Privatization and Antitrust (and other aspects of Competition Policy) — Part 3

This is the third 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, and here’s the second 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.

Earlier, I discussed theoretical reasons to think that whether a service is privatized may not be relevant, and then practical reasons why the irrelevance assumptions may not hold up. Now, I go on to see how other policies, like corporate governance or antitrust, affect the efficiency of privatization.

Privatization and Corporate Governance

But the success of privatization also depends on other policies. Privatization may promote macroeconomic growth, but probably most so when combined with “in-depth institutional reforms.”

Consider, for instance, agency costs within private firms, which most models assume away. This omission is fine when corporate law does a tolerable job of controlling managerial self-dealing, not so fine in countries where the institutions to control self-dealing are underdeveloped. Thus, Bernard Black, Reinier Kraakman, and Anna Tarassova argue that privatization in Russia went awry because there was no brake on asset stripping and other forms of looting by managers —the trouble may thus have been not the oligarchs as such, but a system that let managers maximize their utility without the check of corporate law.

This is perhaps why privatization in the former Soviet republics has enhanced productivity when the new owners were foreign but much less so (if at all) when the new owners were domestic: perhaps foreign companies had better corporate governance in their home countries, or perhaps foreign companies are less vulnerable to political pressure from shady post-Soviet governments.

Not that bad corporate governance is a problem that magically appears when an enterprise is privatized. Governance in public enterprises is also notoriously bad; Daniel Sokol documents that the quality of governance in government postal enterprises varies widely, that governance is better where postal enterprises are commercialized, corporatized, and competitive, and that some of the worst offenders are in advanced economies like the United States. Thus, even bad privatization might be better than business-as-usual public enterprise, though a more sophisticated program, whether better private corporate governance or a corporatized state-owned enterprise, might be more efficient still.

The regulation of nonprofits is also important. Glaeser and Shleifer show that nonprofit status can be valuable: by weakening the provider’s incentives to maximize profits, nonprofit status can be a valuable signal of quality when quality itself is non-verifiable; moreover, altruistic entrepreneurs will tend to be attracted to the nonprofit form. This matters for privatization, given the role of nonprofits (both secular and religious) in public-private partnerships worldwide, and the importance of “faith-based initiatives” in the American debate: Besley and Ghatak show that, when both a provider and the government can make productive investments in a project, and when the provider is altruistic, then the provider should own the project if it values it more than the government does. As a result, nonprofit regulation can play a valuable role in preventing profit-making firms from abusing the nonprofit form.

Privatization and competition

Privatization of natural monopolies

Hart, Shleifer, and Vishny write that “competition strengthens the case for contracting out,” if “consumers can assess quality and have a choice among competitive suppliers.” This might describe auto manufacturers but not, say, electric utilities.

Even if consumers are powerless, as they will necessary be in the case of true natural monopolies, competitive bidding for the right to provide the service—perhaps even between public and private providers—can strengthen the case for privatization. Franchise bidding—where the winner of the auction is the potential provider who commits to offer the lowest prices or (in the case of multidimensional services) the best overall deal—has been proposed to counter monopoly power in utility regulation, and has actually been used in, say, cable television regulation. Of course, if the formula for picking the auction winner (for instance, the rule for what’s the “best overall deal”) embodies a great deal of discretion, the effectiveness of the mechanism depends heavily on the benevolence of the government officials running the bidding process. Moreover, competition at the bidding stage, contract enforcement difficulties, and reputational concerns matter as well.

Cable television is an example where whatever monopoly exists is purely local: the bidding isn’t over cable access for a whole state or country, but over a local area. A similar example in this respect is prisons. Prisons are contracted out one at a time, and no state has contracted out anything like all its prisons: as of 2010, only 8% of prisoners were housed in a public prison, about 7% in state systems and about 16% in the federal system. Of the 30 states that do some contracting out, the median percentage of inmates in private prisons is about 10%, and no state’s percentage is above 45%. The result is a fairly concentrated, but not monopolistic, private prison industry, dominated by a couple of firms. Prison provision as a whole thus consists of “the public sector,” with a 92% “market share,” and a few private firms with an 8% market share—or, to speak more precisely, a public sector in each state with a share between 55% and 100% and some firms sharing the rest.

Private provision, combining a bidding process with strong cost-cutting incentives, has thus increased competition without the need for a special antimonopoly policy. Costs have dropped somewhat, though the magnitude of the drop depends heavily on the accounting assumptions one uses, with some methods yielding savings on the order of 20% and others yielding savings on the order of 3% (or even finding in some years that the public sector is more cost-effective). And while quality studies have been notoriously poor, careful comparative studies haven’t given a strong edge to either sector. Naturally, whatever success the enterprise may have had—and this is hotly disputed—depends on minimally competent monitoring, a working tort and constitutional litigation system, and reputational concerns. Moreover, the full potential of prison privatization is as yet untested, since making compensation depend on performance measures like low recidivism (as opposed to giving flat per diems with penalties for particular listed in-prison incidents) and giving prison companies the flexibility to experiment (as opposed to incorporating the public prison rulebook in the contract) are still in their infancy.

