An analytic history of delivered price litigation: do courts properly distinguish rivalrous from collusive instances?

Cabou, Christian G. ; Haddock, David D. ; Thorne, Michele H. 等

The federal judiciary's present, cautious approach to delivered pricing is appropriate. Using court records and recent contributions to the economic literature, this article evaluates the antitrust history of those practices. Delivered pricing can replace mill pricing naturally in some rivalrous settings, but under special circumstances could be collusive. Thus, delivered pricing has sometimes been adjudged a Sherman Act violation, but only given other evidence of collusion. For a time, however, the legal system threatened to treat delivered pricing as a per se violation of the Clayton and Federal Trade Commission Acts. That reflected poor economic understanding.

Delivered pricing litigation seems to be a pure statutory phenomena; our investigation has unearthed no cases under the common law of restraint of trade.[1] Since the advent of federal antitrust legislation, however, plaintiffs have relied on several distinct acts of Congress to bring delivered pricing practices before the courts.

The earliest of those statutes, the Sherman Act of 1890, forbad conspiracies in restraint of trade. But courts typically have required explicit evidence before they will find a collusive conspiracy. The mere use of a novel and poorly understood delivered pricing technique has not been enough.

Relying on the Clayton and Federal Trade Commission Acts of 1914, however, lawyers formulated various claims that purported to draw an inference of collusion from the mere use of a particular delivered pricing schedule. Under those arguments, no evidence of collusion was to be required. Several important precedents were set in reliance on such claims. That, we argue below, is a leap of logic inconsistent with present economic understanding. Because the courts sometimes have accepted such "theories," ,but at other times have not, the predictability of the law has been eroded, to the cost of defendants and, ultimately, to the cost of consumers.

I. THE DAWN OF DELIVERED PRICING AS AN

ANTITRUST PROBLEM

The History

A number of industrial practices that became evident during the decades following the Civil War were soon followed by statutory antitrust law, which largely supplanted the common law antitrust precedents that had developed over the preceding centuries. The speed and distance over which commercial orders could conveniently be placed and deliveries made steadily increased as local and regional transportation and communication networks expanded to form national systems. That evolution enabled aggressive entrepreneurs to expand local markets into regional ones, regional markets into national ones and national markets into world-wide markets.[2]

In industries where technical scale economies were important, the market expansion ultimately led to large-scale production centers arising at particularly favorable transport hubs, resource nodes, or buyer concentrations. Concomitantly, firms at less favorable sites either failed in the face of the stiffening long-distance competition, or found ways to thread themselves into small niches that had been left poorly served by the large-scale production centers.

In the absence of significant differentiation of the physical product, modern economic theory implies that it would be difficult for more than one firm to survive in a single small market away from the advantaged production centers in a predominantly large-scale industry.[3] So, with no local competitors, such "remote" enterprises as did survive would be able to price discriminate among nearby customers. In that sense the surviving small, remote firms would resemble the norm from an earlier day, when markets were more commonly characterized by small, dispersed, spatial monopolies.

But the manufacturing costs of firms in the new, highly competitive production centers, coupled with the transport costs required for them to reach various points in a market, determined the delivered prices that were available from the production center. For remote firms operating at substantial production-cost disadvantages, such major-market delivered prices would create a cap on the prices that could be charged. So the remote firms' ability to price discriminate need not generate revenues sufficient to provide an incentive for competing firms to enter the neighborhood. Indeed, such discrimination may be essential for the survival of the remaining remote, high-cost firms; monopoly does not guarantee profit.

Thus, for industries where transport cost was especially important, the viable forms of price discrimination began to alter.[4] Regardless of variations in the underlying elasticities of buyers' demands, the geographical pricing structure was increasingly driven by the delivered prices from major multi-firm production centers. Whatever the location of the remote firm, it could not sensibly ask as high a price of customers who were relatively near a major production center as was asked from buyers who were relatively far from the center.

Buyers, of course, typically perceive only a tiny part of the data that equilibrates a market--usually a price quotation at home, perhaps a much lower price quotation to a brother or cousin an equal distance from the production center, though in another direction, and often nothing about quotations from unsuccessful offerors in the other locale. To many of the buyers against whom the remote firm discriminated, therefore, the newly evolving discriminatory price schedules would have seemed arbitrary and maddening.

