Overall, given our state of knowledge, I don’t see a strong case for ending the use of private prison facilities — not without better data, and perhaps, not without trying to see if we can write better contracts that will force the lower-performing private prisons to clean up their acts.

Which is why it's interesting that the issue seems to be suddenly taking off. This may be the swiftest and most high-profile shift in prison policy in many years -- and probably the least important. Privatization is not what makes prison awful, and ending it will not do much to fix our current problems, because as Keith Humphreys has written, “Private prisons are bit players in the sorry drama of mass incarceration.” The overwhelming majority of prisoners are housed in public prisons, where conditions are not obviously better than they are in the private ones.

So why does this issue attract, as Humphreys notes, such “outsize attention"?

I suspect that this is an example of what philosopher Robert Nozick once called “normative sociology”: the study of what the causes of problems ought to be. It seems as if private prisons should be worse than public ones. The idea of profiting off of keeping someone in a cage makes us feel, for want of a better word, kind of squicky. So closing private prisons is a concrete step that can make the public feel better about the fact that so many Americans are locked up in terrible places.

The problem with normative sociology is that it can become a substitute for social science, and for policy improvements that actually make us better off. Most people in this country will never go to prison, which means most of them will never intimately care about whether they’re decent. That makes prison policy a particularly fertile field in which to plant an empty symbolic gesture and walk away. And meanwhile, in other fields that need deep weeding, the real problems can continue to thrive unchecked.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Megan McArdle at

    To contact the editor responsible for this story:
    Jonathan Landman at

    Before it's here, it's on the Bloomberg Terminal.