ICLE Amicus to U.S. Supreme Court in Apple v Epic Games
Interest of Amicus Curiae[1]
The International Center for Law & Economics (ICLE) is a nonprofit research organization that applies economic analysis to legal questions, with particular expertise in competition, intellectual property, and digital markets. ICLE has participated as amicus in earlier stages of this litigation, including the original appeal, the ensuing rehearing and certiorari proceedings, and the appeal from the district court’s contempt order. Its scholars have written extensively about platform competition, investment incentives, and the consequences of judicial error in evaluating business conduct.
Summary of Argument
The 2021 injunction required Apple to allow developers to communicate with customers about alternative purchasing mechanisms, including through links within their apps. It established no commission rate or method of calculating one. Pet. App. 166a–167a. The Ninth Circuit nonetheless found a limit on “prohibitive” commissions in the injunction’s spirit, even though “the text of the Injunction does not address commissions at all.” Pet. App. 34a n.9.
The injunction rested on a finding of informational harm to consumers—that Apple’s anti-steering rules prevented consumers from making informed choices because they were left ignorant of non-Apple purchasing mechanisms. Apple’s revised guidelines alleviated that harm by allowing developers to inform users within the App Store of external payment alternatives. Apple also imposed a commission on those external payments.
The contempt order then treated the commercial attractiveness to developers of external purchasing as the measure of whether the informational harm had been remedied. A commission can affect developers’ willingness to offer links and thus the information consumers receive. But its effect on developers’ incentives does not, by itself, establish whether a lower commission would improve competition and consumer welfare. The injunction left Apple discretion to determine compensation for the services it continued to provide, and a commission did not violate the order merely because a lower rate would have been preferred by some developers.
A business required to change its conduct must still decide how to run its business. When an injunction leaves several permissible ways to comply, choosing a compliance mechanism that also preserves the most revenue is still compliance. The firm need not prefer an alternative because it is less profitable, and its decision to compare the financial consequences of available choices is entirely ordinary. The contrary approach would make the sacrifice of revenue a test of obedience even where an injunction imposed no such requirement.
Apple faced a practical question in determining the exact terms of compliance: how to collect compensation for services that developers continued to use when a user followed a link-out and paid elsewhere. Preserving that compensation is not inconsistent with the obligation to enable developer communications, and, indeed, doing so furthers the competition the obligation was intended to promote.
The underlying California Unfair Competition Law (UCL) judgment makes the competitive effects of commissions particularly relevant. The UCL reaches conduct beyond completed antitrust violations, but the standards applied in this case remain concerned with harm to competition and the benefits of the challenged conduct. A court invoking the competitive purpose of an injunction must account for the procompetitive effects of the defendant’s conduct, including its role in sustaining valuable services. Because those services require financial backing, it does not further competition simply to equate a larger financial advantage for developers with a better competitive outcome.
An external payment changes who processes a transaction. It does not eliminate the software development, distribution, and access to users that helped make the transaction possible. Commissions provide a way to finance the development, provision, and ongoing improvement of those services while allowing developers to enter with low initial charges and pay more as they succeed. They can also prevent developers from free riding by avoiding payment for valuable platform services simply by moving the final purchasing step elsewhere. Consideration of these functions is essential to evaluating a remedy intended to promote competition.
In finding Apple’s commission prohibitive, the courts below relied heavily on the conclusion that developers’ outside transaction costs exceeded the alleged three-percentage-point difference between Apple’s in-app and link-out commissions. Pet. App. 20a, 136a. That comparison may be relevant to assessing the competitive effects of a chosen commission, but it does not establish the value of Apple’s contribution. Nor does it capture the full range of reasons other than short-term cost that a developer might choose another purchasing channel. For these reasons, increasing use of that alternate channel will not necessarily promote competition. A rule that encourages migration by limiting compensation for services still supplied can reduce the resources and incentives to provide those services, ultimately harming competition and consumer welfare.
The resulting rate-setting problem is not trivial, and it necessitates exactly the sort of internal commercial deliberations that the district court treated as evidence of contempt. The Ninth Circuit rejected a permanent prohibition on commissions but contemplated a court-approved fee based on selected costs and uses of intellectual property. Pet. App. 34a–36a. Determining that fee requires decisions about common costs, continuing investment, and the allocation of charges across different users. A general instruction to promote competition answers none of those questions.