Although dissatisfaction with the development of the increasingly subtle forms of spatial price discrimination was perhaps not central to passage of the Sherman Act, buyers eventually tried to use that statute to attack the discrimination, as we discuss below. The success of early Sherman Act attacks on the apparent increasing industrial concentration was limited, contributing to passage of the Clayton Act, which was directly aimed at price discrimination, and the Federal Trade Commission (FrC) Act, which was aimed at that practice among many others.[5] The Law

The Sherman Act outlaws collusion and empowers private plaintiffs to sue for damages. Because many agreements among firms constrain their behavior in economically beneficial ways--indeed, it would scarcely be possible to conduct business without a substantial number of negotiated constraints--courts soon formulated a "rule of reason" to help mark the boundaries between permissible voluntary restraints on business activity and impermissible collusion. But even so, the Sherman Act purports to limit behavior that many (though hardly all) economists find objectionable.

Nearly a quarter of a century after the passage of the Sherman Act, Congress passed both the Clayton and the FTC Acts. The Clayton Act, which dealt directly with price discrimination, shared with the Sherman Act the notion that litigation would take the form of private actions seeking damages for prior behavior. But unlike the Sherman Act, the Clayton Act outlawed behavior that many economists consider benign or even desirable. Price discrimination can either increase or decrease economic surplus.[6] But the statutory language and much of the judicial treatment of defendants' answers to Clayton Act charges seem unrelated to factors bearing on the desirability of the discrimination.

The FTC Act, for its part, differs from the Sherman Act more broadly still. It specifically substituted a publicly financed agency, the Federal Trade Commission, for the private plaintiffs envisioned by the Sherman and Clayton Acts. Moreover, the FTC Act encumbers the Commission with numerous vague and open-ended instructions to which its employees supposedly are to refer when policing the economy. In consequence, many observers argue that the Act does not sufficiently focus Commission attention on activities that have economically deleterious consequences.

The Phenomenon

A mill price is the price of a unit of output that is picked up at the seller's place of business. In contrast, a delivered price applies to output that is shipped by the seller to the buyer. After a slow beginning, delivered pricing antitrust cases have become common, utilizing all three of the major antitrust statutes, and initiated by both private plaintiffs and the FTC.

The most frequently challenged and most avidly debated delivered pricing system is basing-point pricing. A firm that uses basing-point prices offers to deliver the product to consumers' locations according to a price schedule that would be c.i.f. (mill "cost" (price) plus insurance plus freight) if the firm's plant were located somewhere else (the "base").[7] Perhaps the most well-known example in the United States is the Pittsburgh Plus steel pricing from the early part of this century. The delivered prices of steel that was produced in Pittsburgh invariably rose with increased distance from that city. But the delivered prices of steel produced at other locations, such as Chicago, Birmingham, or Pueblo, fell with increased distance from the source if the deliveries were made to points toward Pittsburgh. Thus, a Pueblo producer quoted delivered prices that seemed to be c.i.f. Pittsburgh, so that Pittsburgh was the base of the system. For decades, nearly all economists believed that basing-point prices indicated collusion.8 Many still do.[9]

That consensus has recently been challenged, however.[10] Indeed, Carlton [1983] argues that basing-point prices not only would fail to maximize cartel profits, their use would actually increase the danger of chiseling, and thus of cartel failure. As it happens, basing-point prices can arise in either rivalrous or collusive situations, but only in a rather special environment would it seem to be in a cartel's interest to adopt them. And it is even possible for basing-point prosecutions to be procollusive.[11] Those arguments are reviewed in the following section.

Whatever the Congressional expectation may have been, judicial interpretation of the Sherman Act has been strict where basing-point pricing is concerned. Collusion is outlawed by the Act, but judicial interpretations have established that collusion cannot be inferred merely because an industry is concentrated, or because business practices seem unusual in comparison with other industries where scale economies and transport costs are less substantial. Evidence of collusion will be needed before courts invalidate the practice-under the Sherman Act.

