chevron-down Created with Sketch Beta.

Antitrust Magazine

Volume 38, Issue 2 | Spring 2024

The Critics Are Wrong: How the Robinson-Patman Act Has Been Misunderstood By Its Detractors

Mark Poe


  • There is no evidence that the Robinson-Patman Act causes higher consumer prices or discourages lower prices; those are just ipse dixit attacks by laissez faire lawyers, not economists.
  • In real-world application, private enforcement of the Robinson-Patman Act encourages lower prices to be available to more customers, given that the disfavored plaintiff is invariably asking for the favored purchaser’s price.
  • Congress very clearly adopted the Robinson-Patman Act for the purpose of protecting and preserving small businesses. The judiciary should be faithful to its repeated proclamations that Congress, not the courts, is responsible for choosing desirable economic policy.
The Critics Are Wrong: How the Robinson-Patman Act Has Been Misunderstood By Its Detractors
Alan Thornton via Getty Images

Jump to:

The Robinson-Patman Act (“RPA”) is the favorite punching bag of antitrust commentators. The criticism has led to memorable lines from conservative legal scholars, such as Judge Robert Bork’s description of the RPA as “the misshapen progeny of intolerable draftsmanship coupled to wholly mistaken economic theory,” and Professor Herbert Hovenkamp’s claim that “[t]he chief ‘evil’ condemned by the Act [is] low prices, not discriminatory prices.” The Antitrust Modernization Commission joined this refrain in 2007, proclaiming that “[t]he Act is fundamentally inconsistent with the antitrust laws and harms consumer welfare.”

In conjunction with the July 9, 2021 “Executive Order on Promoting Competition in the American Economy,” the FTC has shown a renewed interest in RPA enforcement. Unsurprisingly, the agency’s move to revive the RPA has prompted opponents to revive their criticisms. For example, the Cato Institute recently published an article saying that the RPA “inhibit[s] more efficient firms from receiving . . . discounts” and that “[i]f enforced again it would no doubt smother efficiencies once more, resulting in higher prices for customers.”

This strongly worded condemnation—relying on citations to luminaries like Judge Bork and Professor Hovenkamp—would lead most readers to assume that the criticisms are supported by robust economic literature and empirical analysis. That is simply not true. As recently as 2014, the then-Deputy Director of the FTC’s Bureau of Economics observed that “[a] formal study of the effects of the Robinson-Patman Act on prices has not been conducted, to my knowledge.” A year later, Professor Christina DePasquale of Johns Hopkins University observed that “[t]here is no shortage of studies detailing the harmful consumer effects of price discrimination,” and discussed some of them. Note the italics: Professor DePasquale (a Ph.D. economist) refers to studies showing the “harmful consumer effects” of price discrimination; not the supposed harmful effects of prohibiting price discrimination—the claim that RPA opponents continually assert.

The familiar criticisms have been repeated with such frequency and certainty that they have become accepted as truisms. But despite more than 40 years of repetition, these truisms are entirely lacking in empirical support. Tracing the lineage of these familiar criticisms back through the literature reveals that the supporting citations are only to prior works voicing the same criticisms. Indeed, the criticisms did not originate in the work of economists at all, but in the works of laissez faire legal scholars. Before squashing a potential revival of the RPA, it is time to query whether the routine condemnations are valid to begin with.

The most familiar criticisms are based on unsupported assumptions or are targeted at straw-men mischaracterizations of the RPA. Furthermore, the criticisms are based on a misunderstanding (or disregard for) Congress’s purpose in adopting the RPA.

The Main Criticisms of the RPA and Their Errors

The chief criticisms of the RPA largely repeat the same themes and are typically supported with cross-citations to each other. The criticisms below are drawn from the AMC Report, a frequently-cited 2000 article by Professor Hovenkamp, and the commentary of Judge Posner.

“The RPA results in higher prices to end consumers.”

