The Volokh Conspiracy - Prison Rape and the Limitations of Government:

This post by Ampersand rounds up the recent blogosphere reaction to the scandal of widespread prison rape. As he notes, conservative, libertarian, and liberal bloggers all agree that more should be done to curb prison rape. Yet, he concludes, "this is a curious case where it appears that everyone agrees, yet nothing ever gets done."

For what it's worth, I agree with everyone else that we should do more to prevent the rape of prisoners. But the government's failure to address the problem is not accidental. Government is responsive to those who have political power, and prisoners are the classic example of a group that has almost no power, and is generally unpopular with those who do. In most states, prisoners don't even have the right to vote, and of course their ability to wield political power in other ways (activism; campaign contributions; lobbying, etc.) is also extremely limited. Most of the general public, by contrast, is either unaware of the problem of prison rape or doesn't care about it very much. And, of course, measures to make it easier for prisoners to sue or otherwise alleviate their plight will be strongly opposed by prison guards unions and other influential interest groups.

This is an extreme case of an important broader lesson about the nature of government: it usually can't be relied on to protect the political powerless or even the relatively weak. As I have blogged in the past, the same point applies (albeit with less force) to claims that a strong government will be good for the poor. Because the poor have little political power, government intervention is more likely to cut against their interests than in their favor - especially when the needs of the poor conflict with those of middle class or wealthy interest groups.

Returning to the prison rape question, we probably cannot adopt here the standard libertarian solution of simply ending government involvement with the issue - at least not without unacceptable social costs. We can, however, reduce that involvement. As I pointed out several months ago, nonviolent drug offenders account for 55% of all federal prisoners and 21% of state prisoners, and probably account for even higher percentages of all incarcerated nonviolent offenders. Eliminating or cutting back on the War on Drugs - which is highly desirable for many other reasons - would have the beneficial secondary effect of greatly reducing the number of people exposed to prison rape.

A second way of reducing (though not eliminating) government involvement in this field is prison privatization. As co-blogger Sasha Volokh demonstrates in a recent paper, replacing government-run prisons with private ones may well reduce the overall lobbying power of the prison industry, and thereby make it easier to both reduce overall incarceration levels and force improvements in prison conditions. Even under privatization, it would still be difficult to force through legislation that reduces incarceration rates or protects prisoners. But it would be easier to achieve this than under the status quo.

Ultimately, efforts to alleviate the problem of prison rape are likely to fail politically unless they include policies that benefit constituencies more powerful than the prisoners themselves. My two suggestions both have this virtue. Constraining the War on Drugs would benefit the many people who want to see limited legal drug use, such as legalized medical marijuana (a proposal popular even in conservative states). Prison privatization will, of course, attract the support of those private firms that hope to get contracts to run private prisons.

UPDATE: I have slightly altered the original title of this post to make my point more clear.

UPDATE #2: Mark A.R. Kleiman responds to this post with a combination of arguments and overrwrought rhetoric here. Kleiman suggests that prison rape is not an example of the general shortcomings of government because rape is relatively rare in federal prisons:

rape remains a rarity within the Federal prison system; the U.S. Bureau of Prisons is generally a higher-performing organization than the state prisons. So if prison rape demonstrates some sort of generalized failure of "government," how come it isn't much of a problem for the biggest government of all?

An obvious response is that federal prisons, unlike state prisons, contain relatively few people convicted of violent crimes. Most murders, rapes, assaults, etc., are state crimes. Federal prisoners are mostly nonviolent small-time drug dealers. A prison population with few people convicted of violent offenses will also have relatively fewer prisoners willing to rape fellow inmates. As of 2003, 52% of state prisoners were serving sentences for violent crimes. compared to only about 10% of federal prisoners (figure calculated from this Bureau of Justice Statistics report). Perhaps even more to the point, the BJS report indicates that some 12% of state prisoners are incarcerated for rape or "other sexual assault," while almost no federal prisoners fall into these categories. I'm no criminologist, but I suspect that a prison population with a large percentage of rapists is going to have a lot more prison rape than one with very few.