But both the Clayton and the FTC Acts greatly relaxed basing-point plaintiffs' burdens of proof. The Clayton Act makes price discrimination illegal, with strictly limited exceptions, and does not require collusion. Various delivered pricing systems-basing-point included--are indubitably discriminatory. And the FTC Act empowers the agency to attack "incipient Sherman Act violations"--practices that have arisen naturally, but which might potentially be anticompetitive, and anticompetitive in some way that need not even be specified very carefully.[12]

In consequence, plaintiffs (often the FTC) ultimately won a number of important basing-point cases that could not have been won under the Sherman Act. The FTC's high-water mark was reached in 1948 with Triangle Conduit, where the Court held that the Commission need not show agreement in order to find a delivered pricing system illegal. Although, due to Congressional pressure, the FTC failed for some years to exploit its Triangle Conduit victory, the Commission has recently resumed an aggressive posture toward delivered pricing, as illustrated by Boise Cascade [1980], Ethyl Corp. [1983], and E.L du Pont de Nemours & Co. v. Federal Trade Comm'n. [1984].

II. WHY DO FIRMS USE BASING-POINT

PRICES?

Most people naively assume that mill pricing is somehow "natural." Or, if for some reason buyers prefer that sellers arrange delivery, perhaps because the sellers receive lower bulk transport rates, then c.i.f. prices must be the "natural" alternative. From that perspective, any discriminatory delivered pricing structure seems suspect, implying, perhaps, some abusive intent.

In fact, however, when local markets are too small to support numerous sellers, spatially discriminatory prices are in some ways more natural than mill or c.i.f. prices, as was understood by spatial economists at least as early as Hoover [1937] and Smithies [1941]. In such environments, firms are apt to discriminate spatially unless it is too costly to estimate demand elasticities in local markets, to deal directly with the commercial carriers, or to prevent shipments from reaching destinations unwanted by the seller. Nevertheless, basing-point pricing often arises in industries with numerous producers at some sites. And that feature continued to puzzle even careful observers, raising anew a suspicion of collusion.

This section will briefly review the rudimentary theory of spatial discrimination. We then review the recent literature that describes how basing-point pricing can arise in a rivalrous environment. The key to the latter is recognition that firms at multi-plant sites do not, in fact, discriminate if those sites are a basing-point system's bases and the firms do not regularly ship out of their "natural markets." The section then reviews a recent argument defining a weakly collusive arrangement under which a geographically isolated subset of firms will find it profitable to collude to adopt basing-point pricing. But, alternatively, basing-point pricing may represent a challenge to a cartel struggling to prevent chiseling. And successful attacks on basing-point pricing may actually increase the likelihood of cartel-like behavior. The section closes with that possibility.

Optimal Noncollusive Spatial Discrimination

Noting that for delivered products marginal cost varies precisely with transportation costs, but marginal revenue ordinarily does not vary precisely with demand, Smithies [1941] showed that transportation costs usually will not vary precisely with profit-maximizing prices for a spatial monopolist. Assuming that firm A owns plant 1 at site I, the line ab in Figure 1.b shows how that plant's marginal cost of producing and delivering a unit of output varies as it ships the product over space. That line slopes upward, because higher transport charges must be incurred to ship units greater distances. Figure 1.a shows the demand curve at each point in the market. Although both figures show delivered price on the vertical axis, the horizontal axis of Figure 1.a shows quantity demanded at a location, whereas Figure 1.b shows the distance of various locations from the production site. Equating marginal cost to marginal revenue at each location in the market implies the spatial pattern of prices shown by the line bc. In the figures the slopes of the two lines differ, implying that maximization of profits requires the firm to absorb some of the freight charges as it ships ever farther afield.[13]

In important ways, firm behavior at many selling points will be similar if firm B now locates plant 2 at site II. In some geographical area (those points to the left of location L) the profit-maximizing price for firm A is below the marginal cost of producing and delivering the product from site II, and so the new plant is not a factor. But either firm could potentially sell at any location between L and L', because each one's marginal cost of producing and delivering to those points is less than the ideal delivered price for the other firm. As Greenhut and Greenhut [1975] note, it is unclear what prices will prevail at such locations, or how the sales at each point will be shared. That will depend on how each firm reacts to the other's presence, in particular, on the conjectures that each firm makes regarding the responses of the other to any initiative in the region between L and L'. It is clear by inspection of the figure that the prevailing delivered price at some locations in the region L to L' may fall substantially below the ideal delivered price from the viewpoint of either firm. Similarly, the delivered prices may lie substantially above the marginal cost from either plant.