Such partial, piecemeal privatization may have also had salutary political economy effects: while public prison-guard unions are big pro-incarceration lobbyists, the evidence that private prison firms lobby for incarceration is quite slim (despite activists’ frequent claims to the contrary), perhaps because privatization, by fragmenting the industry, has introduced a collective action problem in lobbying. “In a roundabout way, . . . privatization is a form of antitrust, and antitrust is a form of campaign finance regulation.”

These have been examples where privatization can be pro-competitive even in the absence of specific anti-monopoly policies. Sometimes, though, if the goal is to prevent the exercise of monopoly power, specific policies may be necessary. For natural monopolies, those who are pessimistic about the possibilities of successful franchise bidding schemes will want to stick to more traditional modes of public utility regulation, like rate-of-return regulation or price caps.

Antitrust and Other Entry-Promoting Policies

Outside the area of natural monopolies, there is of course the option of restructuring enterprises before privatizing them piecemeal for greater competition, or relying on antitrust law. Privatization is frequently criticized for proceeding without effective anti-monopoly provisions, merely replacing a government monopoly with private monopoly—though as I’ve mentioned above, political concerns may sometimes dictate otherwise in practice. The simple fact that monopolies can be sold for more than competitive companies, and thus raise more government revenue, may explain much reluctance to restructure.

In Mexico, the two major airlines, Aeromexico and Mexicana, were privatized in the late 1980s, and de facto merged in 1993. (Their immediate post-privatization history is a good case study in the synergy of privatization with other policies: Aeromexico was able to restructure more effectively because it was privatized out of bankruptcy and wasn’t tied down by its earlier collective bargaining agreements.) Gordon Hanson writes that the merged companies controlled over 70 percent of the domestic air travel market in Mexico. The presence of antitrust would have made a significant difference in preventing the consolidation of the companies’ market power, perhaps by blocking the de-facto merger. But Mexico’s antitrust law didn’t take effect until a few months later.

The existence of robust antitrust law might thus assuage some fears that newly privatized state entities will gain monopoly power. Not that antitrust law is a panacea—in the U.S., a company that starts out as a monopoly, or that grows to be a monopoly through unobjectionable means, is generally free to price how it likes, certainly including monopoly pricing and perhaps even including limit pricing. Continuing antitrust activity in recently privatized areas like telecom in France and postal service in Germany show that regulation doesn’t magically lead to markets mimicking perfect competition. Robert Feinberg and Mieke Meurs write that, in Eastern European transition countries, early antimonopoly efforts had little effect—“[m]any of the largest firms form[ed] a single technological unit, defying efforts of demonopolizers to break them into smaller enterprises,” efforts to split firms along regional lines only resulted in regional monopolies, and existing connections among managers facilitated collusion.

The recently uncovered international air cargo cartel—a conspiracy involving at least 20 airlines—suggests that antitrust may not work perfectly even in its core area of policing price fixing, though its effectiveness would surely increase with higher fines, aggressive leniency programs, and personal sanctions on responsible corporate officers. This particular cartel managed to acquire a surprisingly large number of participants primarily because the alternative to price fixing, for many of the participants, was not just lower profits but bankruptcy. Moreover, the implications of the cartel for privatization policy are ambiguous, since at least two of the participants, China Airlines and Singapore Airlines, are directly or indirectly government-owned.

Add to these problems the inherent danger that antitrust will wrongly condemn legitimate business practices, and that antitrust enforcement in developing countries is likely to be less competent in this regard than in the developed world. Add as well the public-choice concern that companies will use antitrust to shield themselves from their pesky and too-successful competitors. I’m not arguing that antitrust is useless as a complement to privatization, but that there may be other, perhaps even more useful, complementary policies.

Feinberg and Meurs suggest, again with respect to Eastern Europe, that entry-facilitating policies can be even more valuable than explicit anti-monopoly laws. Mexico is a good case study on the importance of entry: in 1991, it deregulated air travel, first, by reducing domestic carriers’ entry and fare restrictions, and, second, by reducing barriers to U.S. carriers. At the time of Hanson’s article, a small, low-cost airline had entered the domestic Mexican market (several more have entered since then), and fares were already significantly lower for international than for internal flights, where U.S. carriers provided significant competition.

In short, free trade and entry liberalization are themselves competition policies, which can be effective complements to, or even substitutes for, antimonopoly laws. One might add to the list a liquid capital market, as well as general deregulation, both of which one expects to facilitate entry.

Next time: privatization as a tool for escaping anti-competitive public enterprise regimes, and the conclusion.