Kleiman also argues that Sasha Volokh's paper on prison privatization, which I cited above, doesn't support my conclusion that privatization might reduce the overall lobbying power of the prison industry. He quotes a part of the abstract, but not this part:

In fact, privatization may well reduce the industry's political power: Because advocacy is a "public good" for the industry, as the number of independent actors increases, the largest actor's advocacy decreases (since it no longer captures the full benefit of its advocacy) and the smaller actors free-ride off the largest actor's contribution. Under some plausible assumptions, therefore, privatization may actually decrease advocacy, and under different plausible assumptions, the net effect of privatization on advocacy is ambiguous.

Readers can judge for themselves whether or not this is consistent with my characterization of the paper as stating that "replacing government-run prisons with private ones may well reduce the overall lobbying power of the prison industry."

Kleiman also argues that the real way to address prison rape is to improve the quality of prison management and to elect politicians who will support reform in this area. These are worthy objectives, but he does not explain how they are to be achieved given that 1) prisoners themselves have almost no political power, and 2) most voters don't seem to care about the issue. In the course of making this point, Kleiman also conflates Republican Party advocacy of "tough on crime" policies with the fostering of prison rape by promoting "hatred of criminals." This ignores the obvious fact that some criminals arguably deserve to be hated, and that it i not inconsistent to support harsh penalties for criminals, while also opposing prison rape.

Finally, Kleiman claims that "Somin, who would like to see the Libertarian Party get out of the way of the project of maintaining Republican electoral dominance, is in effect voting for more incarceration, more brutality, and more rape every time he pulls the Republican lever."

I have indeed argued for getting rid of the Libertarian Party. But any reasonably fair-minded person who read my post on the subject should be able to see that the reason is not to "maintain Republican electoral dominance" but because I believe that the LP is an obstacle to promoting libertarian policies. Absent the LP, libertarians would have greater leverage over the policy positions of the two major parties. If I was as committed to "Republican electoral dominance" as Kleiman supposes, I would not have supported the Democratic takeover of the House of Representatives last fall.

As for the claim that the Republicans are unique in supporting "incarceration" and "brutality" and prison rape, this charge would hold more water if there were proof that the Democratic Party's positions on these issues are significantly different from those of the Republicans. Kleiman himself notes that the Democrats have been "cowed into inaction" on these questions. The fact - which Kleiman correctly notes - that many of them are privately more sympathetic to efforts to address these issues than most Republicans is cold comfort unless and until they actually act on those convictions.