Under plausible assumptions, if an increasing number of plants begin to operate at site I, each will perceive an increasing demand elasticity at each location, and consequently the delivered price will decrease toward the site ! plants' marginal cost of production plus delivery. Viewed over space, the slope of the delivered price schedule steepens, as in Figure 2. There had been greater markups of price over marginal cost near site I than far from it, but the markups are dissipated by the increasingly intense competition. The left border of the region that is shared with the site II plant moves from L to L" Moreover, the delivered price that prevails in the L' to L" region cannot deviate by much from the marginal delivered cost and ideal delivered price from site I, because those two schedules are converging with the increase in the number of site I plants.

Spatial price discrimination, then, will be normal if (1) the demands at different selling points can be separated by a firm, (2) the number of competitors selling at some locations is limited, and (3) the cost that the firm bears to fine-tune its spatial price schedule and deal with delivery is less than the added revenue the firm receives by discriminating.

So spatial discrimination per se need have nothing to do with collusion. But can the particular form of price discrimination that is implied by basing-point pricing arise without collusion? Until recently, most economists thought not.

Rivalrous basing-point prices are indeed plausible. Given enough competitors at site I, mill-net prices there fall to marginal cost of production. Delivered prices from that site then become c.i.f., or else the firms at site I dispense with a delivered pricing system altogether, selling for a mill price at the plant door. But if site II continues as a local monopoly, its optimal pricing schedule in any region that is disputed with the other site may be basing-point, given the proper conditions.[14] Indeed, that may happen even if more than one producer begins to operate at site II.[15]

The demand curve at any point in the market will now be perceived by the site II firm to be horizontal at the price for which deliveries can be had from site I plants, PIg' At prices below that, however, the site II firm perceives that location's entire demand curve. That generates a kink in the demand curve that faces the site II producer for deliveries to that location, and so the marginal revenue curve has a gap, as in Figure 3. If the marginal cost (including freight charges) for the site II producer, [mc.sub.II], falls anywhere within the gap in marginal revenue, the profit-maximizing price for that firm to charge at that location will be [P.sub.Ia] Hence, without any taint of collusion, the site II firm adopts a price that is based on the delivered price that site I producers offer at that location.

More surprisingly, in view of the tenacity of the collusive beliefs spawned by the earlier literature, unless there are special considerations within the industry, basing-point pricing will not arise with collusion. A strong cartel maximizes cartel profits, meaning that its behavior resembles a multi-plant monopolist's, except (possibly) for a lesser ability to optimize capacity. But if that is true, Smithies's [1941] model applies; profit-maximizing prices will nearly always vary in a different pattern than transport cost does. Hence, for basing-point pricing to prevail in an industry, it must be rivalrous or it must arise from a cartel that is disabled in some way.

Weak Coordination and Basing-Point Pricing

One way for a cartel to be disabled is for its influence to be geographically limited-i.e., one that cannot achieve agreement with, detect chiseling by, and/or punish violations of the cartel arrangement by firms that are at a distance. Functionally, such a cartel resembles a spatial monopolist and so resembles the remote firm of the rivalrous basing-point model. Maximizing cartel profits may then require basing cartel members' prices on those of nonmembers elsewhere and then absorbing freight.[16]

In contrast, suppose a cartel is marketwide. Carlton [1983] argues that the cartel then will want to divide market areas among cartel members, not adopt basingpoint prices. The problem is not just that basing-point prices do not maximize cartel profits; they actually enhance opportunities to chisel.

A cartel must first detect chiseling to control it. When prices are unobservable, Stigler [1964] shows that a second-best means to detect chiseling is to observe sales patterns. Changes that could not plausibly reflect stochastic factors will signal chiseling. But if they are strictly adhered to by sellers, several common delivered price systems, basing-point prices included, make price irrelevant for buyers' choices among sellers. Consequently, when the cartel members agree to charge the same prices to buyers at a number of locations, as they would if they colluded to adopt basing-point pricing, then a shift of customers from one seller to the other provides relatively little information about chiseling. The shift could arise from any number of exogenous events that have resulted in even a modest alteration in buyer preferences.

