This post continues my series on my upcoming Stanford Law Review paper on Privatization and the Law and Economics of Political Advocacy (see here for the technical paper). Two posts ago, I set out the basic economic theory—be sure to read that post if you're having trouble following (or if you're disagreeing with) the discussion of free-riding. And in the last post, I discussed various miscellaneous points about the theory and how it applies to privatization. Now, on to the primary case study in the paper: prisons.

* * *

Now let us apply the theory to a real-world industry subject to the "political influence" critique of privatization: prisons. When I speak of "pro-incarceration advocacy," I use the term "advocacy" broadly to include any use of political influence, licit or illicit, including endorsements, political contributions, lobbying, and bribes. And I use the term "incarceration" as a shorthand to include the criminalization of a greater range of behavior, more active enforcement, greater reliance on imprisonment, longer sentences, and less parole; thus, endorsing a politician for being "tough on crime," donating money to a "Three Strikes" initiative, or testifying in favor of a "truth in sentencing" law all count as advocating incarceration.

Consider the main political actors in the prison industry: the private prison firms and the public corrections officers' union. Without privatization, the public sector is the monopoly provider of prison services, and the corrections officers' union enjoys the benefits that flow from serving the whole system. Now suppose that part of the system is privatized. At first, the public sector is clearly the dominant sector, that is, the sector with the largest proportion of total benefits from provision of the service—from 100% of the industry, it has gone down slightly, and the private sector is not big enough to even come close. Because it has shrunk, it is less willing than it used to be to spend money on reforms that would increase the size of the prison pie. And because the private sector is tiny, it free-rides. Provided the public sector stays the dominant sector, privatization will always have this effect—because, in terms of the model presented above, any reform that shrinks the dominant sector will reduce industry-expanding advocacy. As it happens, this proviso is true in the case of prisons. At current levels of privatization, the public sector both has a larger industry share and extracts more benefit from the system than does the private sector.

We can easily perform some rough estimates to verify this (I provide the detailed derivations of the numbers elsewhere):

• Industry share. The private sector has a smaller share of the industry. Of the 1.5 million prisoners under the jurisdiction of federal or state adult correctional authorities in 2004, 7% were held in private facilities; this includes 14% of federal prisoners and 6% of state prisoners. Among the 34 states with at least some privatization, the median private percentage was 8--9%. If we are interested in the private share of marginal prisoners—i.e., how likely a prisoner is to go to a private prison if he is convicted today—the private share becomes larger, mainly because private firms have absorbed much of the recent growth in federal incarceration. A reasonable estimate of the private share of marginal prisoners over the period 2000--2005 yields 6% for state systems, 54% for the federal system, and 22% overall.

• Private-sector profitability. The profits of the private sector are low. If the industry were perfectly competitive—like in textbook models of perfect competition—every firm would make zero economic profit. "Economic profits" measures how profitable a company is relative to other ways of investing one's money; thus, "zero economic profits" means not that firms are making no money but rather that all firms are doing as well as the rest of the market. In such a (hypothetical!) world, firms wouldn't care whether their market were growing or shrinking, because they would be indifferent between running prisons and putting their money into the stock market. This is, of course, unrealistic—the prison industry is oligopolistic, so prison firms do make some profit. But not much: 10% would be a generous estimate of prison firms' profitability.

• Public-sector rents. Public-sector correctional officers, on the other hand, benefit substantially from public provision of prisons, because their wages are quite a bit above what they would be making in the private sector—by about 30--65%. This is a lot of money, because wages are about 60--80% of most prisons' operating expenses.

I make the assumptions—oversimplified but common in the economic literature on firms and unions—that firms maximize profits and that unions maximize total "union rents" (that is, the difference between public-sector and private-sector wages, times the size of the public sector). Trying to put these numbers together rigorously requires a fair amount of algebra, which I provide elsewhere. But it should be intuitively clear that the public sector extracts substantially more benefit from any given prison than does the private sector; and it is likewise clear that the public sector has a greater share of the industry than the private sector.

Thus, the public sector enjoys a greater benefit from prison provision than the private sector does, perhaps by an order of magnitude. This model predicts that the public sector should be doing all the pro-incarceration advocacy, and the private sector should be entirely free-riding. Moreover, privatization reduces the public sector's share of total benefits. So, at current levels, privatization should cause total pro-incarceration advocacy to decrease.