The AMC Report notes that “‘Congress considered it to be an evil that a large buyer could secure a competitive advantage over a small buyer solely because of the large buyer’s quantity purchasing ability.’”The AMC Report then rejoins that “[i]n its operation, however, the Act has had the unintended effect of limiting the extent of discounting generally and therefore has likely caused consumers to pay higher prices than they otherwise would.” Oddly, the AMC did not follow this claim about the “operation” of the RPA with examples in which RPA enforcement either “limited discounting” or “caused consumers to pay higher prices.” Nor did it cite to any study showing either thing. Instead, it cited Professor Hovenkamp, who “explained” that the RPA “was designed to protect small businesses from larger, more efficient businesses,” with the “necessary result [of] higher consumer prices.” But that section of Professor Hovenkamp’s treatise cites nothing at all.

Professor Hovenkamp’s unsupported claim is emblematic of the standard criticisms. Crucial to his claim is the assumption that “larger businesses” are “more efficient” than their smaller competitors, which is asserted as a fact by most critics. But that is plainly a quantitative claim that could be empirically evaluated. Sound policymaking should be preceded at least by an effort to verify the empirical premise and to understand the assumptions underlying the “efficiency” argument.

To evaluate relative “efficiency,” one would first need to define what it means to be “more efficient.” If “efficiency” includes phenomena like employing workers on a part-time basis at minimum wage, while externalizing their healthcare costs to Medicaid and subsidizing their income with taxpayer-funded food stamps, one could certainly question whether that is the type of “efficiency” that our public policy should promote. Yet critics seldom if ever pause to explain what they mean by “more efficient.” Instead, they simply assert that big businesses are more efficient and that “higher consumer prices” will result from RPA enforcement. These pronouncements then get repeated as if they have been empirically established.

As against the complete absence of evidence upon which the RPA critics base their efficiency claim, there are contrary real-world examples. In the RPA cases that our firm has brought, the “favored purchaser” is typically either Sam’s Club or a Costco division called Costco Business Centers, while the “disfavored purchasers” are family-owned grocery wholesalers who cater to the same small retailers that Costco and Sam’s target. The RPA’s critics would presumably hypothesize that the “efficiency” advantages of Costco and Sam’s Club manifest in lower margins on the products they sell, in turn allowing for lower consumer prices. But the evidence in our cases has shown that the opposite is true. Costco typically applies a 12-15% markup on its merchandise, while our clients (who compete with Costco for sales to the same customer base) typically apply a 6-8% markup, sometimes as high as 10%. Nevertheless, Costco is able to sell the same merchandise at a far lower price (and at higher margins), financed by the discriminatory pricing it receives from suppliers. Thus, any cost savings to the consumer are due not to the (wholly assumed) “efficiency” that results from Costco’s bigness, but to the fact that Costco’s bigness permits it to secure a discriminatorily favorable price, while realizing supracompetitive margins. If family-owned wholesalers were given the same price that Costco receives, but continued to charge only a 6-8% markup, customers of those businesses would pay far lower prices. Likewise, if Costco faced price competition from those small businesses, it would almost certainly respond by knocking down its 12-15% markup, meaning that customers of Costco would also pay less.

The legal scholars’ assumption that permitting price discrimination is good for consumer prices is undermined by an economic understanding of a seller’s purpose in engaging in price discrimination, with reference to the familiar chart of supply and demand curves. Economically, the goal of price discrimination is to transfer the consumer surplus to the seller, by charging each customer the maximum amount that it is willing to pay, all the way down the demand curve to the last purchaser to whom the supplier can profitably make a sale. If that last sale yields a normal profit to the supplier, then all of the customers to his left have paid a higher price than they would pay at a single, profitable, non-discriminatory price. It is beyond the scope of this article to evaluate whether a supplier’s ability to capture the “consumer surplus” from intermediate purchasers (i.e., the resellers to whom the RPA applies) would similarly be expected to result in higher end-consumer prices on average, but the point is that no RPA critic appears to have even considered such issues.