In contrast, market-dividing price schedules, such as mill pricing, yield similar indeterminacy only along market boundaries between neighboring production sites. Sales patterns away from a boundary can thus be used to help infer whether or not chiseling is occurring. Thus, even if all colluders at a single site quote identical prices to every customer because no other way to divide the local market has been found, they will retain market boundaries with neighboring cartel members. Market division would seem to be a substantially better cartel strategy than product homogenization.

Procollusive Challenges to Basing-Point

Pricing

Because basing-point prices will often be contrary to collusive interests, the destruction of renegade firms' basing-point schedules can aid a cartel. Assume that a cartel tries to institute some set of marketdividing prices--for simplicity suppose they are mill prices. Since a successful cartel's price quotations exceed marginal cost, a firm at one site can profitably invade at least the margins of another site's "natural market" at those prices. Such an invasion may reflect chiseling on the cartel, or simply opportunism by a non-member. The chiseling/non-member firm would, in effect, be creating a market-intermingling multi-base basing-point system out of the cartel's preferred market-dividing mill pricing system. But in that event the cartel itself would benefit from a successful challenge to the multibase system, for the chiseling/non-member firm will then be compelled to limit its anti-cartel behavior to the area of its own natural market.

A real-world "experiment" provided an informal test of the argument. Many firms interpreted a 1948 decision against cement producers as outlawing discriminatory delivered pricing schedules generally. As a result, a mill pricing structure in the steel industry rapidly evolved, instigated by U. S. Steel, the industry's largest producer.[17] Thus, a previously intermingled steel market became one characterized by geographically localized oligopolies and monopolies. According to the Carlton model, price coordination should have improved. Subsequently, it was reported that steel prices increased by an average of $9 per ton following the switch to mill pricing.[18] Indeed, many buyers of steel petitioned Congress for legislation that would make it clearly legal for steel companies to resume quoting delivered prices.[19]

III. A BRIEF HISTORY OF DELIVERED

PRICING LITIGATION

In the early delivered pricing cases, courts found that there was violation of neither the Sherman Act nor the antecedent common law if no agreement was found among the accused firms, either directly or by circumstantial evidence. For instance, in Cement Mfrs. Protective Ass'n. [1925] and Maple Flooring [1925], the FTC charged manufacturers with Sherman Act violations. In both cases, however, the Supreme Court held for the defendants because it found no agreement. But when the Court accepted evidence of overt collusion to maintain prices, as in Sugar Institute [1936], it held the activities to be illegal restraints of trade.[20]

The Evolution of the Theory of Conscious Parallelism

But the Clayton Act and the FTC Act gave government enforcers much more discretion in applying antitrust laws to delivered pricing. In 1945 the FTC set an important precedent in separate Clayton Act suits against two glucose manufacturers-Corn Products Refining Company, and Staley Manufacturing Company--with no allegation of collusion being raised. It is evident that several delivered pricing systems are discriminatory, but it had not been obvious whether such discrimination was defensible under the Act's "good faith" exception.[21] The Supreme Court held that it was not.

Corn Products [1945] involved two plants of a single manufacturer. One plant was in Chicago, the major production center of the industry, the other in Kansas City. The manufacturer used a mill pricing schedule for shipments from Chicago, as did its Chicago competitors. Actual deliveries, however, might originate from either the Chicago or the Kansas City plant, depending on which had the less severe capacity constraints at the moment. The Court treated the Kansas City plant as though it were making price reductions on shipments toward Chicago in order to compete with its sister plant. That did not seem to the court to meet the good faith exception, since both plants billed against Corn Products quotations. So the Court held such prices to be illegal when the shipments originated in Kansas City.

Staley, a down-state Illinois firm, raised a meeting-the-competition defense in a separate case that was decided on the same day as Corn Products. In a confused opinion, the Court treated the Staley basing-point schedule as an imitation of the pricing schedule of Corn Products's Kansas City plant. The court ruled that a good faith defense under the Clayton Act would not be found when one firm, such as Staley, adopted in total an illegal pricing schedule, such as that of Corn Products's Kansas City plant. The Court's position is perplexing in light of modern understandings of basing-point pricing. The rivalrous basing-point model implies that Staley had not intentionally adopted the pricing schedule of the Kansas City plant at all. Staley's prices competed with the mill prices of the major producers at Chicago, and mill prices are clearly permissible. Nor, for that matter, would Corn Products have wished to base its Kansas City plant's prices on those of its sister plant. Rather, both of its plants were forced to compete with competitors' Chicago prices, or forego profits.