This Court has cautioned against imposing duties that require judges to administer complex commercial terms and that discourage beneficial conduct. See, e.g., Verizon Commc’ns Inc. v. Law Offices of Curtis v. Trinko, LLP, 540 U.S. 398, 414–15 (2004); Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 223 (1993). Those concerns have particular force when contempt supplies the means of imposing a pricing obligation absent from the injunction. The questions the injunction left unanswered create at least a fair ground of doubt about whether it barred Apple’s commission, which rules out civil contempt. Taggart v. Lorenzen, 587 U.S. 554, 561 (2019).
Argument
I. The Injunction Sought to Remedy Informational Harm. It Said Nothing About Commissions
The 2021 injunction rests on a violation of California’s Unfair Competition Law predicated on informational harm to consumers. Apple’s anti-steering rules prevented users from learning that purchasing options existed outside in-app purchase. See Pet. App. 34a n.9. Apple’s revised guidelines removed the source of that harm by permitting developers to tell users—through buttons, links, and calls to action—that other options exist. Once that information reaches consumers, the injunction has accomplished what it set out to do.
A. Apple’s compliance measures must be assessed as a whole
Apple did not simply replace a 30% commission with a 27% commission. Rather, Apple adopted a multi-pronged approach in response to the district court’s initial order and injunction which makes the repeated assertion of a 27% rate factually inaccurate.
First, Apple dropped the anti-steering provisions that the court had found to violate the UCL, so that links to external (non-Apple) purchasing mechanisms would be permitted. Apple also provided a mechanism by which approved developers could “include within their apps buttons or links with calls to action directing users to out-of-app purchasing mechanisms other than IAP.” JA506–507. That, in itself, lowers information costs for developers who wish to alert users to external payment options, for the users themselves, and, by extension, for external payment sites.
Second, Apple adopted a varied schedule of commissions for off-app purchases facilitated by that mechanism. For linked off-app purchases made more than “seven days after a user taps on an External Purchase Link and continues . . . to an external website,” JA519, Apple would charge no commission. For linked off-app purchases made within seven days, Apple would charge either 12% (for developers participating in the App Store Small Business Program) or 27% (for larger participating developers). JA518.[2] The program also specified a 12% rate for qualifying transactions otherwise subject to a 15% in-app commission. JA518.
Third, Apple reiterated recently adopted provisions under a separate settlement that permitted developers to send out-of-app communications to their users about alternative purchasing methods. JA521. These provisions permit developers to request basic contact information from their app users and to “send communications outside of the app to their user base about purchasing methods other than in app purchase.” JA504. Apple’s notice further explained that developers could use those communications to direct users to their websites without incurring a link-out commission at all. JA520–21.
Apple also provided a procompetitive rationale for its multi-pronged approach—one that neither the district court nor the court of appeals refutes—arguing that its approach balances “(1) Apple’s entitlement to a commission for purchases facilitated through the App Store, even if the user waits a small amount of time after leaving the App Store to actually finalize the purchase, with (2) the fact that the longer the time between the link out and the purchase, the more attenuated the connection between the purchase and Apple’s facilitation of the purchase.” JA519–520.
Further, while Apple did adopt a 27% commission on a subset of linked purchases made within one week of linking via apps on the App Store, it is not correct to characterize its compliance as imposing “a 27 percent commission . . . on off-app purchases, where it had previously charged nothing.” Pet. App. 51a. Rather, the 27% commission applied to only a subset of the previously barred purchases, and the effective commission rate was considerably lower.[3]
Moreover, Apple permitted previously barred in-app links to external purchasing mechanisms, with commissions of 0%, 12%, or 27% commissions depending on the timing and type of purchase and the developer’s eligibility for reduced rates. JA518–20. The prior bar on those links was a clear and categorical prohibition tantamount to an infinite price for using that route—it was not a zero-percent commission.
