Much of economics is based loosely upon principles of physics. Social sciences, seeking street cred as a science, looked toward the hard sciences to improve their standing. Out of all the social sciences, economics perhaps did the best job of conveying the notion that it is a hard science and that its beliefs are actual scientific principles.
In Physics, Efficiency is measured in a CLOSED system
Efficiency in physics measures how much energy is preserved by a system. The greater the preservation of energy, the greater the efficiency of the system. However, most physical processes cannot be reversed, especially those that involve things such as electrical generation or heat. Most energy processes are not fully efficient: In electrical generation, for example, the efficiency of a unit can be measured by its heat rate.
According to Philip Mirowski, the law of conservation of energy prohibits the notion that energy is lost in a system. That is true if the system we are examining is a closed system. The laws of thermodynamics, upon which much of economics is based, assumes a closed system. For example, the first law of Thermodynamics states that energy cannot be created nor destroyed, only transferred in form. That transfer is unidirectional when discussing matters such as combustion. Compare with the notion of Pareto Optimality, in which no situation leads to any improvement without a transfer.
In antitrust, the analysis of conduct is quite distinct from that of physics, as it tends to ignore any losses outside the system.
Antitrust does not examine the entire effect of a merger. Instead, what is examined is the change to a “relevant market,” which is scrutinized to determine injury to consumers within that market. Thus, the merger creates changes in the relevant market only, and only those changes that remain in the market are considered positive or negative. Every other negative change is beyond the scope of examination. As an example, an efficiency to a merger in a relevant market might be massive layoffs. The effect of those layoffs are beyond the scope of antitrust, but may well count as positive for purpose of antitrust analysis.
This notion is not new to economics. Greg Mankiw had a parable related to a heroin addict. He wrote: “In some circumstances, policymakers might choose not to care about consumer surplus because they do not respect the preferences that drive buyer behavior. For example, drug addicts are willing to pay a high price for heroin. Yet we would not say that addicts get a large benefit from being able to buy heroin at a low price (even though addicts might say they do). From the standpoint of society, willingness to pay in this instance is not a good measure of the buyers’ benefit, and consumer surplus is not a good measure of economic well-being, because addicts are not looking after their own best interests.”
Mankiw’s point might be dismissed as a concern about a rationality failure in one particular market (the market for heroin). However, we could instill a different hypothetical here and reach the same result. Assume a single-product economy that produces widgets. Every worker is employed in producing widgets. A merger takes place, and two of the companies become more efficient by laying off workers. Fewer consumers of widgets means less demand, which in turn begets more mergers, laying off more workers. In such a situation, what is occurring is a wealth transfer.
However, because it is all internalized into one market, it should be a wealth transfer that antitrust cares about. It does not. Agencies would consider the layoffs efficient, no matter that the ultimate result would be collapse of the economy, because it only looks at one merger at a time without regard to follow-on mergers. The hypothetical here would be an approximation of a “closed system,” one in which there is no change in energy. Even here, however, there is loss in the sense of the workers who are lost to the system. It is possible that antitrust enforcers might start to care at around a “3 to 2” merger, but by that point it is far too late.
To those who might suggest my example ignores the perils of buyer power, again, antitrust would likely only care when the market moved from 3 to 2. Moreover, it would bring Section 1 wrath about any labor cartel, absent unionization. One might imagine one of the “efficiencies” here would be destruction of the labor movement.
The losses outside of the system of reference (in antitrust the “relevant market)” is something only occasionally analyzed by antitrust, but only for the benefit of merging parties.
One such example is “out of market efficiencies.” As the Commentary to the Guidelines states:
Inextricably linked out-of-market efficiencies, however, can cause the Agencies, in their discretion, not to challenge mergers that would be challenged absent the efficiencies. This circumstance may arise, for example, if a merger presents large procompetitive benefits in a large market and a small anticompetitive problem in another, smaller market.
In other words, even if there is an anticompetitive issue in the relevant market, large enough positive transfers accumulating into another market may trump those anticompetitive effects. Perhaps the best example of this is in the airline industry: A merger that creates monopoly in 13 smaller (rural) routes might not be challenged if it allows the merged airlines to vigorously compete with foreign carriers on international (business) routes.
To some degree, what the Guidelines is endorsing is akin to a wealth transfer that it typically believes to be beyond the scope of its analysis.
A final example of out of system losses are efficiencies in general. As the merger guidelines commentary notes:
Merger-specific, cognizable efficiencies are most likely to make a difference in the Agencies’ enforcement decisions when the efficiencies can be expected to result in direct, short-term, procompetitive price effects. Economic analysis teaches that price reductions are expected when efficiencies reduce the merged firm’s marginal costs, i.e., costs associated with producing one additional unit of each of its products. By contrast, reductions in fixed costs—costs that do not change in the short-run with changes in output rates—typically are not expected to lead to immediate price effects and hence to benefit consumers in the short term. Instead, the immediate benefits of lower fixed costs (e.g., most reductions in overhead, management, or administrative costs) usually accrue to firm profits.
These costs are typically labor costs. As such, the job losses create drags on local economies.
