Privatization and Antitrust (and other aspects of Competition Policy) — Part 4

This is the fourth and last in a series of posts serial-blogging a draft chapter I’ve written on privatization and competition policy (including antitrust) for a forthcoming Stanford University Press book on Competition and the Role of the State. Here’s the first part, here’s the second part, and here’s the third part. What follows represents the current draft, though sans footnotes for the blog format. Sorry, I don’t have it posted on SSRN yet, so I can’t link to the full document, but drop me a line if you’d like the Word document.

In this final post, I explore how privatization can help escape anti-competitive public enterprise regimes; then I conclude.

Escaping Anti-Competitive Public Enterprise Regimes

Antitrust shouldn’t be seen as merely a form of damage control—a way of reining in firms in the transition from public-spirited but inefficient public ownership to rapacious and anti-consumer private ownership. The reality is that public ownership has dysfunctionalities of its own, and can even be anti-competitive.

In an early model, Sam Peltzman argued, for instance, that public enterprises subsidize voter-taxpayers with prices below the profit-maximizing level, using higher prices for nonvoters to cross-subsidize voters, and price-discriminating less among voters. This was pre-Sappington-Stiglitz, so Peltzman gave insufficient consideration to whether the government could regulate private enterprises to the same effect, but the earlier discussion suggests that an explicit incomplete-contracts model could plausibly show that governments can more easily control public firms than private ones.

In any event, it’s empirically observed that municipalities often engage in profit-making activities and “exert[] law-making power to exclude competitive challenges. . . . [M]onopoly profits frequently provide a politically preferable alternative to taxes as a general revenue source.” Public enterprise, after all, is presumptively somewhat responsive to the median voter in its own jurisdiction, but the median voter doesn’t necessarily care about social welfare.

In light of this, that various legal doctrines (at least in the U.S.) exempt governments from an otherwise pro-competitive regime is an independent argument in favor of privatization.

The most obvious example is the state action exemption from antitrust law. In Parker v. Brown, the Supreme Court wrote: “We find nothing in the language of the Sherman Act or in its history which suggests that its purpose was to restrain a state or its officers or agents from activities directed by its legislature.” Municipalities don’t automatically get this exemption because, unlike states, they’re not sovereign, but they can get a pass if their anti-competitive activities were authorized by the state “pursuant to state policy to displace competition with regulation or monopoly public service.” The policy must be “clearly articulated and affirmatively expressed,” but in Town of Hallie v. City of Eau Claire, the Supreme Court unanimously suggested that a state statute can “clearly contemplate” municipal anti-competitive conduct even if such conduct is merely “a foreseeable result” of permitted municipal activity.

Hallie justified this position with a “presum[ption], absent a showing to the contrary, that the municipality acts in the public interest”—partly because of the public scrutiny, disclosure regulation, and electoral accountability that municipalities have and private firms don’t. Electoral accountability, of course, is all well and good, but as I’ve just mentioned, one way for municipal officials to be perfectly faithful agents of their constituents would be to make monopoly profits at the expense of outsiders (potentially reducing taxes), which was the case in Hallie itself. How clear is clear and how foreseeable is foreseeable is a continuing area of ambiguity under antitrust state action doctrine.

The state action doctrine is problematic in other ways as well: the exemption has required an element of active state supervision, but how active is active likewise has never been fully spelled out. Moreover, when the challenged actor is a municipality rather than a private organization, the “requirement” is actually optional and merely serves “an evidentiary function,” to ensure “that the actor is engaging in the challenged conduct pursuant to state policy.” Between fully private actors and municipalities, there exists a “gray area consisting of hybrid state or local entities” where the vitality of the active supervision requirement is decided case by case.

Couldn’t we just fix the doctrine instead, rather than cutting this Gordian knot through privatization? Perhaps. But suppose there were no state action exemption. Or suppose we required more “clear articulation” and “active supervision,” limited the set of entities subject to the exemption, denied the exemption for market participants, and the like. Regular antitrust law may still be a bad fit for entities that aren’t traditional profit maximizers. Suppose, for instance, one sued the U.S. Postal Service for predatory pricing. Under current U.S. antitrust doctrine, one may still lose if the alleged predator wouldn’t plausibly be able to recoup its lost profits from the predation period. This is a plausible attitude to take when one is alleging predation by a profit-making firm; but a state-owned enterprise like the U.S. Postal Service may be acting to maximize revenues or employment rather than profits, so a predation claim that would be implausible against a private firm may well be plausible in the postal context. Sokol suggests alternative pricing tests that would facilitate the application of antitrust law in these cases, but grants that constructing an administrable test might be difficult.

The Dormant Commerce Clause is another example where public actors are given privileges that private ones lack. In C & A Carbone, Inc. v. Town of Clarkstown, the Supreme Court invalidated a flow control ordinance that required all trash producers (i.e., you and me) in the town to use a particular favored private trash processing facility; the ordinance, the Court wrote, was protectionist and undermined a national market in garbage. But in United Haulers Ass’n v. Oneida-Herkimer Solid Waste Management Authority, which differed only in that the favored facility was publicly owned, the Court upheld the ordinance on the grounds that government ownership was different: “Unlike private enterprise, government is vested with the responsibility of protecting the health, safety, and welfare of its citizens. These important responsibilities set state and local government apart from a typical private business.” Note how the Court fails to come to terms with Sappington-Stiglitz; and Justice Alito, in dissent, points out as much (though not in precisely those terms). Yes, the facility in Carbone was nominally private, but everyone regarded it as the town’s facility. The contractor had agreed to build the facility for free and transfer it to the town five years later for $1. The town guaranteed a minimum waste flow, allowed the contractor to charge an above-market tipping fee, and committed to make up any deficit in tipping fees if the waste flow was less than the guaranteed amount. “[F]or all practical purposes,” the facility “was owned by the municipality,” and the Court’s failure to treat the two cases similar was “exalt[ing] form over substance.”