“The RPA condemns low prices by discouraging discounting."

Beyond unsupported claims of “efficiency,” the other device that RPA critics use to claim that the RPA results in “higher consumer prices” is to assume that the supplier’s baseline price is the higher, disfavored price. Upon that assumption, they erect the straw man that the RPA prohibits discounts. Only under that view can one claim that the RPA “discourage[s] discounting that otherwise would lead to lower consumer prices” and conclude that the “chief ‘evil’ condemned by the Act [is] low prices.” As a professor of economics at Princeton University observed, however, “[t]his line of argument should not be a convincing one.” After all, the script is easily flipped by using the critics’ same tactic, but assuming that the baseline price is the one given to large national retailers (such as Walmart or Costco in the example above)—after all, that price is presumably above the supplier’s marginal cost, and it allows the supplier to make at least some amount of profit. Under that paradigm, the “necessary result” of RPA enforcement would be “lower consumer prices,” i.e., those that result when the disfavored customers receive a lower price than they could otherwise negotiate for.

In other words, it is just a matter of framing for the critics to claim that the RPA prohibits “discounts.” Section 2(a) says nothing about “discounts”; what it prohibits is “discriminat[ion] in price.” Under the view laid out here, one can just as easily frame the RPA as prohibiting suppliers from charging “higher prices” to small businesses.

As between the two paradigms, there is far more evidence supporting the validity of the paradigm urged here. As an initial matter, could anyone doubt that lower prices are achieved through bargaining between the seller and a strong buyer, as opposed to unilateral price-setting by the supplier? And if one agrees that bargaining between a seller and strong buyer is more likely to yield lower prices, does anyone doubt that large, national retailers are better positioned to drive a hard bargain than is a single-location independent grocer in Winnemucca, Nevada? If large resellers are able to negotiate better prices, then prohibiting discrimination from that price baseline gives all buyers access to the large reseller price, and would undoubtedly result in lower consumer prices to more people. This is because “[i]f the bargaining is done by the stronger of two buyers, then the lower price becomes the price paid by everyone.”

Real-world observations of the RPA “[i]n operation” support the view that the proper baseline price is the one given to the big buyer. Of the hundreds of private-enforcement RPA cases where the plaintiff seeks prospective relief, the plaintiff appears to have always demanded to be given the lower, favored price. Success in those cases would “necessar[ily] result” in lower prices, not higher ones, as the critics proclaim.

That pattern is supported by the cases brought by the author’s firm in which a settlement has included a price component. At least in our cases, not once has the favored purchaser’s lower price been taken away; in every case that lower price has been extended to the disfavored purchaser. It is also supported by anyone familiar with the business practices of major buyers—those firms would not tolerate a supplier raising its prices 10-20% to match the price it extends to independent competitors.

Of course there are complications to a simplistic view that prohibiting discrimination would result in all purchasers receiving the large-buyer price. Most obviously, suppliers likely use the higher prices charged to disfavored purchasers to subsidize the lower prices they charge to the favored. But that is hardly a reason to suspect that the RPA results in higher prices to consumers on average. Rather, it seems consistent with the RPA’s fears: that in a world that permits price discrimination, smaller businesses will find themselves subsidizing the profitability of their larger competitors. More to the point of this article, the RPA’s critics do not seem to have even considered these issues.

“The cost-justification defense is inadequate to protect discounts that are actually cost-justified."

The RPA allows a supplier to discriminate in pricing where the lower price to the favored customer is “justified” by cost savings that the supplier realizes from that transaction. This is the so-called “cost justification defense” set out in the proviso to the RPA’s section 2(a), which allows “due allowance for differences in the cost of manufacture, sale, or delivery resulting from the differing methods or quantities in which such commodities are . . . sold.” For example, if a buyer performs its own packaging of a supplier’s product, or accepts delivery FOB at the supplier’s facility instead of having the supplier bear the shipping cost, the supplier is permitted to discount its price to that extent.