	[Ilya Somin, February 17, 2007 at 4:41pm] Trackbacks Prison Rape and the Limitations of Government: This post by Ampersand rounds up the recent blogosphere reaction to the scandal of widespread prison rape. As he notes, conservative, libertarian, and liberal bloggers all agree that more should be done to curb prison rape. Yet, he concludes, "this is a curious case where it appears that everyone agrees, yet nothing ever gets done." For what it's worth, I agree with everyone else that we should do more to prevent the rape of prisoners. But the government's failure to address the problem is not accidental. Government is responsive to those who have political power, and prisoners are the classic example of a group that has almost no power, and is generally unpopular with those who do. In most states, prisoners don't even have the right to vote, and of course their ability to wield political power in other ways (activism; campaign contributions; lobbying, etc.) is also extremely limited. Most of the general public, by contrast, is either unaware of the problem of prison rape or doesn't care about it very much. And, of course, measures to make it easier for prisoners to sue or otherwise alleviate their plight will be strongly opposed by prison guards unions and other influential interest groups. This is an extreme case of an important broader lesson about the nature of government: it usually can't be relied on to protect the political powerless or even the relatively weak. As I have blogged in the past, the same point applies (albeit with less force) to claims that a strong government will be good for the poor. Because the poor have little political power, government intervention is more likely to cut against their interests than in their favor - especially when the needs of the poor conflict with those of middle class or wealthy interest groups. Returning to the prison rape question, we probably cannot adopt here the standard libertarian solution of simply ending government involvement with the issue - at least not without unacceptable social costs. We can, however, reduce that involvement. As I pointed out several months ago, nonviolent drug offenders account for 55% of all federal prisoners and 21% of state prisoners, and probably account for even higher percentages of all incarcerated nonviolent offenders. Eliminating or cutting back on the War on Drugs - which is highly desirable for many other reasons - would have the beneficial secondary effect of greatly reducing the number of people exposed to prison rape. A second way of reducing (though not eliminating) government involvement in this field is prison privatization. As co-blogger Sasha Volokh demonstrates in a recent paper, replacing government-run prisons with private ones may well reduce the overall lobbying power of the prison industry, and thereby make it easier to both reduce overall incarceration levels and force improvements in prison conditions. Even under privatization, it would still be difficult to force through legislation that reduces incarceration rates or protects prisoners. But it would be easier to achieve this than under the status quo. Ultimately, efforts to alleviate the problem of prison rape are likely to fail politically unless they include policies that benefit constituencies more powerful than the prisoners themselves. My two suggestions both have this virtue. Constraining the War on Drugs would benefit the many people who want to see limited legal drug use, such as legalized medical marijuana (a proposal popular even in conservative states). Prison privatization will, of course, attract the support of those private firms that hope to get contracts to run private prisons. UPDATE: I have slightly altered the original title of this post to make my point more clear. UPDATE #2: Mark A.R. Kleiman responds to this post with a combination of arguments and overrwrought rhetoric here. Kleiman suggests that prison rape is not an example of the general shortcomings of government because rape is relatively rare in federal prisons: rape remains a rarity within the Federal prison system; the U.S. Bureau of Prisons is generally a higher-performing organization than the state prisons. So if prison rape demonstrates some sort of generalized failure of "government," how come it isn't much of a problem for the biggest government of all? An obvious response is that federal prisons, unlike state prisons, contain relatively few people convicted of violent crimes. Most murders, rapes, assaults, etc., are state crimes. Federal prisoners are mostly nonviolent small-time drug dealers. A prison population with few people convicted of violent offenses will also have relatively fewer prisoners willing to rape fellow inmates. As of 2003, 52% of state prisoners were serving sentences for violent crimes. compared to only about 10% of federal prisoners (figure calculated from this Bureau of Justice Statistics report). Perhaps even more to the point, the BJS report indicates that some 12% of state prisoners are incarcerated for rape or "other sexual assault," while almost no federal prisoners fall into these categories. I'm no criminologist, but I suspect that a prison population with a large percentage of rapists is going to have a lot more prison rape than one with very few. Kleiman also argues that Sasha Volokh's paper on prison privatization, which I cited above, doesn't support my conclusion that privatization might reduce the overall lobbying power of the prison industry. He quotes a part of the abstract, but not this part: In fact, privatization may well reduce the industry's political power: Because advocacy is a "public good" for the industry, as the number of independent actors increases, the largest actor's advocacy decreases (since it no longer captures the full benefit of its advocacy) and the smaller actors free-ride off the largest actor's contribution. Under some plausible assumptions, therefore, privatization may actually decrease advocacy, and under different plausible assumptions, the net effect of privatization on advocacy is ambiguous. Readers can judge for themselves whether or not this is consistent with my characterization of the paper as stating that "replacing government-run prisons with private ones may well reduce the overall lobbying power of the prison industry." Kleiman also argues that the real way to address prison rape is to improve the quality of prison management and to elect politicians who will support reform in this area. These are worthy objectives, but he does not explain how they are to be achieved given that 1) prisoners themselves have almost no political power, and 2) most voters don't seem to care about the issue. In the course of making this point, Kleiman also conflates Republican Party advocacy of "tough on crime" policies with the fostering of prison rape by promoting "hatred of criminals." This ignores the obvious fact that some criminals arguably deserve to be hated, and that it i not inconsistent to support harsh penalties for criminals, while also opposing prison rape. Finally, Kleiman claims that "Somin, who would like to see the Libertarian Party get out of the way of the project of maintaining Republican electoral dominance, is in effect voting for more incarceration, more brutality, and more rape every time he pulls the Republican lever." I have indeed argued for getting rid of the Libertarian Party. But any reasonably fair-minded person who read my post on the subject should be able to see that the reason is not to "maintain Republican electoral dominance" but because I believe that the LP is an obstacle to promoting libertarian policies. Absent the LP, libertarians would have greater leverage over the policy positions of the two major parties. If I was as committed to "Republican electoral dominance" as Kleiman supposes, I would not have supported the Democratic takeover of the House of Representatives last fall. As for the claim that the Republicans are unique in supporting "incarceration" and "brutality" and prison rape, this charge would hold more water if there were proof that the Democratic Party's positions on these issues are significantly different from those of the Republicans. Kleiman himself notes that the Democrats have been "cowed into inaction" on these questions. The fact - which Kleiman correctly notes - that many of them are privately more sympathetic to efforts to address these issues than most Republicans is cold comfort unless and until they actually act on those convictions.