Privatization and the Law and Economics of Political Advocacy, Part 3 -- The Model:

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). This installment is not connected to privatization specifically at all, much less prisons, but gives (in plain English) the basic economic theory behind public goods and free riding.

* * *

I now present the main model I use to predict how industry actors will react to privatization. The central feature of the model is that industry-increasing advocacy is a public good. Privatizing part of the industry therefore introduces a collective action problem: Unless everyone in the industry cooperates with each other, they will together spend less on industry-increasing advocacy than a single firm would if it covered the whole industry, because a portion of their expenditures will benefit their competitors.

This intuition should not be surprising, as it is standard in the literature on public goods. When a good is private, everyone pays for, and enjoys, only his own consumption. By contrast, when a good is public, in the classic model, everyone benefits from the total amount, and this amount is determined by the total amount of contribution.

If we benefit from our national defense, we benefit from the full amount, not from the chunk we paid for; we cannot be excluded from the full benefit, no matter how little we paid; and the total amount of national defense is just determined by how much money Congress allocated to national defense from the Treasury. A tax-funded program that improves air quality benefits everyone who breathes the relevant air, whether or not they contributed to the program; and the total improvement is just determined by the amount of resources directed toward that goal.

Similarly, contributing to a candidate’s campaign benefits all of his supporters; and it is not too implausible to say, as an approximation, that to the extent the money he raises and spends affects his probability of winning, it is only the total amount of money that matters.

In all these cases, the temptation to free ride off one’s fellows’ contributions is strong—so strong that the category of “public goods” is standard among economists as a case of “market failure.”

To explore the basic model, consider a monopolist, who's willing to invest some amount of money in lobbying to increase the size of his industry. To determine that amount, he weighs the benefit that his money can buy—the expansion of the industry is worth something to him, and money can help his policy pass—against the cost of the lobbying.

If that firm is broken up into two smaller firms—say a 90% incumbent firm and a 10% splinter firm—the larger incumbent isn’t willing to spend as much as it used to be, because the costs of lobbying are the same while the benefits are 10% less than they used to be. And the smaller splinter firm won’t be willing to spend anything, because it will be satisfied free-riding off the larger incumbent’s lobbying. Thus, splitting up an industry tends to decrease industry-expanding lobbying.

The rest of this post will illustrate this intuition graphically.

liberty (mail) (www):
Similarly, contributing to a candidate's campaign benefits all of his supporters; and it is not too implausible to say, as an approximation, that to the extent the money he raises and spends affects his probability of winning, it is only the total amount of money that matters.

Unless his winning only matters to the extent that once he has won he repays his donors with favors. If, for example, politicians aren't ideological and it doesn't really matter if they are (D) or (R), instead the winning politician simply uses his power to leverage favors for his donors. Then it isn't a public good anymore and each contribution is a private good that only benefits the contributor.
4.5.2007 12:58pm
liberty (mail) (www):
Of course, in your example you could say that the benefits to one contributor would be the public good which benefits the other (would-be) contributors. However, this requires that the favors granted are sharable. If the favors are market power (or no-bid contracts or subsidies) for individual firms, rather than an increase in incarceration, this would not be the case.
4.5.2007 1:01pm
Adam R (mail):
This assumes that the industry actors do not cooperate. It is quite easy to cooperate one's lobbying efforts, and as far as I know legal to do so. It may technically be more economically rational for the small company to try to free ride off the big company, but in real life people (entities) quite often contribute to a common enterprise even when strict economic logic suggests that they should try to free ride. We often hear talk of various industry lobby's as a whole, e.g. the gun lobby, the beef lobby, the auto manufacturers' lobby, etc., which unless the preception is totally wrong, suggests that in fact cooperation is quite common and easy.
4.5.2007 1:43pm
Isn't it possible that by splitting the market into two competitors, the theoretical increased productivity of both firms — and the inability of the monopolist to continue restricting its goods to artificially drive up prices — will make the market larger in absolute terms? And thus isn't it possible, theoretically, that the 90% firm might spend more than $1 if the marginal benefit to the firm has moved because of the increase in the size of the market?
4.5.2007 2:21pm
Sasha Volokh (mail) (www):
bjr: Yes, if you think (1) the split will make prices go down and (2) the market is sensitive to the fall in price. In the case of prisons, where "split" means "privatization," (1) is probably true, though this is controversial. In the antitrust case where a monopoly becomes a duopoly, (1) is generally true because duopoly prices are lower than monopoly prices. In the most general case, it's not clear whether (1) is true, because my industry shares aren't antitrust-style "market shares"; for instance, if the company had a national market and the two firms are divided into geographic areas in which each is a monopolist, there may be no change in price.

As for (2), the effect of price decreases on the amount of the service is probably quite weak in the case of prisons; the cost of incarceration has increased a lot over several decades, but the political system hasn't seemed very responsive to that. For what it's worth, Bruce Benson, a libertarian, opposes prison privatization because he believes it's more efficient -- he wants to keep incarceration more expensive so fewer drug offenders get behind bars.

Anyway, yes, in the general case, one would want to take into account that additional complexity, where the split might make price go down which might make the industry increase.

Adam R: Stay tuned to a future post for a discussion of whether the sectors will cooperate. Just to anticipate: In this case, if the firms are identical, then cooperate just restores the status quo, and the split-up has zero effect. But in the prison context, the private sector is more competitive, and in fact its profits are fairly low. So even if the two sectors cooperated, the total profit would still be lower than before privatization; so advocacy would still increase.

liberty: This is why I said in this post (and will restate at greater length a couple of posts from now) that industry-increasing advocacy, that is, advocacy for the sort of reform that increases the size of the industry (e.g., the prison sector) generally, is a public good.

Contributing in favor of getting a particular project is often a private good for an individual firm, and contributing in favor of privatization generally is a private good if you consider the private sector as a bloc. Back to the prison case: If the number of prisoners increases, it's in practice harder for a provider to lobby in such a way that the increase benefits him specifically.
4.5.2007 2:56pm
Doesn't this model of a "monopolist" fail to accurately capture the nature of almost any organization? While certain firms may present a united front in their marketing to the public, there are almost always internal splits that will result in divisions that can and will pursue their own lobbying efforts. For example, think of the tension between those pursuing cellular technology at AT&T in the early 1980's, versus the more-established land line division.

True, there could be unitary monopolists, especially when a monopoly is young (i.e. the founder is alive), or when the monopoly is small (geographic), or if there is a very tightly integrated board of directors, but even in those situations, the model breaks down over time. I will grant that you're distilling a model that will probably prove useful in specific circumstances where a discrete analysis of a particular market/legal situation is required; I wonder about its broader application.
4.5.2007 3:17pm
Sasha Volokh (mail) (www):
Random: Yes, I plead guilty, I'm abstracting away from any agency issues within the firm and treating the firm as a profit-maximizing "black box." That doesn't mean I believe it, but (1) the black box view can be useful sometimes, and (2) in any event, I'm responding to people arguing against privatization who do say that the problem is the private prison firms that advocate incarceration for simple profit-maximizing reasons. So I need to assume simple profit-maximizing here to be able to show that even under that simple model you don't necessarily get more advocacy under privatization.
4.5.2007 5:02pm
Sasha, I know nothing about this subject, but arguably the increase in prison privatization itself is a reaction to the increased cost of incarceration, and suggests the "political system" -- i.e., state and local government -- is indeed responding to this price increase.

I look forward to discussing this with you at the next IHS weekend seminar on prison reform, where I expect to hear your next lecture on Yeats -- "All changed, changed utterly; a terrible beauty is born."

(Ben R.)
4.5.2007 8:00pm