The consumer welfare standard is used in modern antitrust enforcement to evaluate a merger between two firms. However, its original definition was corrupted in the 1970’s and has led to under-enforcement of antitrust law.

The ProMarket piece by Bush and Glick on antitrust’s consumer welfare principle leads to one question: How could an enforcement theory explicitly embracing the welfare of consumers become a “concerted and unified push” to protect high producer profits? The answer lies in the history and changing meaning of “consumer welfare.” That term was used already in the Progressive Era  to reflect the increasingly important role of consumers in economic theory about distribution.  John Kenneth Galbraith then used it in a 1954 article on Countervailing Power and his popular book with that title.  He defined it as “any type of economic behavior which lowered the prices of products to the consumer” without reducing quality.

In the 1960s even Robert Bork equated consumer welfare with high output, associating it with antitrust’s rule of reason in 1966 (“impact on output” as “primary criterion”) and resale price maintenance in 1968 (goal is “maximum output that can be achieved”). In 1973 Ralph Nader and Mark Green also spoke  of conduct which maximizes the output of goods and services as “optimizing consumer welfare.”  In his dissent in the 1972 Topco case, Chief Justice Burger complained that adopting a new per se rule would be harmful to “the welfare of consumers.” In the 1976 Sylvania decision, which the Supreme Court  affirmed, the Court of Appeals described the legislative history of the antitrust laws as reflecting an exclusive concern with consumer welfare, requiring it to “discourage restriction of output without hampering efficiency.”

Prior to Bork’s publication of The Antitrust Paradox in 1978, a thin but consistent line of thought related antitrust’s concern with consumer welfare to high market output, low prices, and competitive profits.  Then Bork read Oliver E. Williamson’s 1968 essay on antitrust and efficiency and changed his mind.  The effect was to divorce the concept of “consumer welfare” from high output and low prices.

“It was not merely naïve, but wrong-headed in just about every way.”

Williamson’s model hypothesized a merger or other practice that simultaneously increased a firm’s market power but also produced “cost savings.”  He described his model as “naïve.”  Costs were simply a horizontal line that made no distinction among fixed, variable, or marginal costs.  The practice in question produced a monopoly “deadweight” loss triangle that harmed consumers, and also an enlarged producer profits rectangle that reflected cost reductions.  Williamson did not specify the source of these cost reductions, although he referred to them repeatedly as “economies of scale.”  He concluded that the merger was efficient if the rectangle was bigger than the triangle.  Further, a small efficiency gain would be sufficient to offset a large price increase.  Indeed, Williamson concluded that under common assumptions a 4% efficiency gain would be enough to offset a 20% price increase.

Thus was born the view that a merger or other practice that led to an actual output reduction and a significant increase in profits could be lawful if even modest cost savings resulted. Williamson noted that this was consistent with economic “general welfare” theory, which viewed wealth transfers as a “matter of indifference” and thus to be ignored.  A dollar to a producer was as good as a dollar to the consumer.

Bork adopted this model in The Antitrust Paradox (Chapter 5), but with some differences. Williamson had correctly named the model “welfare tradeoff,” reflecting what it actually did.  Bork called it the “consumer welfare” model.  In Bork’s illustration (Paradox, p. 107) the firm[s] in question reduced output by approximately one half and raised prices by roughly one third, but also produced cost savings. The actual output decrease and price increase could be larger or smaller, depending on the amount of market power created or the magnitude of efficiencies. Disagreeing with Williamson, Bork argued that efficiencies are “incapable of proof.”  As a result, in most cases the defendants should not have the burden of showing them.  This was Bork’s way of addressing one problem with general welfare theories that require balancing of cardinal values — the insurmountable costs of measurement.

Bork’s book was perfectly timed, coming just at the onset of the Reagan Era neoliberal revolution in economic policy.  The use of “consumer welfare” in antitrust discussions exploded.  During the period 1960-1980 the term had appeared in 11 cases and 54 law review articles.  In the period 1980-2000 it appeared in 218 cases and 1853 law reviews.  This was hardly a conspiracy, but rather a revolution.  The Bork view has enabled dissenters on the Supreme Court to proclaim a “consumer welfare” principle even as they would approve a horrific price increase in drugs resulting from delayed entry of generics.  The majority cited it in support of a practice that resulted in higher credit card prices every time it was used.

Most users of the term do not acknowledge the distinction between true consumer welfare and Bork’s corruption of it.  Anecdotally, much of today’s opposition to consumer welfare as an antitrust principle appears to come from people who equate it with Bork’s idea and perhaps are not even aware of alternatives.

Not everyone followed along, however.  Some writers defended a “true consumer welfare” approach in rejection of Bork’s position.  The most notable dissent is the 2010 Horizontal Merger Guidelines, §10, which allows an efficiency claim only if there is no output reduction at all – that is, there is no tradeoff.

And what of the Williamson/Bork version of the “consumer welfare” principle itself?  It was not merely naïve, but wrong-headed in just about every way.  First, the model assumed a market that was competitive prior to a practice and monopolized thereafter.  But most antitrust practices don’t create monopolies.  For example, a merger of a 30% firm and a 10% firm is often challengeable even though the resulting market share is 40%.  To the extent this merger facilitates collusion, the higher prices show up in the entire market, but any efficiency gains benefit only the 40% output of the merging firms.  As a result, the model understates the social cost of monopoly by two and a half times.

Second, costs in the Williamson/Bork model were a black box.  It simply assumed that all inputs were competitive.  Any restraint that reduces product output by half reduces the demand for labor in proportion, and if this results from a change in an antitrust rule it could affect the entire economy.  If any market power exists over labor, that results in an additional welfare loss that the Williamson/Bork model did not consider.  In any event, the labor outcome was quite consistent with the neoliberal position, which included suppression of labor as one of its goals.

A particularly severe defect in the Williamson/Bork model was its assumption of significant efficiencies that result even as a practice reduces the firms’ output substantially.  When in the real world does that happen? Ronald Coase referred to this as “blackboard economics,” or the idea that if you can draw a picture of something it must represent economic reality.  The most common source of efficiencies is economies of scale, but these usually accrue at higher rather than lower output.  And what about fixed costs?  Because a fixed cost does not change with output, it goes up as output goes down.  A merger that cut the parties’ output by half would result in much greater per unit fixed costs.  Fixed costs in Bork’s example must have been insubstantial.  But then what is the source of the monopoly?  This is not to say that a merger with such a tremendous output reduction and simultaneous decrease in per unit costs is impossible, but it would be exceptional.

This critique of Bork’s corruption of the consumer welfare principle does not justify overreacting and thinking that efficiencies are irrelevant. They absolutely are, and the health of the economy depends on competition to attain them.  Here, the best approach to efficiencies is that in the Merger Guidelines, with some modification for rule of reason offenses generally:  First, efficiencies must be rigorously proven, with the burden on the person asserting them. Second, they must be shown to be reasonably necessary to facilitate the challenged restraint.  Third, they must be sufficient so that no welfare tradeoff is necessary; that is, they must leave consumers unharmed.  If we do that, then we can start calling this a consumer welfare principle.

Learn more about our disclosure policy here.