A common refrain from RPA critics is that this defense is inadequate, because it is “difficult and costly to meet the requirements of the cost-justification defense.” Judge Posner puts it more starkly, claiming that the defense is “virtually impossible” to meet, and “[t]hus the Act has in practice undoubtedly operated to suppress price differences that were justified by differences in cost.” In keeping with the other critics, Judge Posner has not identified any of the examples that “undoubtedly” exist, but makes his point just by proclamation. Moreover, no one has endeavored to explain why it is so hard for a supplier to show that its lower price to a favored purchaser is justified by the cost savings it realizes. This shouldn’t be difficult; calculating cost-of-goods-sold is fundamental to business accounting.

In reality, the difficulty in establishing the cost justification defense is not that the arithmetic is hard. Rather, it stems from the fact that the supplier did not base its favorable pricing on cost savings at all, but instead based it either on the price that the favored purchaser demanded, or on an arbitrary figure like ten or twenty percent. In that circumstance, no one should be surprised to find it “virtually impossible” for the supplier to show that the discount had any relationship to its cost savings. That is hardly grounds to condemn the defense as inadequate. Rather, it is the inevitable result of a supplier awarding a favored purchaser a non-cost-justified price.

Another real-world data point is useful here. In the only two cases that the author’s firm brought where the supplier has produced relative-cost information, the supplier has internally recognized that due to the exacting packaging and delivery requirements of Sam’s Club and Costco, it is more costly to sell to these giants than to their independent competitors. It is certainly possible that there are contrary examples, but unless the critics can identify even one, they should acknowledge that they are just guessing.

Courts Should Not Set Economic Policy That Is Contrary to Congress’s Expressed Intentions

The RPA’s critics fall into two camps: those who call for repeal, and those who would have the courts “interpret” the RPA into a nullity. After fifty years of unsuccessful repeal efforts, the RPA faces a much greater risk from the “reinterpret” camp. On that front, Justice Gorsuch’s recent opinion in Epic Systems Corp. v. Lewis should be a sufficient response: “This Court is not free to substitute its preferred economic policies for those chosen by the people’s representatives. That, we had always understood, was Lochner’s sin.” With specific respect to the RPA, Justice Scalia took a similar view: “The short answer to this argument is that it should be addressed to Congress.”

Beyond the fact that Congress is the branch charged with “regulat[ing] Commerce . . . among the several States,” the judiciary is not equipped to competently establish economic policy. None of the current Supreme Court justices, for example, appears to have any rigorous training in economics. And they make decisions based on 150 pages of legal briefing, written by lawyers who by design are not providing objective analysis aimed at maximizing social welfare, but are urging whatever “economic policy” favors their client in the case. Given those constraints, the judiciary is particularly vulnerable to being misled by citations to pithy economic claims from famous legal scholars, while page limits prevent robust exploration of the validity of those claims.

The proposal most often advanced by the “reinterpret” crowd is to encourage the judiciary to ignore Congress’s expressed intention of protecting small businesses from discriminatorily high prices, invoking the “oft-quoted chestnut” that the purpose of the antitrust laws is “the protection of competition, not competitors.” But in enacting the RPA, Congress took the view that competition was promoted by protecting competitors. That purpose has been repeatedly recognized by the Supreme Court. In Morton Salt, the Court noted that the RPA “was intended to justify a finding of injury to competition by a showing of ‘injury to the competitor victimized by the discrimination.’” That intention was taken from the Senate’s committee report, which described the Clayton Act’s prior price-discrimination prohibition as having been “too restrictive, in requiring a showing of general injury to competitive conditions.” The committee report also explained that instead of requiring proof of general injury to competition, the RPA was intended to protect “the competitor victimized by the discrimination,” based on Congress’s belief that doing so would “catch the weed in the seed [to] keep it from coming to flower.”

