Assistant Attorney General Thomas Barnett of the antitrust division of the Department of Justice writes an interesting summary of what some would consider neoliberal views on antitrust. You can find his essay Maximizing Welfare Through Technology Innovation on the DOJ’s website. Last year an article about his rejection of Google’s claims against Microsoft appeared in the New York Times. This article I’ve written is an analysis of his substantive claims and rationale behind the more laissez faire position on trusts. It is also a critique of these views, arguing that Barnett and his colleagues have been nearsighted in their analysis of the collusive efforts of capitalists, workers and cartels.

For Barnett and the antitrust officials at the DOJ to adopt the approach, “first, do no harm” articulates a sort of Hippocratic, oath-like attitude toward maximizing society’s welfare. Barnett explains that in the context of antitrust, the principle should be interpreted as a careful intention to not “kill the goose that lays the golden egg.” This comes under his broader discussion of three different types of efficiency, static, incremental-dynamic, and leap frog-dynamic.

Under the administration of Barnett, the antitrust division finds the dynamic efficiencies, especially where “entirely new forms of production” (that is, to “leap frog”) are realistic potentialities, to be the primary incentive for not punishing firms with greater market power. This is quite unlike the view that earlier administrations had taken to, as evidenced by the Philadelphia Bank case, where two banks with relatively little market power were broken up by the antitrust authorities in the 1970s.

The type of competition which Barnett says antitrust officials should not mistake for a violation of antitrust law is what Joseph Schumpeter terms “creative destruction”. This sort of competition is a long run phenomena, dealing with “new technologies”, “new sources of supply”, “new types of organization”, and “new commodities”. It is the constant restructuring and reorganization of the market through competitive dynamism.

For this reason, here quoting Judge Easterbrook, Barnett believes that “aggressive, competitive conduct by a monopolist” should not be mistaken for anticompetitive conduct or performance since it is “highly beneficial to consumers.” Competitive and exclusionary conduct looks the same from the court’s point of view. This is the difficulty, Barnett says. But instead of condemning the monopolists as social welfare-destroyers, courts should “exercise appropriate caution” when dealing with them.

He adds that the courts should to the best of their ability “prize and encourage” dynamic efficiency through antitrust enforcement. That is, prize and encourage monopoly market structure. Because monopoly structure does not determine monopoly performance, the drive towards greater exclusion is interpreted as an exemplification of sort of pro-competitive performances that do not warrant the DOJ’s scrutiny toward the market’s structure. Exclusionary conduct, in short, is welfare-maximizing. If two firms compete head-to-head in an industry, arguably, every bit of exclusionary conduct (provided a legal system conducive to competition) helps the firms to perform more efficiently, and thus more desirably.

Even if there is only one firm in an industry, the threat of potential competition through the sorts of dynamic efficiencies discussed by Schumpeter ensure the market performance will be efficient and desirable in the long run. If the monopoly firm is not dynamically efficient, it will be taken over by new entrants that are dynamically efficient, and the process of creative destruction continues.

Yet for this same reason, allegedly, Barnett believes that cartels should be the cause of tremendous worry. “Their entire purpose is to avoid disruptive forces of competition,” he says. To live the “quiet life”—as Nobel Laureate John Hicks wrote—is their ultimate end. Barnett adds that cartels are not only dynamically inefficient, they are statically inefficient as well. This is to say that cartels not only decrease social welfare in the short run (static), but in the long run as well (dynamic). This is “double the ‘calamity’” and deserving of “severe sanctions available against them under U.S. law.”

Now, if one were to attack Barnett’s position on trusts, what would it look like? One position, that of the Marxist, might argue that antitrust forces like these only stave off the destruction of the capitalist system. By enforcing the laws on cartels, the DOJ ultimately keeps away the inevitable tendency toward the paroxysm of the capitalist system. I’m not sure whether the appropriate Marxist thing to do is to breakup the trusts (cartels, monopolies, oligopolies, etc.) and give the industries to the workers, or whether to allow capitalism the ultimate free-for-all it needs to force workers into revolt. I don’t consider myself a Marxist so my confusion is warranted to some degree.

At any rate, the modern liberal, on the other hand, would argue that the monopoly structure is itself evidence of poor performance and poor conduct. To the liberal, market structure determines conduct which in turn determines performance. This was embodied in the work of Joe Bain during the 1970s, and to this day modern liberals have made essentially the same argument with regards to monopoly market structure. Since the 70s, countless empirical and econometric studies have shown this falsifiable hypothesis to in fact be false: market structure does not determine market performance.

