Total Repeal of Antitrust Legislation: A Critique of Bork, Brozen, and Posner Walter Block* T he premise underlying laissez-faire capitalism is that the only actions which should be illegal are those which involve an initiation of aggression against another person or his property. Antitrust law is clearly in violation of this prin- ciple, because it prohibits business practices no one even alleges constitute such depredations. The economists mentioned in the title of this paper are widely and properly celebrated for upholding the virtues of the free marketplace. However, there is one lacunae in their defense: antitrust legislation. Although they have done yeoman work in helping us to understand the beneficial effects of much com- mercial conduct which is prohibited by these enactments, their critique of this law is less than full. They each see a small but important role for the Antitrust Division of the Justice Depart- ment. They advocate reduction in the power and scope of this law, but not, unfortunately, total repeal. It is as if they are a football team which has succeeded in bringing the pigskin to the three yard line, but can make no further progress. This paper is an attempt to help them over the goal line. To continue our football analogy, the present paper will not comment on the 97 percent of their work which is responsible, *Walter Block is associate professor of economics at the College of the Holy Cross andexecutive editor of the Review ofAustrian Economics. The author wishes to thank two anonymous referees for helpful suggestions, many of which are incorporated here into this paper. The usual caveat of course applies. I have liberally footnoted this paper with the writings of Murray N. Rothbard on the topic of monopoly and antitrust, but these few citations are far from adequate to express the degree to which I rely on his pathbreaking work in this field. The Review ofAustrian Economics Vo1.8, No. l(1994):35-70 ISSN 0889-3047 36 The Review of Austrian Economics Vol. 8, No. 1 in large part, for the scholarly contribution to the cause of keeping antitrust law from being even more intrusive than it now is. In focusing on the 3 percent of disagreement, this paper may give the impression that there are large differences of perspective in public policy conclusions between these authors and their present critic. Nothing could be further from the truth. Robert Bork Merger Robert Bork (1978) maintains that some entrepreneurial choices in the market lead to efficiency, while others merely serve to restrict output. His main thesis is that antitrust has thus far insufficiently distinguished between these two situations. This is important, he contends, because if consumer welfare is to be enhanced, the restriction of output must be prohibited, while wealth enhancing activities must be promoted (or at least al- lowed.)' This can be shown by a consideration of Bork's "two vectors" hypothesis, representing, at least in the first instance, a merger. This is depicted in Figure 1. As Bork explains: The diagram assumes that the merger reduces the long-run aver- age costs of the two firms from ACi to AC2 but that the increased market power created by the merger results in a restriction of output so that the rate moves from Qi to Q2. We then see that consumers have lost outputfor which they would have been willing to pay an amount above cost equal to the area labeled Ai-and have gained in resource savings an amount equal to the area A2. Obviously, if Aa, the cost savings, is larger than A1, the dead-weight loss, the merger represents a net gain to all consum- ers. IfAi is larger than A2, a net loss results. This diagram can be used to illustrate all antitrust problems, since it shows the relationship of the only two factors involved, allocative inefficiency and productive efficiency. The existence of these two elements and their respective amounts are the real '~ork'sd iscussion of "allocative efficiency" depends intimately on the "bench- marks" provided by the purely competitive model. But the entire notion of "output restriction" depends logically upon some reference point: "restriction" relative to what?" I owe this point to an anonymous referee. Block: Total Repeal of Antitrust Legislation Figure 1 Source: Robert Bork, The Antitrust Paradox: A Policy at War with Itself (New York: Basic Books, 1978),p. 107. issues in every properly decided antitrust case. They are what we have to estimate-whether the case is about the dissolution of a monopolistic firm, a conglomerate merger, a requirements con- tract, or a price f ~ naggre ement . . . It must also be remembered that there need not always be a tradeoff [between A1 and Az]. In most cases, in my opinion, economic analysis will show that one of the areas does not exist, and a decision of the case is therefore easy. Some phenomena involve only a dead-weight loss and no, or insignificant, cost savings. That is the case with the garden-variety price-fixing ring. Output is restricted so that Q2 is to the left of Qi, creating the area Ai, but there is no downward shift of costs, no line AC2, and hence no area A2. (p. 108; material in brackets added by present author) One problem with the foregoing is that it pushes the courts into the role of determining whether or not any particular type of industrial organization or contract is or is not "cost saving." But the judiciary has no comparative advantage in making any such 38 The Review of Austrian Economics Vol. 8, No. 1 determinations2 Its members are not selected on the basis of being able to do so. Their salaries and promotions are not in any way tied to success in distinguishing efficient arrangements from inefficient ones. Failures are not punished with demotions. Achievement is not automatically rewarded with