In 1992, the U.S. Supreme Court held in Eastman Kodak Co. v. Image Technical Services that a firm without market power in photocopiers might still possess market power in photocopier parts and service. The Court’s logic turned on opportunistic hold-up: Kodak could profit by trading short-run exploitation of locked-in customers for long-run losses in equipment sales. That tradeoff, the Court concluded, could establish antitrust market power.
Benjamin Klein’s 1993 article, “Market Power in Antitrust: Economic Analysis after Kodak,” alls this a category error. Hold-up is real; Klein helped define it in “Vertical Integration, Appropriable Rents, and the Competitive Contracting Process” (1978), co-authored with Robert Crawford and Armen Alchian.
But hold-up is not market power. The Court took the framework Klein helped build and pressed it into service for a task it was never meant to perform.
That misstep carries a broader lesson for law & economics. The Court in Kodak relied on sound economic concepts—hold-up, switching costs, lock-in—but aimed them at the wrong legal question. Law & economics demands more than importing good economics into legal disputes. It requires matching the right economic concept to the right legal question. Klein’s contribution lies in doing exactly that—and in understanding both sides well enough to know the difference.
Mistaking Hold-Up for Market Power
Independent service organizations (ISOs) had long repaired Kodak copiers—often at lower cost and higher quality than Kodak. In the mid-1980s, Kodak restricted access to replacement parts, effectively tying parts and service and pushing ISOs out. The legal question: could Kodak, which lacked market power in the equipment market, still wield market power in the aftermarket for parts and service?
The Court said yes—in principle. A firm without power in the foremarket might still exercise power in the aftermarket if buyers face high switching costs, significant information costs, and the kind of lock-in that makes single-brand aftermarkets viable. The dissent, led by Justice Antonin Scalia, took a different view. In a competitive equipment market, prices should adjust to reflect any aftermarket exploitation, protecting buyers ex ante at the point of purchase.
Klein argues that both sides missed the mark—but the majority’s mistake runs deeper. The dissent leaned on a model of perfect competition: fully informed buyers, zero switching costs. That model doesn’t describe real markets, and Klein agrees the Court was right to reject it.
The majority, though, made a different error. It identified a plausible hold-up story and treated it as evidence of antitrust market power.
That move collapses an important distinction. Hold-up appears everywhere: landlords and tenants, employers and employees, manufacturers and dealers. If hold-up equals market power, then routine contract disputes become antitrust cases. That can’t be right.
Klein’s core point turns on timing. As he puts it, “the seller’s market power must be determined at the point in time when the tie-in contract was agreed to, not at the time when a ‘hold-up’ is occurring” (p. 58). The majority focused on whether Kodak could exploit buyers after they made equipment-specific investments. That shows hold-up potential, but says nothing about whether Kodak had market power when buyers entered the arrangement.
Buyers need not predict precise aftermarket prices to protect themselves. They only need to recognize that switching costs exist and that hold-up is possible. From there, they can rely on reputation or contractual safeguards. Hold-up may redistribute surplus from locked-in customers, but it does not restrict output in the equipment market.
Why Kodak’s Tie Looks Like Metering, Not Muscle
If hold-up doesn’t explain Kodak’s tie, what does? Klein points to discriminatory marketing. The record is murky. Kodak claimed the parts tie dated to 1975, but ISOs may have purchased parts freely before the 1985 policy shift. Did customers anticipate the restriction? The evidence is unclear. Even so, Klein argues price discrimination fits the facts better.
The idea traces to Aaron Director and Edward H. Levi’s “Law and the Future: Trade Regulation” (1956). That paper displaced the old “leverage” theory of tying with a metering explanation. IBM tied punch cards to its machines not to extend monopoly power, but to meter usage it could not directly observe. Klein applies that logic to Kodak, with a more tailored mechanism.
Start with two customer segments: self-service buyers, who maintain their own equipment, and purchased-service buyers, who rely on Kodak. Within the purchased-service group, the tie allows further discrimination based on type and urgency of demand. Customers who value fast on-site repair and single-vendor accountability pay more through bundled service pricing. Kodak sets equipment and parts prices with the self-service segment in mind. Service pricing then acts as a residual, extracting more from purchased-service buyers.
The self-service exception does the real work. Why not simply overprice parts and meter that way? Because self-service customers would respond by substituting toward more labor-intensive servicing to avoid high parts prices, distorting the parts–labor margin. By tying service and exempting self-servicers, Kodak avoids that distortion.
This arrangement also reflects customer preference. Purchased-service buyers value bundled accountability. The self-service carveout suggests Kodak could not force unwilling buyers into the bundle. Customers chose among aftermarket options when they bought the equipment. The fact that many chose Kodak’s bundle suggests the arrangement created mutual gains.
The deeper point: price discrimination neither requires nor implies antitrust market power. Most firms face downward-sloping demand curves and can price above marginal cost. Economists often treat that as evidence of market power. Klein rejects that move. A firm’s ability to price above marginal cost given its own demand curve—what he calls “individual pricing discretion”—is not the same as antitrust market power.
Consider branded grocery products. Lester Telser estimated brand-level demand elasticities for orange juice, coffee, beer, and similar goods, finding most fall between 2.5 and 5. That implies markups of 25% to 67% over marginal cost. Later work finds similar results for cereals. By the Lerner Index, these firms appear to have substantial market power. Yet they hold small market shares, face elastic supply from rivals, and operate in markets with free entry. No one treats them as monopolists.