Of course, there have been more modern restatements of the structure-conduct-performance paradigm, and yet there have also been restatements to counter this as well. For the most part, I don’t agree with the liberal’s basic contention that structure determines performance. It seems to me, as it does to Barnett, that performance determines structure. Especially when one takes into account the dynamic aspect of competition, the cost-benefits to dynamically efficient firms will lead to market structures in which inefficient and poorly performing firms must either vanish or simply refrain from entering a market.

Probably one of the less talked about areas of trusts are labor trusts, or unions as they are more commonly known. In general, modern liberals favor unions and disfavor monopolies. Yet a union is in fact a monopoly on the labor force. It seems to me that this is a one-sided view. If I would allow the performance of labor market to determine the structure of the labor market without interfering with its marketplace, I should also allow the performance of the capitalists’ market to determine the structure of the industry without interfering with the marketplace.

Another point about labor forces is that they are more likely to be unionized when there are monopolistic structures than when there is perfect competition, since a firm’s variable costs (among other things, this primarily denotes wages) are not fixed in the short run or in the long run. They can be cut if needed, and in a perfectly competitive market cuts to variable costs are needed more due to elasticities and lack of market power. In monopoly market structures, therefore, it would seem the level of cooperation (collusion, unionizing, etc.) increases relatively simultaneously with capitalists as it does with the labor force.

Because we ought not interfere with the aggregation and power of the capitalists, according to Barnett’s analysis, why ought we to interfere with the aggregation and power of the labor force? I am not suggesting this is Barnett’s position. I am merely using this example to make my argument more clear. It is within a labor union’s interest to be dynamically efficient and exclusionary just as it for the monopoly capitalists. The union’s performance keeps the firm in a position of market power and dominance, and, so long as the union has market power, the threat of striking or other predatory practices keeps the unions’ profits high. With the background analysis that I have provided, I now wish to explain why I think Barnett’s argument is ultimately unsatisfactory.

When capitalists collude with other capitalists, how is the DOJ to determine whether this is competitive or anticompetitive? Barnett suggests that performance, as opposed to structure, is the indicator the antitrust division should be concerned about. However, Barnett points to the need to break up cartels as the DOJ’s highest antitrust priority. Yet existence of cartel is a structure; it is not an indication of performance. This is an important point. To say that monopoly structure is to be “prized and encouraged” because it may be performatively welfare-maximizing, while at the same time say that cartel structures are to be “severely sanctioned” because it is a structure which cannot possibly perform in a welfare-maximizing way is a non-sequitur.

Perhaps an economist or an attorney general has already come along and said this, but we must ask ourselves what reasonable distinction can be drawn between a collusive capitalist who merges, acquires, sets prices, restricts output, earns profit, (etc.) within a monopoly market structure, and a collusive capitalist who merges, acquires, sets prices, restricts output, and earns profit within a cartelized market structure? Another way to put this would be to ask, What is the difference between two capitalists, one in a monopoly structure and the other in a cartel structure, both whose “entire purpose is to avoid disruptive forces of competition”? To use John Hick’s well-known phrase we can likewise ask, What is the difference between the “quiet life” of the monopolist, the “quiet life” of the cartel, and the “quiet life” of the unionized worker?

The argument goes that even if there are no other static competitors in a monopoly market, that the threat of potential competition through dynamic efficiency implores the market to become efficient. Firms may also be checked by gateway entrants from other markets who can enter the industry at the significant cost advantages. All attempts at exclusionary practice, arguably, lends itself toward greater social welfare through competition. Yet the greatest source competition, I would argue, comes from a firm’s competition with the labor market. It seems this would be especially evident when there are no other market challengers.

Let me restate this last point. The labor market is the greatest source of competition, since, even if firms exclude to the point where even gateways from other markets have been exhausted, the competitive pressure from the labor market will continue to ensure revenues to capitalists do not exceed the cost of production by far. The standard microeconomic argument that “economic profits attract entrants into the market” and thus lower prices can equally be said to apply here. Instead of attracting entrants, however, it attracts greater demand for wage increases, which in this example would appear to be a form of inflation.

In short, it seems to me Barnett is having the same sort of dynamic misconceptions about collusion and the long run economy that attorney generals of an earlier era had, albeit, now applied to much greater forms of market power. Barnett and his colleagues may believe the DOJ in the 1970s had it wrong when it broke up two banks in Philadelphia with insignificant market power, yet attorney generals now show the same unfounded concern for collusive activity on a greater scale in this era. This shows the extent that the DOJ’s views on dynamism reaches. Perhaps an economist or an attorney general will soon come out and say that we cannot make the analytic case against cartels for the same reasons that we cannot make the analytic case against the monopolies, as we once thought. But as of now such argument to my knowledge has not reached fruition.