Mobile Edition
Print
Get the Mag
Weekly UpdatesIn contrast, the view of regulation and antitrust embodied in AT&T is not merely that they are two independent means for limiting the exercise of market power. Rather, regulation creates circumstances under which antitrust becomes relevant, in ways that antitrust would not if a firm with monopoly power were not regulated. In that sense, regulation does not diminish demand for antitrust, but should boost it. By economics textbook definitions, regulation and antitrust are complements--not substitutes, as implicitly and erroneously held in Trinko.
Regulation boosts demand for antitrust because in limiting the ability to exercise market power directly through setting the price that the market will bear--activity which is perfectly legal under the antitrust laws--it creates incentives to enter and to suppress competition in regulated markets that would not otherwise exist. Absent regulation, a firm with a monopoly prefers competition in vertically related markets, since lowering price in those markets boosts demand and thus profits for the monopolist's service. Hence, practices that nominally limit competition in those markets--be it vertical integration, tying, or exclusive franchising--would, to a first approximation, be entered into by a monopolist if they would enhance profits through efficiency, since the monopolist already has the ability to extract market power. (31)
But when the ability to exercise monopoly power is eliminated by regulation, a firm has an incentive to get around that restraint. In the telecommunications context, it could enter an unregulated market for a service (e.g., long distance) that requires access to its regulated service (e.g., the local network), and deny or give lower-quality access to its competitors. This reduces or eliminates competition in the unregulated market, allowing it to raise the price. The source of the profit is control over access to the regulated product combined with the regulation. Absent the regulation, the firm would lack this incentive to discriminate; it would make more money charging a high price for access to competitors in the unregulated market. This concern is not unique to telecommunications; it stands behind policies to limit the degree to which electricity-generation companies, in a competitive sector, can control operations of the monopoly transmission and distribution facilities needed to deliver energy to consumers. (32)
As noted above, a second concern is what in telecommunications is called "cross-subsidization." That practice entails charging costs of labor, equipment, or services used to provide unregulated services to the accounts of a regulated service. If the rates for that regulated service are based on costs, and if regulators are not able to determine how that labor, equipment, or other services are used, charging these costs to the regulated side of the business raises the rates. On paper, the regulated side of the firm earns only the regulated profit rate. The profits from the higher price are realized from sales of the unregulated service where the costs have been shifted.
Not only can cross-subsidization essentially force ratepayers to pay rates above what regulators intend, the ability to shift costs may deter entrants into the unregulated markets, who may find it unprofitable to compete against a rival who can charge costs to regulated firms. This predatory threat is not credible without the regulation that allows a firm to raise prices through cross-subsidization; absent regulation, the firm cannot shift the cost of those subsidies.
The shift in attitudes from AT&T to Trinko reflects a declining influence of this view that regulation and antitrust are complements, not substitutes. This has not only affected antitrust in telecommunications in the quarter century since the AT&T divestiture. In January 1982, during the press conference at which Assistant Attorney General William Baxter announced the settlement of the AT&T case through divestiture, he also said that the other longstanding antitrust case against IBM "'is without merit and should be dismissed.'" (33) The difference between IBM and AT&T was that the latter was regulated and the former was not. In contrast, nearly twenty years later, in announcing a similar proposed divestiture remedy in the government's case against Microsoft, Assistant Attorney General Joel Klein said:
Klein erred in analogizing Microsoft in 2000 to AT&T in 1982 in that he failed to appreciate that AT&T was regulated and Microsoft was not. In the end, and not surprisingly, the Microsoft divestiture was rejected by the D.C. Court of Appeals. (35)
IV. "ESSENTIAL FACILITIES" AND TRINKO: TWO SIDES OF THE SAME COIN
One of the hallmarks of Trinko was its rejection of the EF doctrine. An almost surely unintended irony went unnoticed, at least by the Court. Another hallmark of Trinko is that the same entity--in this case a sector-specific regulator--should both set prices when competition does not work, and also apply structural and behavioral remedies to the sector to protect competition where it could work. The EF doctrine does the same, substituting only antitrust enforcers and courts for regulatory agencies as the institutions charged with carrying out both tasks. If the basis for opposing the EF doctrine is that antitrust agencies are poor price regulators--a position with which I agree, for reasons discussed in the final section of this Article--one should also be skeptical of the similar cohabitation of antitrust and regulation envisioned by Trinko.