The FFC had argued that, by imposing "artificially" high costs on buyers, the price discrimination that flowed automatically from basing-point prices could potentially have an anticompetitive effect on the candy manufacturing industry, the main destination of glucose shipments. What the Court did not recognize was that such discrimination can be procompetitive. For example, plants in some inhospitable 1ocations are unviable unless they price discriminate. If such plants discount their prices below precise basing-point prices, or if they offer other advantages to attract nearby buyers,[22] their customers, even those "against" whom the firm discriminates, are benefited; the customers are better off than if the plant did not operate at all.[23]

Emboldened by its success in the glucose cases, in 1948 the FTC won an action against a cement manufacturers' trade organization, the Cement Institute, whose members used a multiple-base system. The Court remarked that, although there was evidence of overt agreement in that case, the FTC might not need to show agreement to win a delivered price action under the FYC Act. That speculation soon became law in the pivotal Triangle Conduit [1948] decision. To decide that case against the defendants, the Court need not have even considered the legality of spontaneously arising delivered pricing systems. The Triangle Conduit defendants had sometimes submitted identical bids for government contracts, and the court could have focused directly on that practice.[24]

Nearly immediately, several bills were introduced in Congress to reverse the Triangle Conduit holding and declare delivered pricing systems legal, absent collusion.[25] Though none of the bills were adopted, the FTC reacted promptly to the Congressional prodding.[26] The Commission announced that it would henceforth require evidence of at least tacit agreement before lodging "conscious parallelism" charges, such as the one used successfully to attack Triangle Conduit. The FTC's position regarding conscious parallelism then stabilized for decades.

New Targets and Renewed FTC Threats

During the last quarter of the present century, however, the FTC has revived its attack on delivered pricing systems (Boise Cascade 1980; Ethyl Corp. 1983; du Pont 1984). Contrary to the position it had articulated over the previous quarter century, the FTC once again is bringing actions against firms using delivered pricing schedules.

The FTC's previous clear position regarding the legality of delivered pricing, and the Congressional debates that preceded the Commission's adoption of that position, were noteworthy to the court that heard Boise Cascade [1980]. The court noted that in 1949 a majority of Senate subcommittee members investigating the Commission's pricing policies had found that the rationale of Cement Institute and Triangle Conduit applied only to situations where conspiracy was present. A subcommittee report had concluded that "the Commission appears to have written off the theory that 'conscious parallel action,' absent conspiracy, constitutes an unfair method of competition under the [FTC Act]" (Boise Cascade 1980, 576).

The Boise court found no evidence of overt agreement. Indeed, the record revealed active price negotiations between buyers and sellers, generating prices that consistently were discounted rather than being precise basing-point prices. The Court held that industry-wide adoption of an "artificial" price system is not a per se violation, and that mere widespread use of a pricing practice says little about guilt or innocence.

Under essentially the same set of facts, however, a special jury verdict in the class action suit In re Plywood [1981] found a violation of the Sherman Act. That case was subsequently settled while a Supreme Court decision on appeal was pending.[27] Rather weak evidence had been presented at trial that lower-level field personnel of two of the defendant southeastern plywood producers had discussed pricing trends in the industry. A "bright-line" rule against discussion of prices by competitors may well be advisable. Such discussions may be perfectly innocent, but they are required for a collusive agreement, and are cheaper to prove. Hence, if such discussions have no benign explanation (a rebuttable presumption), economizing behavior by the judiciary would result in their being banned outright.[28]

In du Pont [1984], the Commission recently unsuccessfully attacked another delivered pricing practice that apparently had arisen without agreement among industry members. There were three challenged facilitating practices: (1) thirty-day advance notice of price changes during which period notified customers could buy at the lower price level;[29] (2) a mostfavored nation clause that made available to certain buyers post-contractual price decreases;30 and (3) a delivered pricing system. Although the FTC charged that the business practices facilitated price parallelism, there was no evidence indicating an attempt to collude. The court held that, without evidence of at least tacit agreement, the Commission must show an anticompetitive intent, or a lack of any legitimate business reasons for the challenged practices. The court found that the FTC had not met that burden.