The House took the same view. It repeated almost verbatim the Senate report’s finding that the need for “a showing of general injury” was “too restrictive,” and that “[t]hrough this broadening of the jurisdiction of the [Clayton] act, a more effective suppression of such injuries is possible and the more effective protection of the public interest at the same time is achieved.”

The critics seem to argue that Congress was wrong in concluding that competition is promoted by ensuring a level playing field to competitors. But judges are not entitled to declare a congressional policy “wrong.” And it would be an egregious intrusion on the separation of powers for the judiciary to “interpret” a statute to mean exactly what Congress rejected (e.g., the need for a plaintiff to prove “general injury to competitive conditions”).

Indeed, the Court’s foundational antitrust precedent recognizes that the judiciary is not permitted to undermine Congress’s avowed intent. The “oft-quoted chestnut” that antitrust laws are intended for “the protection of competition, not competitors” first appeared in Brown Shoe Co. v. United States, which involved the merger provisions in section 7 of the Clayton Act—the very act that Congress deemed to have been “too restrictive” in its price-­discrimination prohibitions, thus requiring an amendment (that is, the RPA) to protect competitors themselves. Even with respect to section 7, Brown Shoe cautioned that “we cannot fail to recognize Congress’ desire to promote competition through the protection of viable, small, locally owned businesses”—and recognized that “[w]e must give effect to that decision.” It therefore affirmed the district court’s order prohibiting the merger at issue. Unfortunately, the judiciary appears to have forgotten these passages, carrying forward only the “chestnut” that antitrust law protects “competition, not competitors,” as if that were the holding of the case, when in fact it was the opposite.

A similar incremental creep is seen in the refrain that the RPA must be construed to be consistent with the broader purposes of other antitrust laws. That general principle was first set forth in Automatic Canteen Co. v. FTC, where the FTC interpreted the “knowingly” requirement in section 2(f) to reach a seller’s constructive knowledge that it had received a discriminatorily favorable price. The Supreme Court rejected that interpretation, requiring actual knowledge consistent with the Sherman Act. In doing so the Court explained: “it is our duty to reconcile such [agency] interpretation, except where Congress has told us not to, with the broader antitrust policies that have been laid down by Congress.” Over the ensuing years, the proviso “except where Congress has told us not to” has been dropped, leaving only the admonition that courts are bound to “construe the Act ‘consistently with broader policies of the antitrust laws.’” Only time will tell whether that instruction—and the Court’s own precedent on it—is heeded.


Professor Hovenkamp notes that “[v]ery few statutes have survived such long-lived and unrelenting criticism as has been directed against the Robinson-Patman Act.” As explained above, the main tenets of that unrelenting criticism are empirically unsupported, and made rhetorically persuasive only through the use of straw men and framing devices.

Nevertheless, the RPA persists. To understand why, it is useful to consider the constituencies on both sides. While the RPA’s critics often claim support from “economists,” the specific economists they have in mind are rarely identified. The most stalwart opponents turn out not to be economists per se, but legal scholars such as Judge Posner and Professor Hovenkamp, and the twelve lawyers on the Antitrust Modernization Commission, eleven of whom were either then or now corporate defense lawyers or in-house counsel of major corporations that would likely be either “suppliers” or “strong buyers” in the RPA context. On the other side, small-business groups remain stalwart proponents of the RPA, as exemplified by the recent launch and advocacy of the Main Street Competition Coalition, described as a coalition of “[i]ndependent grocers, pharmacies, restaurants, convenience stores and farmers.”

It is understandable why the RPA has persisted in the face of decades of repeal efforts. Surely the average American (and legislator) would agree with the general proposition that small businesses should share a “level playing field” with their giant chain competitors. And if they delved into the economics of the issue, they would be understandably skeptical of a pricing practice whose very purpose is to transfer consumer surplus from purchasers to suppliers, particularly when even critics like Professor Hovenkamp recognize that “[a]ll forms of persistent price discrimination transfer wealth away from consumers and toward sellers.”