Thus Apple did not implement a 27% charge on transactions for which it previously charged nothing; rather, it opened a previously prohibited purchasing channel, with a maximum commission of 27% and an effective rate it expected to be lower.
B. Developer savings alone do not establish consumer benefit
Apple’s Notice of Compliance should be assessed as a whole, considering both its stated rationale and the informational harm the injunction was designed to remedy. JA500–522. The Ninth Circuit’s question—”whether charging a 27% commission has a prohibitive effect, in violation of the Injunction,” Pet. App. 20a—cannot be answered by isolating that rate from the other terms Apple adopted.
The expanded opportunities to communicate and the varied commission rates cannot simply be presumed to undermine the injunction’s informational objective. The 27 percent figure was the highest rate in a schedule that also included lower rates and commission-free purchases. Apple could reasonably expect those features to produce a lower effective rate. JA518–20. Treating the headline rate as controlling fails to address the Ninth Circuit’s own admonition that “[t]he district court must be mindful of the threshold upon which commissions become ‘prohibitive.’” Pet. App. 34a n.9.
Whether Apple’s response “thwarted the Injunction’s goals, and continued its anticompetitive conduct,” Pet. App. 51a, must also be assessed against the injunction’s express terms and the details of Apple’s approach. But the contempt order treated the commercial attractiveness of external purchasing to developers as the full measure of whether the informational harm had been remedied. It reasoned that Apple’s commission left external purchasing so financially unattractive that developers would decline to offer it, leaving consumers with a right to information but nothing to be informed about. Pet. App. 69a–87a.
A developer’s reluctance to offer an in-app link because of the commission can affect how readily consumers learn about an alternative. But a link both communicates an alternative and provides a direct route to a purchase; it is not the only way to inform users that purchasing is available through the developer’s website. Apple also permitted out-of-app communications directing users to developers’ websites without triggering a link-out commission. JA520–21.
The Ninth Circuit properly distinguished the mere existence of a commission from its effect on developers’ willingness to offer links. Pet. App. 34a n.9. In the courts’ analysis, however, developer savings became a proxy for consumer benefit. A lower commission may encourage more linking, but that alone does not establish that it benefits consumers when the commission also finances services they continue to use.
Recasting the injunction’s informational premise as a guarantee of a commercially attractive alternative does not enforce it. It sets a new commercial term the injunction never fixed, essentially modifying the injunction without proper process.
II. Choosing a profitable way to comply is still compliance
An injunction that does not enjoin a company from doing business entirely implicitly contemplates that it will undertake activities necessary to ensure its continued operation so long as they do not conflict with the terms of the injunction. This is especially true for the injunction at issue here, which is premised on increasing competition by permitting developers to direct users to outside purchasing options while still maintaining Apple’s platform, including its In-App Purchasing (IAP) as a viable alternative.
Incorporating a new operating procedure—here, enabling developers to direct customers to alternate payment mechanisms, Pet. App. 166a–167a—may entail any number of adjustments to other activities not addressed by the injunction and necessary to ensure effective continued operations. Managers implementing compliance with an injunction have two related responsibilities: identify the changes the order requires and decide how best to make those changes while preserving the quality and profitability of the business as a whole. Both are constituent parts of compliance.
This is especially true for managers of firms, who owe a fiduciary duty to exercise their considered business judgment to act in the corporation’s best interests. See Commodity Futures Trading Comm’n v. Weintraub, 471 U.S. 343, 348–49 (1985). And it is no less true when exercising their business judgment in choosing among lawful means of complying with a court order. See Schenker v. E. I. du Pont de Nemours & Co., 329 F.2d 77, 78–79 (2d Cir. 1964) (upholding directors’ chosen method of implementing a court-ordered divestiture as “a reasonable exercise of business judgment”). The business must continue to assess costs, respond to customers, and earn revenue. Financial analysis does not become evidence of evasion simply because, while determining how to comply with an injunction, the firm is also attempting to avoid a loss.