Hence, subject to the position the FTC had adopted in reaction to Congressional pressure in the late 1940s, the courts have held that evidence of agreement, tacit or overt, will be required in conscious parallelism cases.

IV. CONCLUSION

In this article we examined the impact of the three major antitrust statutes on delivered pricing litigation. Given the records that have survived, we cannot conclude that the Sherman Act standing alone has been seriously misused with respect to the delivered pricing cases.[31] Admittedly, many observers have been critical of the course of statutory American antitrust law. But if the set of delivered pricing cases are representative, the problem seems to arise not from the Sherman Act, but from later statutes, the Clayton and FTC Acts in particular.

The Clayton Act claims, among other objectives, to halt "bad" price discrimination while preserving the "good" variety. But Congress did not give (and likely could not have given) much useful guidance in discerning one from the other. In consequence, the courts have been forced to find their own course, and they seem to have chosen an unfortunate, muddled route if decisions like Staley [1945] are typical. But at least the Clayton Act per se is passive--courts are not empowered to search for suspected violations and then assume an advocate's role at public expense. Instead, private parties who believe they have been injured must, at their own expense, approach the courts for redress. That automatically creates a connection, however tenuous, between the expected costs and expected benefits of antitrust litigation.

The FTC Act is more pernicious. In addition to being phrased in language of disturbing vagueness, the FTC Act created a publicly financed advocate that is empowered to look for injuries where even the supposed victims may not perceive them. It is illustrative that some defense witnesses in Boise Cascade were customers of the defendant firms, just as some earlier witnesses asking Congress to legalize delivered price systems were steel customers. That power is a dangerous power in an agency; if nobody is behaving illegally, and the FTC admits as much, the agency's budget will be reduced. Such an expectation inspires the FTC to find work to do whether or not there is work to be done. As a consequence, the FTC has innovated a stream of new "theories" of illegality to which defendants have been forced to respond, at substantial cost to the economy. One such claim involves "conscious parallelism." After Triangle Conduit [1948], in fact, it seemed that the courts might accept the mere existence of a discriminatory delivered pricing system as conclusive evidence of illegality. That path has not been taken to this moment, though the FTC has urged it anew.

In many different delivered pricing cases the courts seem to have been seriously confused, struggling to find a logical path through a theoretical quagmire. And well they might have been. Quite apart from the flabby statutory language which the courts have confronted, our present understanding of the economics of delivered pricing is almost entirely the product of the past half-century, with a significant part the product of just the past decade. It is unreasonable to criticize earlier courts merely because the judges were ignorant of matters that nobody else understood at that time. Today, however, one should do better; economists now know a great deal more about delivered pricing than they knew when most of the important precedents were set. It would be shameful if that knowledge did not inform future decisions.

REFERENCES

Averitt, Neil W. 'The Meaning of 'Unfair Methods of Competition' in Section 5 of the Federal Trade Commission Act." Boston College Lain Review, January 1980, 227-300.

Barzel, Yoram. 'Measurement Cost and the Organization of Markets.' Journal of Law and Economics, April 1982, 27-48.

Benson, Bruce L., Melvin L. Greenhut, and George Norman. "On The Basing-Point System.' American Economic Review, June 1990, 584-88.

Burns, Arthur Robert. The Decline of Competition. New York: McGraw-Hill, 1936.

Carlton, Dennis W. "A Reexamination of Delivered Pricing Systems.' Journal of Law and Economics, April 1983, 51-70.

Chandler, Alfred D., Jr. Scale and Scope: The Dynamics of Industrial Capitalism. Cambridge: Belknap Press, 1990.

Clark, John Maurice. "imperfect Competition Theory and Basing-Point Problems.' American Economic Review, June 1943, 283-300.

Daugherty, Carroll R., Melvin G. de Chazeau, and Samuel S. Stratton. The Economics of the Iron and Steel Industry. New York: McGraw-Hill, 1937.

DeCanio, Stephen A. "Delivered Pricing and Multiple Basing Point Equilibria: A Reevaluation.' Quarterly Journal of Economics, May 1984, 329-49.

Gilligan, Thomas W. 'imperfect Competition and Basing-Point Pricing: An Application to the Softwood Plywood Industry.' Working paper, Hoover Institution, 1990.