promotion, or with the awarding of bigger, more important or precedent setting cases. Why, then, should we expect this behavior from the courts? Indeed, if Bork himself is to be believed on this issue, jurists, all throughout the history of antitrust, have made findings which show them to be either unconcerned, or incompetent with regard to this issue. Says the author of this book: most of the mergers the Supreme Court strikes down and the "price discriminations" the Robinson-Patman Act is intended to stamp out. .. are examples .. .which involve only efficiency gain and no dead-weight loss. (pp. 108-9) A second difficulty has to do with the interpretation of the demand curve. In Bork's neoclassical construal, the demand curve is seen as an existing entity. True, this author concedes that "we do not know the location of any of the sides of the triangular area Ai,"(one of which is the demand curve), but this is only an inconvenience. "They are what we have to estimate" (p. 108) is the way to get around this annoyance. But this will not do.3 Demand curves are not "out there," ready to be measured by the modern econometric tools of analysis. Rather, they are, except for one dot (P2Q2, in this case), hypothetical alternatives which never come into play. Demand curves answer the question, Suppose that everything else in the universe were exactly the same as it now is, with the one exception that price, instead of being at Pq, is at some other level; then, how much would the customer be willing to buy at that other price. In the event, the price however, was P2 %his applies, in spades, to the administrative commissions which do the lion's share of antitrust work, such as the Federal Trade Commission. At least there is some check and balance on the former; they are subject to recall, and great scrutiny in the process of their initial appointments. These controls are greatly attenuated, if not virtually nonexistent, in the case of the civil service. I owe this point to an anonymous referee. 3~orkalso states: "Passably accurate measurement of the actual situation is not even a theoretical possibility; much less is there any hope of arriving at a correct estimate. . . . Nobody knows these curves. Even the companies involved do not. The clarity of the graphs . . . misleads many people" (p. 108).In short, Bork himself knows that the demand curves are not "out there." The real mystery is why Bork conveniently forgets his own excellent, and essentially Austrian criticism when he addresses himself to policy issues. (I owe this point to an anonymous referee.) Block: Total Repeal of Antitrust Legislation 39 and the consumer wished to purchase Q2. That is all we know, or indeed, can know. The other points on this demand curve never come into play at all. They are contrary to fact conditionals. There is no sense in the notion that we can "estimate" them. There is no doubt that economists can look at other instances (other times, places, people) where different quantities of this item were pur- chased at different prices (and even attempt to control for the fact that the prices and quantities of substitutes and complements have altered, to say nothing of changing incomes, inflation, em- ployment and even the weather) and in that way trace out a "demand But this has little or nothing to do with what is depicted in that diagram. The point is, a demand curve is a unique non-repeatable hypothetical "event." All attempts to "measure" it are thus doomed to failure. So far, we have been implicitly assuming that it is legitimate to make interpersonal comparisons of utility. It is now time to relax this assumption. In point of fact, this methodology is not tenable. It is perfectly reasonable to maintain that all trade benefits both parties in the ex ante sense. This is the reason they engage in such an activity, and this conclusion is part of the bedrock of the science of economics. It is quite another matter, however, to deduce from the failure of a trade to take place (in the free marketplace, Q1Q2 remains unsold, because it is not offered for sale) that had occured, the buyer's welfare would have exceeded the loss to the seller. This contrary to fact conditional implies that interpersonal comparisons of utility indeed can be made-without offering any evidence or reason for such an asser- tion-and moreover that the consumer's benefit exceeds the pro- ducer's loss. The latter contention would remain unproven even if interpersonal utility comparisons were valid in the first place. And yet, unless this assertion is true, the value of A1 would be negative, not positive as claimed by Bork. If a "garden-variety price-fixing ring"4 succeeds in raising prices from Pi to Pa, there is thus nothing within the strict science of economics that can be used to show that this will reduce social (as opposed to consumer) welfare. Still another fallacy of the two vectors approach lies behind the very drawing of the cost curves in this diagram, ACi and AC2. There is nothing untoward about using them for textbook illus- tration purposes only. Bork, however, is attempting to justify $he use of the "ring"i n this context is rather pejorative; it is akin to describing the ancient and honorable profession of price fixing along the lines of a car theft "ring." 40 The Review of Austrian Economics Vol. 8, No. 1 antitrust, a legislative enactment which can fine or even jail businessmen for the "crime" of price fixing, on the basis of this analysis. Under such circumstances it is reasonable to look more closely into these cost curves, an integral part of the analysis. Cost, in economic theory, is not by any means limited to out of pocket expenses, even including implicit rent. These are part of the concept, but in its most sophisticated interpretation, cost