Tying arrangement

From CEOpedia
Tying arrangement
See also


Tying arrangement - is a sales practice in which a seller agrees to supply one product (the tying product) only on condition that the buyer also takes a second, distinct product (the tied product) from the same seller, or at least agrees not to buy it from anyone else. The US Supreme Court put the problem this way: "the essential characteristic of an invalid tying arrangement lies in the seller's exploitation of its control over the tying product to force the buyer into the purchase of a tied product that the buyer either did not want at all, or might have preferred to purchase elsewhere on different terms" (Jefferson Parish 1984)[1].

Ties are everywhere. A car comes with tyres, a phone ships with a dozen apps nobody asked for. Mostly harmless. Competition authorities start to worry only when the seller has real market power over the first product and uses it to win sales of the second that it could not win on the merits. That narrow band of cases produced some famous antitrust litigation, from salt-dissolving machines in the 1940s to Windows and Android.

Types of tying

  • Contractual tying - the condition is written into the contract. A machine lease obliges the lessee to buy all supplies from the lessor; a franchise agreement makes the franchisee buy ingredients from the franchisor (see franchising).
  • Technical tying - the tied product is physically or digitally built into the tying product, so the buyer cannot get one without the other. Think of a browser integrated into an operating system, or a printer whose chip rejects third-party cartridges. The design does the work.
  • Requirements tying - the buyer must purchase all of its needs for the tied product from the seller, usually a consumable used with durable equipment. Since quantities vary, the tied product becomes a meter of how intensively each customer uses the equipment.
  • Pure bundling - the products are sold only as a package, in fixed proportions.
  • Mixed bundling - the products are available separately and as a discounted package. Not tying in the strict sense, though a steep discount can have similar effects.

In tying the tying product is unavailable on its own, while the tied product may be sold separately; pure bundling is the special case with fixed proportions. Package pricing is covered in price bundling and bundled pricing; this article concentrates on the conditional sale and its legal treatment.

Economics of tying

Price discrimination through metering

The oldest explanation has little to do with conquering a second market. It's about measuring customers. When a machine is used with a consumable (salt, punch cards, ink), heavy users usually value the machine more than light users. The seller can let the consumable reveal that value: sell the machine cheaply, price the consumable well above cost, and heavy users pay more in total. Ward Bowman read many tying clauses as devices for counting how much each customer used the equipment[2], and Meyer Burstein worked out the profit logic of tie-in sales in 1960[3]. This is a form of price discrimination, and it only works if the buyer can't get cheaper consumables elsewhere. That's what the tie is for.

Leverage theory and the Chicago critique

For decades courts saw it differently. The "leverage" theory held that a firm with a monopoly in one product could use tying to create a second monopoly in the tied product and collect two monopoly profits instead of one.

Economists associated with the University of Chicago (Aaron Director, Ward Bowman, later Robert Bork and Richard Posner) attacked this with the "single monopoly profit" argument. If customers use the monopolised product together with a competitively supplied one in fixed proportions, they care only about the total price of the pair. The monopolist can already extract everything through the price of its own product; every extra dollar charged for the tied good must be given back on the tying good, or customers walk away. Bowman concluded that where tying is profitable, the explanation usually lies in price discrimination or cost savings, not in the creation of a second monopoly[2].

Powerful critique. It also rested on assumptions: fixed proportions, a competitive tied market, no scale economies, a static world.

Strategic foreclosure: Whinston (1990)

Michael Whinston showed what happens when some of those assumptions are relaxed[4]. In his model a monopolist in market A faces a rival in market B, where there are economies of scale. If the monopolist commits to tying (technical integration makes the commitment credible), it competes harder for every B sale, since losing a B customer now means losing the A sale too. The rival, left with a smaller market, may not cover its fixed costs and exits or never enters. Tying becomes a profitable exclusion strategy.

Whinston stressed the limits. When the tied good is used only with the monopolised good, the foreclosure incentive is weaker, and the welfare effects of tying are ambiguous in any case. Later work showed how excluding a complementary rival can protect the primary monopoly itself, an argument central to the Microsoft litigation. The policy upshot? Tying is neither always harmless nor always harmful, which pushed economists toward case-by-case analysis.

Tying in United States law

Section 1 of the Sherman Act (1890) prohibits contracts in restraint of trade. Section 3 of the Clayton Act (1914) makes it unlawful to sell or lease goods "on the condition, agreement, or understanding that the lessee or purchaser thereof shall not use or deal in the goods ... of a competitor" where the effect "may be to substantially lessen competition or tend to create a monopoly in any line of commerce". Ties involving services fall under the Sherman Act only.