In Physics, transfers are often unidirectional and cannot be undone.
The second law of Thermodynamics states that the entropy of any isolated system (the unavailability of a system’s thermal energy for conversion into mechanical work) always increases. In terms of unidirectional movement, this means that “a system can only be oriented in one direction in time precisely because it cannot go back the way it came, if its path involved the dissipation of heat.” In other words, most system decisions are not reversible.
For antitrust, this would mean the processes of concentration cannot be undone in the way that neoclassical economics envisions. As Mirowski points out, this is a serious issue for economics.
Moreover, serious consideration of the notion of irreversibility would clash with the dictum that the market can effectively undo whatever man has wrought. Finally, and most pertinently from the vantage point of a discipline seeking to emulate physics, the concept of energy in thermodynamics is thoroughly unpalatable when cooked down into the parallel concept of utility in neoclassical economic theory. For example, the parallel would dictate that utility/value should grow more diffuse or inaccessible over time, a figure of speech possessing no plausible allure for the neoclassical research program.
This would suggest that allowing a market to concentrate would be incredibly difficult to undo. In other words, Type II errors are very serious because they are not undoable. This comes as no surprise to antitrust enforcers, who are always concerned about post-merger remedies. It is impossible to unscramble the eggs.
Physicists know what they are seeking to measure
I pause here only to recognize that there is a problem of WHAT is to be measured in markets. Much work has been done on the measurement problem in social sciences. Such a discussion would be the basis of its own post. Measurement problems abound not only because of the inability to make interpersonal comparisons of the utility consumers receive from consumption but also in terms of productive efficiency. As Mirowski states:
The metrics of the efficiency of inputs in a production process are rarely identical with the units in which the commodities are bought and sold. Oil is sold by the barrel, but its efficiency as an input depends on its BTU rating, or perhaps its sulfur content in milligrams per liter, or its Reynolds number, and so on. Further, this metric will vary from process to process, even if one is looking at the same barrel of oil….Such distinctions are critical for any serious representation of a production field, because they raise the possibility that, as long as the axes of the field formalism are confined to commodity space, the field cannot be analytically defined….We hasten to add that this entire discussion is a disquisition on the problems of neoclassical economic theorists, and not the practical actors in the actual economy. They do not separate their world into airtight divisions of substitution versus innovation; nor do they keep tabs on marginal products (unless they have been to business school); nor do they have difficulty keeping track of the boundaries of their economic activities. Instead, they actively constitute the identity of their economic roles and artifacts as they go along.
The problem is exacerbated in antitrust if Professor Herb Hovenkamp is correct and consumer welfare means more than one thing.
Incomplete Measurement Favors Concentration and Externalization of Economic Injuries
Thus, that to which antitrust enforcement agencies apply the notion, the “relevant market,” is not a closed system. A firm claims efficiencies, typically expressed by an economist making over $1000 an hour, it typically means job loss, layoffs, and the foisting of costs outside the system of examination. This is what Professor E.K. Hunt called the “invisible foot:”
The “invisible foot” ensures us that in a free-market . . . economy each person pursuing only his own good will automatically, and most efficiently, do his part in maximizing the general public misery. . . . To paraphrase a well-known precursor of this theory: Every individual necessarily labors to render the annual external costs of the society as great as he can. He generally, indeed, neither intends to promote the public misery nor knows how much he is promoting it. He intends only his own gain, and he is in this, as in many other cases, led by an invisible foot to promote an end which was no part of his intention. Nor is it any better for society that it was no part of it. By pursuing his own interest he frequently promotes social misery more effectually than when he really intends to promote it.
In other words, market transactions promote pervasive negative externalities. In examining only the relevant market and consumer welfare, the Antitrust Enforcement agencies are assuring they ignore effects foisted upon the remainder of the economy, the externalization of costs associated with the merged firms’ transaction. Those negative externalities are also considered efficiencies within the relevant market, and their exportation outside of the market is considered a value-less transaction. It could be the case that the positive effects inside the system are outweighed by the negative effects outside the system as those effects are externalized. It could be the case as well that the positive effects are overestimated, as might be the case in a merger claiming efficiencies of system integration that take 4 times longer than projected.
Consumer Welfare Theory isn’t Science—It’s A Policy (an Ethical Consideration)
If one changes the lens back to physics, the framework of analysis would be considered wrong because there cannot be loss of energy to the system. Thus, the focal point of the analysis is always a singular framework that tends to benefit the merging parties except in the most extreme cases. Thus, the “science of antitrust economics” is incomplete, because it disclaims losses to the system without further analysis. By focusing on consumer welfare in one relevant market, it assures a most narrow antitrust goal that in effect assures a large externalization of costs. Indeed, it solicits it. As an example, the consideration of out-of-market efficiencies without any consideration of out-of-market costs is indefensible outside of some ethical argument in favor of consolidation.
Efficiency in physics is a valueless concept, the purpose of which is to describe as the work performed per quantity of energy. It is the maximization of something to be measured, output, subject to the amount of input available. It is examined on a system basis. It is not a position of advocacy. Nor is it identical to the view economists have of the term.