This, then, is a case where nominal privatization, even under conditions of (otherwise) complete contracting, would make a real difference in the state’s ability to pursue anti-competitive policies. The case is even clearer when public actors are direct market participants—in which case doctrine gives them complete leeway to favor in-state interests. Privatization would thus directly serve competitive goals, since regulating a private market to achieve the same result would most likely violate the Dormant Commerce Clause.

These are all cases where privatization would further the goals of national competition. Again, perhaps this is a second-best solution relative to just fixing the offending doctrines. But Congress has “shown little desire” to meddle with antitrust; Dormant Commerce Clause doctrine is constitutional; more generally, as I’ve noted previously, theoretically second-best fixes are often practically the first-best option for policymakers.

Moreover, one can say the same of any regime that treats public and private organizations differently, even when the regime has little to do with competitiveness. For instance, the Supreme Court, in College Savings Bank v. Florida Prepaid Postsecondary Education Expense Board, has expressly rejected a market participant exception to state sovereign immunity. This means that a lawsuit against a state or one of its instrumentalities subject to sovereign immunity would fail where a lawsuit against a similar private organization might succeed—even if the organization were blatantly engaged in false advertising of financial instruments, which was the allegation in Florida Prepaid. The ability to violate generally applicable laws with impunity, if nothing else, operates as a barrier to entry against private competitors.

One could say the same of various mini-immunity doctrines, like public prisons’ (but not private prisons’) immunity from tort law with respect to their discretionary activities, or the rule that grants qualified immunity under § 1983 to public state corrections officers but denies it to private ones. (Of course, one should also take account of contrary doctrines that treat the public sector more onerously, like the procedural requirements of the Freedom of Information Act, the Administrative Procedure Act, or the requirement of voter approval of bond issues.)

Nor are these merely contingent quirks of U.S. law: public enterprises are often treated more leniently under the prevailing regulatory regime, whether by explicit doctrine or by favoritism and non-enforcement. If the underlying regulatory regime (the one that the public enterprises are evading) is itself unjustified, one may think that the more non-enforcement, the better—but it’s unlikely that a regime of selective non-enforcement, where the sole beneficiaries are public enterprises, is even second-best. Thus, privatization can serve the salutary role of moving specially exempted public enterprises into the more stringent “default” regime.

None of these are knock-down arguments against government ownership: the standard theoretical arguments discussed earlier still apply, so government ownership can still be more efficient than private ownership in many cases. The foregoing considerations are merely examples of institutional details that, for better or worse, differ as between the public and private sectors. To the extent these details are hard to change, the most efficient ownership structure will differ from the efficient ownership structure that one would otherwise support.


It’s clear, then, that privatization is no cure-all: an economic policy that encourages entry and discourages monopolistic behavior, hard budget constraints, good corporate governance, liberalized trade, and the like, are all necessary for privatization to produce the most benefits.

In light of this, one might ask whether, if these complementary policies are adopted, privatization is required at all. State-owned enterprises have found unexpected flexibility in various transition countries when faced with economic reform—hard budget constraints, restrictions on credit, competitive markets (including liberalized trade and currency convertibility), and bankruptcy laws have spurred improvements in corporate governance and competitiveness. Even in the U.S., government enterprises can be seen perfoming better when faced with a more competitive environment: witness the improvements in public schools in the face of school vouchers or other accountability mechanisms, or improvements in municipal services when public- and private-sector providers compete with each other under a regime of competitive neutrality.

Certainly, this is better than nothing. Can we do even better by adding privatization?

Shleifer says yes: it’s true that mere corporatization—shifting control rights from politicians to managers without privatizing—can stimulate restructuring, but this depends on how much de facto control or influence over public firms politicians still retain. “In a country like Russia,” he writes, “where the mechanisms of political influence are numerous and the politicians’ demand for influence is high, corporatization by itself is a rather weak measure.” Privatization—giving managers (or outside shareholders) cash-flow rights in addition to control rights—depoliticizes firms still further because it strengthens managers’ and shareholders’ opposition to government control (both by creating a political class of owners, and by adding foregone profits to the managers’ costs of following government dictates) and because it weakens the avenues by which government can control firms.

And what’s true “[i]n a country like Russia” is no less true in a relatively advanced rule-of-law state like the U.S. Recent experience with companies in which the government has recently taken an ownership stake, like GM, shows that government firms are characterized by government control no matter how strenuously the government insists otherwise. Moreover, this is the sort of control that would be much harder to pass through Congress if it were to be imposed regulatorily on an entire private industry. Independent private stakeholders who can selfishly assert their own interests—and demand potentially expensive bribes if they are to do the government’s bidding—are necessary not just in Russia but everywhere.