Leading cases

Case Year Tying product Tied product Outcome
International Salt Co. v. United States 1947 Patented salt machines (Lixator, Saltomat) Salt and salt tablets Tie unlawful; summary judgment for the government
Jefferson Parish Hospital District No. 2 v. Hyde 1984 Hospital services Anaesthesiology services No per se violation; insufficient market power
Eastman Kodak Co. v. Image Technical Services 1992 Replacement parts for Kodak equipment Repair service Claim allowed to go to trial; aftermarket power possible
United States v. Microsoft (D.C. Cir., en banc) 2001 Windows Internet Explorer Per se ruling vacated; remand under rule of reason
Illinois Tool Works v. Independent Ink 2006 Patented printhead systems Ink No presumption of market power from a patent

International Salt (1947). The country's largest producer of industrial salt leased its patented machines on condition that lessees buy all salt used in them from the company. In 1944 it sold about 119,000 tons of salt, worth roughly $500,000, for those machines. The Court held that "it is unreasonable, per se, to foreclose competitors from any substantial market"[5]. It rejected the quality-control defence: the company could have set specifications any supplier had to meet instead of naming itself as sole vendor.

Jefferson Parish (1984). A hospital's exclusive contract with one anaesthesiology firm meant its surgical patients had to use that firm. Dr. Hyde, an excluded anaesthesiologist, sued. The Court kept the "per se" label but added conditions: two separate products (judged by separate consumer demand), forcing, and sufficient market power in the tying product. Around 30% of parish residents used the hospital, which the Court found too little to justify per se condemnation[1]. Four concurring justices, led by Justice O'Connor, wanted to drop the per se rule altogether.

Eastman Kodak (1992). Kodak stopped selling parts to independent service firms, in effect tying parts to its own repair service. It argued that without power in the equipment market it couldn't exploit aftermarket customers. The Court disagreed, at least at summary judgment. Evidence that Kodak controlled nearly 100% of the parts market and 80-95% of the service market was enough to proceed, and the Court pointed to "significant information and switching costs" that weaken the link between aftermarket prices and equipment sales[6].

Microsoft (2001). The en banc D.C. Circuit upheld much of the Section 2 monopoly maintenance case but vacated the District Court's per se finding that Internet Explorer was illegally tied to Windows. Platform software was a novel category, the court said, and "simplistic application of per se tying rules carries a serious risk of harm"[7]. Plaintiffs could pursue the claim under the rule of reason on remand; the government chose not to. The court confined its reasoning to platform software.

Illinois Tool Works (2006). Older cases had presumed that a patent on the tying product confers market power. The Supreme Court abandoned that presumption: "the mere fact that a tying product is patented does not support such a presumption"[8]. Market power must now be proved.

Tying in European Union law

Article 102 TFEU (formerly Article 82 EC) prohibits abuse of a dominant position; point (d) lists "making the conclusion of contracts subject to acceptance by the other parties of supplementary obligations which, by their nature or according to commercial usage, have no connection with the subject of such contracts". Before the software era the EU courts had condemned ties in Hilti (nail guns, cartridge strips and nails) and Tetra Pak II (carton-filling machines and cartons).

Microsoft (2004 and 2007)

On 24 March 2004 the European Commission found that Microsoft had abused its near monopoly in PC operating systems in two ways: by restricting interoperability with rival work group servers and "by tying its Windows Media Player (WMP), a product where it faced competition, with its ubiquitous Windows operating system". It imposed a fine of €497.2 million and ordered Microsoft to offer PC makers a version of Windows without the media player[9]. The Court of First Instance (now the General Court) upheld the tying finding on 17 September 2007 in case T-201/04. The judgment set out what became the standard EU test: the tying and tied products are two separate products; the undertaking is dominant in the tying market; customers cannot obtain the tying product without the tied product; and the practice forecloses competition.

Google Android (2018-2026)

On 18 July 2018 the Commission fined Google €4.34 billion for three restrictions imposed on device makers and mobile operators. The tying element: Google "required manufacturers to pre-install the Google Search app and browser app (Chrome), as a condition for licensing Google's app store (the Play Store)". The others were payments for exclusive pre-installation of Google Search and a ban on selling devices running non-approved Android forks[10]. On 14 September 2022 the General Court (case T-604/18) largely upheld the decision, annulled the part concerning the revenue-share payments and set the fine at €4.125 billion. On 2 July 2026 the Court of Justice dismissed Google's further appeal (case C-738/22 P), so that fine stands.

Conditions for illegality

The wording differs between legal systems; the questions are roughly the same:

  • Two separate products. Is there distinct demand for the tied product? Courts ask whether customers buy the items separately and whether firms without market power sell them separately. Shoes and shoelaces: one product. Hospital services and anaesthesia: two, according to Jefferson Parish.
  • Market power or dominance in the tying product. Without it, buyers can go elsewhere and the tie cannot force anything. In Jefferson Parish a market share of about 30% was not enough for per se treatment; EU law requires dominance.
  • Coercion. The buyer must be unable to obtain the tying product without the tied one, or must face terms so unattractive that separate purchase is not a realistic option.
  • Foreclosure or anticompetitive effect. The tie must affect a not insignificant volume of commerce or harm competition in the tied market. Both US and EU law let the defendant argue that efficiencies outweigh the harm.

Example: metering through printer cartridges

The numbers below are hypothetical and chosen to illustrate the mechanism.

A printer costs $50 to make. Cartridges cost $5, for the firm and for independent suppliers alike. Every cartridge printed is worth $25 to a customer. Over the printer's life, 3,000 light users will use 4 cartridges each and 1,000 heavy users 20 each. The firm cannot tell them apart.