Klein labels this phenomenon “individual pricing discretion.” It is widespread, competitive, and typically benign. Antitrust market power is something else: the ability to raise market-wide prices by restricting output. That distinction matters. It separates conduct that warrants antitrust scrutiny from conduct that reflects ordinary competition.
As Klein puts it:
One would not want to refer to the pervasive examples of price discrimination in the real world as implying that “market power” or “monopoly power” in any relevant economic or policy sense also is pervasive. Instead, all it means is that most firms in the marketplace possess some “individual pricing discretion.”
Why the Lerner Index Leads Courts Astray
What, then, should courts measure? Klein’s answer may look like a detour into measurement theory. It isn’t. If the Lerner Index points courts in the wrong direction, the entire framework for identifying market power needs a reset.
Klein engages directly with William Landes and Richard Posner’s “Market Power in Antitrust Cases” (1981). He doesn’t dispute the math. He disputes the target. Landes and Posner use market share as a proxy for a firm’s own elasticity of demand. Klein argues market share should capture something else entirely: a firm’s ability and incentive to restrict market-wide output.
That distinction does real work. A differentiated product has two components. One is firm-specific: brand loyalty, product features, and other factors that produce a downward-sloping demand curve and allow pricing above marginal cost. The other is market-wide: the firm’s place in a broader market where output interacts with rivals.
A firm can enjoy substantial pricing discretion over its own product and still lack any ability to move the broader market. A branded grocery company may face relatively inelastic demand from loyal customers. But if its market share is small and rival supply is elastic, it cannot raise market-wide prices. That firm has individual pricing discretion. It does not have market power in any sense antitrust law should recognize.
The same logic applies beyond antitrust. Klein illustrates the point with Reuben Kessel’s study of physician pricing. Kessel observed that doctors charged wealthier patients more and inferred that the American Medical Association enforced a cartel. Klein rejects that inference. Doctors price-discriminate because patients face switching costs—familiarity, continuity of care, reluctance to start over—that give each physician some pricing discretion over existing patients. But that says nothing about whether the profession can restrict output and raise market-wide prices.
The Court in Kodak was right to focus on lock-in and switching costs. It erred in treating a firm’s ability to exploit those features as antitrust market power.
From Kodak to Epic: Getting the Question Right
Klein’s distinction has started to surface in doctrine. In Epic Games v. Apple (2023), the 9th U.S. Circuit Court of Appeals articulated a four-part test for single-brand aftermarkets. Plaintiffs must show that aftermarket restrictions were not generally known at the time of purchase, that significant information costs prevented accurate life-cycle pricing, that meaningful switching costs exist, and that standard market-definition principles support the proposed market. If buyers knowingly accepted the terms, courts have no competition problem to solve. The Epic test stays within Kodak’s framework, but it gives operational form to Klein’s insight. The key question is not whether lock-in exists, but whether buyers understood it ex ante.
The framework, however, remains contested. In Surgical Instrument Service Co. v. Intuitive Surgical (No. 25-1372), the Federal Trade Commission (FTC) argued as amicus before the 9th Circuit that when a defendant already has market power in the foremarket, courts need not require proof of the Kodak lock-in factors to define the aftermarket. Klein signed an International Center for Law & Economics (ICLE) amicus brief opposing that position.
The FTC’s approach would apply the Kodak factors only when the foremarket is competitive—and dispense with them when the defendant already has foremarket power. Klein would see that as a familiar mistake: treating lock-in as a substitute for proof of antitrust market power. Retaining customers is not the same as raising market-wide prices.
When Theory Meets the Wrong Legal Question
I find the conceptual distinction persuasive. The empirical questions prove tougher. Klein’s price-discrimination account and the hold-up story both predict higher aftermarket prices for locked-in buyers. They diverge on why Kodak did it: to exploit locked-in customers, or to price efficiently across segments. The data alone rarely settles that dispute.
Klein writes in the Director-Levi tradition, which treats tying as efficient marketing more often than exclusion. Michael Whinston showed in 1990 that bundling can exclude rivals under certain conditions, particularly where firms can commit and deter entry. But Whinston’s model starts with foremarket power. Without it, there are no monopoly profits to protect through tying. On Klein’s account, Kodak lacked that power. Even when such theories apply, courts should still demand evidence of actual market-wide effects.
The real value of Klein’s paper is not its bottom-line conclusion. It is the lesson in translation. The Court in Kodak relied on sound economics but answered the wrong legal question. Klein understood both disciplines well enough to spot the mismatch.
That turns out to be the hard part. Economists often assume the challenge lies in getting the theory right. Klein shows the harder task is knowing which theory fits which legal question. The economics of hold-up was sound. The translation wasn’t.
Further Reading
- Benjamin Klein, “Market Power in Antitrust: Economic Analysis after Kodak,” Supreme Court Economic Review, Vol. 3 (1993)
- Benjamin Klein, Robert G. Crawford & Armen A. Alchian, “Vertical Integration, Appropriable Rents, and the Competitive Contracting Process,” Journal of Law and Economics, Vol. 21 (1978)
- William M. Landes & Richard A. Posner, “Market Power in Antitrust Cases,” Harvard Law Review, Vol. 94 (1981)
- Aaron Director & Edward H. Levi, “Law and the Future: Trade Regulation,” Northwestern University Law Review, Vol. 51 (1956)
- Benjamin Klein & Keith B. Leffler, “The Role of Market Forces in Assuring Contractual Performance,” Journal of Political Economy, Vol. 89 (1981)
- Michael D. Whinston, “Tying, Foreclosure, and Exclusion,” American Economic Review, Vol. 80 (1990)