The Trinko Court dismissed the EF doctrine on four grounds: (1) the Court has never recognized such a doctrine in the past; (2) unavailability of access (and presumably not just a high price) is essential; (3) the EF doctrine does not apply when "a state or federal agency has effective power to compel sharing and to regulate its scope and terms"; and (4) the EF argument was not distinct, at least in this case, from a "general [section] 2 argument." (36) Outside the regulated context, this is a sound finding. Without the court getting involved in setting prices, ordering access to an essential facility will not improve economic performance or competition. If the owner of an essential facility can charge the monopoly price for access, merely ordering that access be made available does nothing fundamental about market power. In addition, such an order may forego efficiencies from vertical integration if the owner of the facility restricts access to itself. Only in the presence of regulation, as seen above, where the price is already set below the monopoly level, could using antitrust to ensure access improve market performance. It could do so through a divestiture or other separation that would thwart a regulated firm's incentives to evade price controls and exercise market power by favoring vertically related affiliates through discrimination or cross-subsidization. (37)
The role of regulation is important for the EF doctrine in terms of crafting remedies. It turns out to be crucial, at least hypothetically, in deciding how one would determine if a facility is "essential." For antitrust purposes, the question is whether a firm's ownership of a facility in question, such as a network of telephone lines to customers' premises, gives it market power over a good or service, such as providing telephone service. Defining when a firm has market power is a long-standing conundrum. (38) The primary problem is that a firm with market power will raise prices up to the point where buyers start to view others' products as substitutes, so the presence of substitutes is consistent with market power rather than an indicator of competition. (39) High profits or prices fail because they can also result from unanticipated increases in demand or rents accruing during peak demand periods. (40) The methods used to identify markets in merger cases, based on whether consumers would turn to substitutes if a set of firms were to institute a small but significant, non-transitory increase in price (SSNIP), are appropriate for seeing if a merger would increase prices above current levels, but for the reasons listed here, cannot tell us if an individual firm possesses market power. (41)
In light of these difficulties, one has to go back to first principles. The behavioral distinction between a firm with market power and a firm lacking it is that the former acts as if it can increase price by reducing output, while the latter takes price as a given. This suggests that to ascertain whether a firm has market power, we ask the following question: what would a firm do facing a small but significant, non-transitory reduction in price? (42) If a firm has no market power, it would likely reduce output; at most, it would keep output constant. Taking prices as a given, it would be producing up to the point where price equals marginal cost. At lower prices, a firm lacking significant market power would thus reduce production, unless it happens to be operating at full capacity, in which case it would still find it profitable to produce to its limit. On the other hand, if a firm has market power, it would hold output down in order to keep prices up. Were it to face a ceiling on the price it could charge below the price it is charging it would increase output, as it no longer has anything to gain by holding back supplies.
The good news is that we have a theoretically sound test for whether an individual firm has market power and, specifically, whether its facilities are "essential." Instead of hypothesizing whether a theoretical cartel could impose a SSNIP, as in the Merger Guidelines, (43) one posits what would happen if a hypothetical regulator were to impose a small but significant, non-transitory reduction in price. If output goes up, the firm has market power; if output does not go up, it does not have market power. The bad news is that the validity of the test is matched by its impracticality. One is extremely unlikely to have any sort of natural experiment in which a price ceiling was imposed, where one could observe whether or not the firm reduced output. More to the point, rather than identifying essential facilities to see whether regulation is justified, the test looks to see whether regulation is justified (by increasing output) to identify essential facilities--putting the policy cart before the theoretical horse, as it were.
FEED