An injunction must identify required and forbidden conduct with specificity and reasonable detail, which necessarily leaves discretion over implementation on various dimensions. See Fed. R. Civ. P. 65(d)(1); Schmidt v. Lessard, 414 U.S. 473, 476 (1974) (“[Rule 65(d)] was designed to prevent uncertainty and confusion on the part of those faced with injunctive orders, and to avoid the possible founding of a contempt citation on a decree too vague to be understood.”). The firm knows more than the court about its operations, and it can use that knowledge to identify less costly ways to comply. Preserving that discretion allows the injunction to achieve its stated objective without unnecessary losses to the business and the customers it serves.
Where several alternatives may satisfy the same obligation, managers must assess the full consequences of each form of compliance. The underlying court-ordered obligation may affect how rivals compete and how users behave, and the commercial and operational consequences of those changes may require numerous adjustments throughout the defendant’s business that go well beyond the text of the injunction.
It was thus predictable that compliance would require internal deliberations over the precise mechanism to employ and the follow-on adjustments that mechanism might entail. These included, e.g., how to collect commissions on purchases made outside Apple’s payment system, Pet. App. 73a; whether to change the amount or basis of its charges, Pet. App. 78a; how to adapt its commission discounts and partner programs to the new payment arrangements, 3-ER-433; how to allocate responsibility for subscription management and refunds, JA509, JA511; how to revise its communications with users, Pet. App. 145a; and how to manage developer accounting, audits, interest charges, and payment offsets, JA520. Its implementation discussions also addressed the software changes, developer materials, and review procedures needed to meet the anticipated compliance deadline. 3-ER-443.
Indeed, Apple’s planning documents illustrate the work of translating the injunction into practice. One presentation treated persistent, discoverable links and in-app price comparisons as requirements, while identifying placement, frequency, design, and pricing language as matters still to be resolved. 3-ER-385, 387. The injunction did not specify those details, Pet. App. 166a–167a; rather, Apple had to determine how to give effect to the order on those dimensions and incorporate the resulting changes into its existing services.
The necessity of such considerations was, in fact, even contemplated by the district court. In its order denying Apple’s request for a stay of the injunction, the court acknowledged that “the injunction may require additional engineering or guidelines,” Order Denying Apple’s Motion to Stay Injunction Pending Appeal at 3, Epic Games, Inc. v. Apple Inc., No. 4:20-cv-05640-YGR (N.D. Cal. Nov. 9, 2021) (“Stay Order”), and that it could “envision numerous avenues for Apple to comply with the injunction,” specifically noting that Apple could “take steps to protect users, to the extent that Apple genuinely believes that external links would create issues.” Id. Tellingly, the court added that it was not “here to micromanage.” Id.
Consider Apple’s specification of lower commissions for small developers and qualifying subscription renewals as part of its compliance plan. JA518. Having elected to collect commissions on linked-out purchases, Apple also had to decide how its existing discounts would apply to those transactions. The injunction did not prescribe that treatment. For a profit-making enterprise obligated to provide a return to its shareholders, that decision inevitably included an evaluation of the revenue consequences of the various possible arrangements. Considering those consequences was thus an inherent function of compliance, not an indication of noncompliance.
Permitting in-app links to outside purchases also necessitated a reconsideration and revision of how to convey Apple’s resulting responsibilities and users’ risks when users opt to leave the App Store to make external payments. Thus, Tim Cook asked the compliance team to revise the customer notice to reflect that “Apple’s privacy and security standards do not apply to purchases made on the web.” Pet. App. 145a. Enabling those links required Apple to decide how to explain the consequences of outside payments to users, even though the injunction did not specifically address that issue.
In considering different commission rates, Apple also assessed how its pricing rationale would compare with “existing discounting approaches in other jurisdictions” and whether it could explain those rates “without implicitly devaluing our IP / proprietary technology.” Pet. App. 81a. Implementing the injunction had consequences both for Apple’s day-to-day operations as well as for its pricing and commercial relationships elsewhere, and incorporation of those effects was a function of compliance with the order, not of its evasion.
Because the injunction left compensation available, Apple’s decision to preserve it cannot itself show that the firm frustrated the order. Nor does the existence of a cheaper option for developers establish that the more remunerative choice was impermissible. That inference would assume the very pricing obligation in dispute: that Apple had to forgo otherwise available compensation to make the external alternative more attractive. As long as Apple was still providing services, it had to decide whether and how to collect compensation for them through other means. The injunction specified the communications Apple must permit; it did not establish a price for the platform services that remained in use. Pet. App. 166a–167a. Evaluating the revenue consequences of alternative collection arrangements was a necessary and predictable part of implementing the order.