Greenhut, John, and Melvin L. Greenhut. "Spatial Price Discrimination, Competition and Locational Effects." Economica, November 1975, 40119.

Haddock, David D. "Basing-Point Pricing: Competitive vs. Collusive Theories.' American Economic Review, June 1982, 289-306.

"On the Basing-Point System: Comment.' American Economic Review, September 1990, 95762.

Hoover, Edgar M. "Spatial Price Discrimination.' Review of Economic Studies, October 1937, 182-91.

Karlson, Stephen H. "Competition and Cement Basing Points: F.O.B. Destination, Delivered from Where?" Journal of Regional Science, February 1990, 75-88.

Lancaster, Kelvin. Variety, Equity, and Effciency: Product Variety in an Industrial Society. New York: Columbia University Press, 1979.

Latham, Earl. "The Politics of Basing Point Legislation." Law and Contemporary Problems, Spring 1950, 272-310.

Machlup, Fritz. The Basing-Point System: An Economic Analysis of a Controversial Pricing Practice. Philadelphia: Blakiston, 1949.

McGee, John S. "Cross Hauling--A Symptom of Incomplete Collusion Under Basing-Point Systems." Southern Economic Journal, April 1954, 369-79.

Mund, Vernon A. "The Development and Incidence of Delivered Pricing in American Industry." Law and Contemporary Problems, Spring 1950, 141-58.

Robinson, Joan. The Economics of Imperfect Competition. London: Macmillan, 1933.

Seager, Henry R., and Charles A. Gulick, Jr. Trust and Corporation Problems. New York: Harper & Brothers, 1929.

Smithies, Arthur. "Monopolistic Price Policy in a Spatial Market.' Econometrica, January 1941, 63-73.

"Aspects of the Basing-point System.' American Economic Review, December 1942, 705-26.

Stigler, George J. 'A Theory of Delivered Price Systems." American Economic Review, December 1949, 1143-59.

"A Theory of Oligopoly.' Journal of Political Economy, February 1964, 44-61.

Stocking, George W. "The Economics of Basing Point Pricing." Law and Contemporary Problems, Spring 1950, 159-80.

Thisse, Jacques-Francois, and Xavier Vives. "On the Strategic Choice of Spatial Price Policy." American Economic Review, March 1988, 122-37.

United States Congress. Hearings before the House Select Committee on Small Business. 81st Congress, 1st Session, 1949.

United States Senate. Report of Senate Committee on Interstate Commerce. 63rd Congress, 2nd Session, 1914.

Cases Cited

Boise Cascade Corp. v. Federal Trade Comm'n., 637 F.2d 573 (9th Cir. 1980).

Cement Mfrs. Protective Ass'n. v. United States, 268 U.S. 588 (1925).

Corn Products Refining Co. v. Federal Trade Comm' n., 324 U.S. 726 (1945).

E. L du Pont de Nemours & Co. v. Federal Trade Comm'n., 729 F.2d 128 (2d Cir. 1984), vacating Ethyl Corp. et al., 101 F.T.C. 425 (1983).

Federal Trade Comm'n. v. Cement Institute, 333 U.S. 683 (1948).

Federal Trade Comm'n. v. Gratz, 253 U.S. 421 (1920).

Federal Trade Comm' n.v. Motion Picture Advertising Service Co., 344 U.S. 392 (1953).

Federal Trade Cornm'n. v. Staley Mfg. Co., 324 U.S. 746 (1945).

Fort Howard Paper Co. v. Federal Trade Comm' n., 156 F.2d 899 (7th Cir.), cert. denied, 329 U.S. 795 (1946).

In re PlywoodAntitrust Litigation, 655 F.2d 627 (Sth Cir. 1981).

Maple Flooring Mfrs. Ass'n. v. United States, 268 U.S. 563 (1925).

Milk & Ice Cream Can Inst. v. Federal Trade Comm' n., 152 F.2d 478 (7th Cir 1946).

Sugar Institute, Inc. v. United States, 297 U.S. 553 (1936).

Triangle Conduit & Cable v. Federal Trade Comm'n., 168 F.2d 157 (7th Cir. 1948), aff'd. by an equally divided Court, 336 U.S. 956 (1949).