is equivalent to the next best opportunity foregone by making any particular choice. As such, cost can only be a subjective notion (Buchanan 1969; Buchanan and Thirlby 1981; Mises 1963). The next best opportunity foregone by the choice to sell Qi need not be anywhere close to Pi. In any case, it can never be known by a third party, for example by the Antitrust Division of the Justice Department, the government bureau charged with punishing or incarcerating price fixers. More radically, the cost of all saleable items is actually zero, and therefore can have no effect in any case. States Rothbard (1962, p. 604): there is no such thing as costs (apart from speculation on a .higher future price) once the stock has been produced. Costs take place along the path of decisions to produce-at each step along the way that investments (of money and effort) are made in factors. The allocations, the opportunities foregone, take place at each step as future production decisions must be taken and commitments made. Once the stock has been produced, however (and there is no expectation of a price rise), the sale is costless, since there are no advantages foregone by selling the product (costs in making the sale being here considered negligible for purposes of simplification). There- fore, the stock will tend to be sold at whatever price is obtain- able. There is no such thing, then as "selling below costs" on stock already produced. One can go even further. Not only is the sale costless when it occurs, it may even occur at less than zero costs. For example, if I have piled up a horde of tomatoes, or shoes, or steel, or tires, and I cannot find a customer for them, then, at least in a strictly private property rights-no trespassing world (Rothbard 1973, p. 1082), I will have to pay for their removal. Under such condi- tions the costs of the sale will be negative. That is, if the disposal costs are $500 (I have to pay $500 to rid myself of this unwelcome Block: Total Repeal of Antitrust Legislation 41 stock), then, ceteris paribus, I should be willing to sell it to a customer at any negative price above this level. For example, if I sell at $200, then I make a profit of $300. Even though I have to pay a customer $200 to cart away my merchandise, I am better off by $300 because the private sanitation hauler would have charged me $500. Sovereignty Yet another problem arises with regard to the issue of indi- vidual versus consumer sovereignty. Let us allow Bork to articu- late this thesis in his own words. As far as consumer welfare is concerned, he states it as follows: (antitrust) can only increase collective wealth by requiring that any lawful products, whether skis or snowmobiles, be produced and sold under conditions most favorable to consumers. (p. 91) Productive efficiency, like allocative efficiency, is a normative concept and is defined and measured in terms of consumer wel- fare. (pp. 104-5) But this rendition of the goal is problematic. Why should the goal of antitrust be to enhance consumer welfare alone? Why, for that matter, should the aim of any public policy be so narrowly defined? If it is taken for granted that some sort of welfare be maximized by legislation, why not attempt to maximize total welfare, that is, the welfare derived by both producer and con- sumer. It is possible to employ a reductio ad absurdum in this regard. If we really want to enhance the welfare of consumers only, as opposed to both consumers and producers, all sorts of other enactments become justifiable which would not have been oth- erwise. For example, if there were any producer's surplus (economic rents earned by manufacturers) then these should be summarily seized, and handed over to consumers. Needless to say, however, no warrant for any such action has ever been given. This policy, moreover, is internally inconsistent, for it will tend to counteract Bork's own goal of augmenting consumer welfare. We cannot safely ignore people as producers if we are attempting to maximize their well-being as consumers. People are people, and typically play a dual role as both consumers and producers. If we hurt them in one role, they are necessarily hurt in the other as well. 42 The Review of Austrian Economics Vol. 8, No. 1 Rothbard's remarks (1962, pp. 560-61) seem to be addressed directly to the Bork hypothesis, although they were written al- most two decades beforehand: We have seen that in the free market economy people will tend to produce those goods most demanded by the consumers. Some economists have termed this system "consumers' sov- ereignty." Yet there is no compulsion about this. The choice is purely an independent one by the producer; his depend- ence on the consumer is purely voluntary, the result of his own choice for the "maximization" of utility, and it is a choice that he is free to revoke at any time. We have stressed many times that the pursuit of monetary return (the consequence of consumer demand) is engaged in by each individual only to the extent that other things are equal. These other things are the individual producer's psychic valuations, and they may counteract mone- tary influences. An example is a laborer or other factor-owner engaged in a certain line of work at less monetary return than elsewhere. He does this because of his enjoyment of the particular line of work and product and/or his distaste for other alternatives. Rather than "consumers' sovereignty," it would be more accurate to state that in the free market there is sovereignty of the individual: the individual is sovereign over his own person and actions and over his own property. This may be termed individual self-sovereignty. To earn a monetary return, the individual producer must satisfy con- sumer demand, but the extent to which he obeys this