The Ninth Circuit expressly stated that the district court had not found contempt merely because Apple selected the most advantageous option for its business. Pet. App. 16a. That distinction must have practical consequences for how a court evaluates commercial deliberations and their resulting decisions. Discussions about protecting revenue, calculating the cost of different options, and retaining a return on services supplied are consistent with ordinary efforts to comply. Managers can plan around an identified obligation and choose among ways of satisfying it. The possibility that another arrangement would leave more money with developers supplies no additional standard for determining what compliance requires.
This Court has recognized that “[e]conomic hardship can influence choice only as among two or more effective remedies.” United States v. E. I. du Pont de Nemours & Co., 366 U.S. 316, 327 (1961). That holding arose in the context of the Court’s rejection of an ineffective remedy, chosen by a court to avoid severe economic consequences for the defendant. What the Court rejects is the ineffectiveness, not the economic consideration, and the same distinction is relevant to managers implementing a presumptively effective remedy: Economic considerations may inform the choice among compliant alternatives. Once the firm has identified permissible alternatives, it has every reason to select the one that works best commercially.
Attention to commercial self-interest does not conflict with the competitive purpose of the court’s remedy. Indeed, it furthers it. The prospect of earning revenue encourages firms to compete by improving their products, attracting customers, and sustaining services that others value. See Trinko, 540 U.S. at 407–08. The injunction left Apple free to compete for users’ business alongside outside purchasing options. Requiring it to disregard costs and returns when choosing among compliant arrangements would weaken its incentive to improve the services on which that competition depends.
III. Apple’s Commissions can advance the UCL’s purpose of protecting competition
The competitive purpose of this injunction must be understood in light of the UCL standards that produced it. The judgment addressed restrictions on developers’ ability to inform users about alternative purchasing options. The district court found that those restrictions impaired informed choice and competition in the relevant market, and it ordered Apple to permit the specified communications. JA373–79, JA403–05. Apple’s response to the injunction complied with its express terms and addressed the underlying competitive harm identified in the decision.
The UCL reaches conduct that threatens an incipient antitrust violation or otherwise significantly harms competition. JA487 (citing the Supreme Court of California in Cel-Tech Commc’ns, Inc. v. L.A. Cellular Tel. Co., 20 Cal. 4th 163, 186–87 (1999)). It also reaches “unfairness to consumers . . . us[ing] the balancing test, which ‘weigh[s] the utility of the defendant’s conduct against the gravity of the harm to the alleged victim.’” JA487 (quoting Progressive W. Ins. Co. v. Super. Ct., 135 Cal. App. 4th 263, 285 (2005)).
The Ninth Circuit applied both standards here, explaining that they “are not mutually exclusive.” JA487. Under either approach, a conduct’s benefits to competition and consumers are just as much a part of its competitive effects as are its costs. As the Supreme Court of California noted, when fashioning equitable remedies under the UCL, the court must consider “the equities on both sides of the dispute.” Cortez v. Purolator Air Filtration Prods. Co., 23 Cal. 4th 163, 180 (2000). To ask only whether Apple could do more to favor developers would truncate the competitive inquiry that justified the original judgment and injunction.
A. Commissions finance valuable services and encourage continued investment in them
Apple’s IAP processes payments and collects compensation for the broader services Apple supplies. The Ninth Circuit described Apple’s developer agreement as providing tools and “access to Apple’s vast consumer base” in exchange for “a $99 fee and an ongoing 30% commission on developers’ iOS revenue.” JA418. Payment processing is one of those tools. But moving payment processing to another provider does not eliminate developers’ access to consumers and use of other tools, nor does it obviate the need for an annual fee and continuing commission.