Without a tie, cartridges sell at the competitive $5, so owning the printer is worth 4 × (25 − 5) = $80 to a light user and 20 × (25 − 5) = $400 to a heavy one. Only the printer price can be set.

Strategy Printer price Cartridge price Buyers Firm profit Consumer surplus Total surplus
A. No tie, low printer price $80 $5 4,000 $120,000 $320,000 $440,000
B. No tie, high printer price $400 $5 1,000 (heavy only) $350,000 $0 $350,000
C. Tie (requirements), metering $4 $24 4,000 $424,000 $16,000 $440,000

Strategy A earns (80 − 50) × 4,000 = $120,000. B earns (400 − 50) × 1,000 = $350,000, but light users drop out. C sells the printer below cost and earns on cartridges. Each light user pays $4 + 4 × $24 = $100, exactly what printing is worth to them. Each heavy user pays $4 + 20 × $24 = $484. Profit is (4 − 50) × 4,000 + (24 − 5) × 32,000 cartridges = −184,000 + 608,000 = $424,000. Heavy users keep 20 × (25 − 24) − 4 = $16 each, or $16,000 in total.

Three things stand out. Tying earns 21% more than the best untied strategy ($424,000 against $350,000). Total surplus under C equals A and beats B, because metering keeps light users in the market. Yet consumers lose: their surplus falls from $320,000 under A to $16,000. And the scheme collapses once $5 third-party cartridges appear, since everyone would buy the $4 printer and refill cheaply, leaving the firm $46 down per printer. That explains the fight against compatible cartridges through contracts, chips and warranty terms. Legality then turns on market power and the elasticity of demand for equipment, the Kodak question.

Advantages of tying

Not every tie is a scheme:

  • Lower costs. Selling two items together can be cheaper than selling them separately, with one transaction instead of two.
  • Quality control and reputation. A machine may fail with poor consumables, and the customer blames the machine maker. Franchisors argue that requiring specific inputs protects the brand. Courts accept this mainly when a specification would not do equally well.
  • Convenience. Customers get a working system without assembling it. That was Microsoft's argument about browser integration.
  • Wider access. Metering lowers the entry price and brings in buyers who would never pay a lot up front.

Limitations and risks

  • Foreclosure of rivals. Under Whinston's conditions, tying can push efficient competitors out of the tied market and raise barriers to entry in both markets.
  • Reduced consumer choice. Buyers who prefer another tied product pay twice or settle for the inferior one; consumer protection rules may step in too.
  • Exploitation of locked-in customers. Switching costs and poor information let the seller raise aftermarket prices after the sale, as in Kodak.
  • Weaker innovation in the tied market. With distribution closed by default, rivals invest less, as the Commission argued in Microsoft and Android.
  • Ambiguous welfare effects of metering. It may raise output or merely shift surplus from consumers to the firm. An empirical question.
  • Legal uncertainty. What counts as "separate products" shifts with technology, and the line between product design and technical tying is hard to draw. Dominant firms face real litigation risk and large fines.

Related approaches

Several practices sit nearby. Price bundling and bundled pricing deal with packages and discounts rather than conditions. Exclusive dealing forbids the buyer from dealing with rivals without requiring the purchase of a second product. Product line pricing covers the broader decision of how to price related items, including the "razor and blades" approach of cheap equipment and expensive refills. Resale price maintenance and vertical integration are other ways to control downstream markets. In platforms, network effects amplify a tie, because pre-installation hands the tied product a user base rivals struggle to match.

References

Footnotes

  1. ↑ Jefferson Parish Hospital District No. 2 v. Hyde, 466 U.S. 2, Supreme Court of the United States, 1984
  2. ↑ Bowman W.S., Tying Arrangements and the Leverage Problem, The Yale Law Journal, 1957, Vol. 67, p.19
  3. ↑ Burstein M.L., The Economics of Tie-In Sales, The Review of Economics and Statistics, 1960, Vol. 42, p.68
  4. ↑ Whinston M.D., Tying, Foreclosure, and Exclusion, American Economic Review, 1990, Vol. 80, No. 4
  5. ↑ International Salt Co. v. United States, 332 U.S. 392, Supreme Court of the United States, 1947, LII text
  6. ↑ Eastman Kodak Co. v. Image Technical Services, Inc., 504 U.S. 451, Supreme Court of the United States, 1992, LII text
  7. ↑ United States v. Microsoft Corp., 253 F.3d 34, D.C. Circuit (en banc), decided 28 June 2001, section IV
  8. ↑ Illinois Tool Works Inc. v. Independent Ink, Inc., 547 U.S. 28, Supreme Court of the United States, 2006, LII text
  9. ↑ European Commission, Press release IP/04/382, Brussels, 24 March 2004
  10. ↑ European Commission, Press release IP/18/4581, Brussels, 18 July 2018
Tying arrangement — recommended articles
Price bundling — Bundled pricing — Price discrimination — Market power — Monopoly — Unfair competition — Barriers to entry
Retrieved from CEOpedia