ex- pected monetary return, and the extent to which he pursues other, nonmonetary factors, is entirely a matter of his own free choice. The term "consumers' sovereignty" is a typical example of the abuse, in economics, of a term ("sovereignty") appropriate only to the political realm and is thus an illustration of the dangers of the application of metaphors taken from other disciplines. "Sovereigntyn is the quality of ultimate political power; it is the power resting on the use of violence. In a purely free society, each individual is sovereign over his own person and property, and it is therefore this self-sovereignty which obtains on the free market. No one is "sovereign" over anyone else's actions or exchanges. Since the consumers do not have the power to coerce producers into various occupations and work, the former are not "sovereign" over the latter. Block: Total Repeal of Antitrust Legislation 43 To this it may be added, in order to bring it into more direct relevance with Bork, that not only do the "consumers not have the power to coerce producers into various occupations and work," but in the free society they do not have the power to coerce the producers to locate at Q1, as opposed to their preferred point, Q2. How does Bork describe the distance Q2Q1? He claims that this is a quantity of product the consumers are willing to purchase, at a price above the costs of production, and yet, because of nefarious or at least questionable doings on the part of the seller, the customer is disappointed in this desire. The area between Q2 and Qi, above the cost curve ACi and below the demand curve is defined as Al, the dead-weight loss. This is the amount of welfare that could have been enjoyed by the consumer, but is not. It is only by focusing on the buyer at the expense of the seller that Bork is able to characterize A1 as a region of dead-weight loss. In order to see this, imagine for a moment that this author had subscribed to the notion of individual, not consumer sover- eignty. If so, then how could we most accurately characterize the distance Q2Q1? No longer can we depict this merely as an amount of quantity that the consumer wishes, but is unable to buy. For under our present assumptions, there are two sides to this trans- action, not just one. Now we can more accurately delineate Q2Qi as a quantity that the consumer wishes to purchase, alright, but also as an amount that the manufacturer does not wish to sell. Similarly, our description of A1 can no longer be one of unambi- guous "dead-weight loss." Now, it must be characterized as an amount of welfare contended over by two different parties. If the sale takes place at Qi, yes, Bork is correct5; the consumer will gain this amount of. welfare. But the producer will also lose (presumably, he is unwilling to sell any more than Q2 because past that point, his marginal revenue lies below his marginal cost). So, it is by no means an unambiguous dead-weight loss A1 which must be set against a clear gain in cost savings of A2; rather, A1 is a loss only to one side of the trade, but a gain to the other. It is possible for the Borkian side of this debate to articulate several objections to the Rothbard perspective on individual ver- sus consumer sovereignty. First, it might be maintained, follow- ing Hutt (1940),that producers are themselves consumers. For example, whenever a seller acts in a way other than to maximize 5~ubjectot further objections to be made below. 44 The Review of Austrian Economics Vol. 8, No. 1 money returns, he is really "buying7's ervices from himself. There- fore, the concept of consumers' sovereignty is wide enough to incorporate both producers and consumers. If we adopt this way of looking at the matter, there are now two sets of consumers. The first, call them the consumer-consumers, are the people for whom Bork drew his demand curve. These are the ones who are purportedly suffering from the output restric- tion from Qi to Q2. The second, call them the producer-consumers, the ones engaged in this (unwarranted, improper, according to the neoclassical school) restricting of output. These two sets of consumers, according to Bork, are acting incompatibly with one another. As Rothbard (1962, p. 562) trenchantly states, In the aforementioned general sense, "consumptionn rules in any case. But the critical question is: which "consumer?" The market consumer of exchangeable goods who buys these goods with money, or the market producer of exchangeable goods who sells these goods for money? The point is, noticing that the producer, too, engages in con- sumption does not help one bit in determining whether we should force, through the majesty of the law, the producer-consumer to locate at Qi instead of Q2, in behalf of the consumer-consumer. Rather, it sets up an infinite regress. A second possible objection Bork could resort to was used by Hutt. As Rothbard notes (quoting Hutt), this is to distinguish between when a producer withholds his person or property out of a desire to use it for enjoyment as a consumers'good" . . . in which case it . . . "is a legitimate. ad, in keeping with rule by the consumer. On the other hand, when the producer acts to withhold his property in order to attain more monetary income than otherwise . . .then he is engaging in a vicious infringement on the consumers' will." (Rothbard 1962,p. 563) This, too, however, has been answered by Rothbard. He notes that it is not difficult, but rather impossible, to distinguish between these two motives. Secondly, the only reason more profit can be earned at P2Q2 than at PiQi is because of the inelasticity of demand between the two points. But this arises out of con- sumer (consumer-consumer, that is) choice! If the consumers were unhappy with this state of affairs, they could