In the absence of a commission, a developer could, for example, make an app available for download without charge and earn revenue later through purchases of content used in the app, effectively cutting off Apple’s mechanism for compensation for distribution, software, security, and access. The district court recognized this distinction and explained that Apple could collect a commission even without requiring use of IAP, although another collection method could be more costly to administer. JA347–48 & n.617. The distinction between payment processing and other attributes of Apple’s system does not disappear when a customer follows a link to complete a purchase outside Apple’s App Store. The app still uses the platform, and the developer may still have acquired the customer through Apple’s distribution services. Ensuring Apple receives compensation for those contributions does not become economically baseless when a different firm happens to handle the final payment.
In addressing Apple’s app-distribution and in-app-payment-processing restrictions under the UCL, the district court noted that “Apple has proffered mostly valid and non-pretextual procompetitive justifications. To a large extent that makes the conduct more than ‘not anticompetitive’ but potentially beneficial to consumers.” JA372–73. Specifically, the court found that Apple’s app distribution restrictions “increase security in the ‘broad’ sense by allowing Apple to filter fraud, objectionable content, and piracy during app review while imposing heightened requirements for privacy.” JA270. The court also recognized that “security and privacy have remained a competitive differentiator for Apple,” JA271, and noted that “[t]he witnesses are unanimous that user security and privacy are valid procompetitive justifications.” JA272.
The court also agreed “with the general proposition that Apple is entitled to be paid for its intellectual property,” and that Apple “invests enormous sums into developing new tools and features for iOS.” JA277. It also recognized that “[b]oth Apple and third-party developers like Epic Games have symbiotically benefited from the ever-increasing innovation and growth in the iOS ecosystem.” JA87.
Having observed that the evidence at trial did not resolve the question of the extent to which Apple’s investments in its ecosystem were specific to the App Store “with precision,” the court strained to qualify that lack of precision, finding that “the record is devoid of evidence that Apple set its 30% commission rate as a calculation related to the value of its intellectual property rights.” JA278. More curious still, the court therefore found “that with respect to the 30% commission rate specifically, Apple’s arguments are pretextual, but not to the exclusion of some measure of compensation.” JA278. That was, at least, a recognition that the UCL does not require a commission rate of zero.
The court nonetheless discounted the connection in its initial judgment, reasoning that “[a] UCL action is equitable in nature,” and that “[c]ourts have ‘broad discretion’ to fashion equitable remedies to serve the needs of justice.” JA373 (internal citations omitted).
Whatever mechanism the injunction leaves available, the underlying economic problem does not disappear once anti-steering provisions are removed. This Court has recognized that vertical price restraints can properly be used to prevent free riding. Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 890–91 (2007) (explaining that absent vertical price restraints, discounting retailers can “free ride on retailers who furnish services” such as showrooms and demonstrations, and finding that it is proper to impose restraints that alleviate the problem by encouraging retailers to “invest in tangible or intangible services or promotional efforts that aid the manufacturer’s position as against rival manufacturers”).
Left unaddressed, free riding weakens the incentive to supply the services that made the underlying transaction possible in the first place and to keep investing in them. Id. Courts have recognized that vertical restraints—including anti-steering rules that restrict point-of-sale merchant steering—can curb negative externalities by preventing retailers from free-riding on network investments. See Ohio v. Am. Express Co., 585 U.S. 529, 547 (2018) (quoting Leegin, 551 U.S. at 890–91, for the proposition that “vertical restraints can prevent retailers from free riding and thus increase the availability of ‘tangible or intangible services or promotional efforts’ that enhance competition and consumer welfare”). The injunction here forecloses that particular tool. It does not, however, eliminate the economic problem a restriction on steering might otherwise have addressed.
A developer that uses Apple’s distribution, discovery, curation, and security-review services to acquire and retain users and then completes the resulting sale through a channel that contributes nothing toward the cost of those services captures the value of Apple’s investment without paying for it. This is the same dynamic that concerned this Court in Leegin. A commission on such purchases addresses that problem by collecting compensation when the developer earns revenue from using Apple’s services. Its effect on developers’ willingness to offer links must be assessed together with its role in financing those services. If anything, that approach is less restrictive of developer and consumer conduct than the vertical restraint this Court sustained in Leegin, because it regulates only compensation for services actually used, not communication about, or access to, an alternative purchasing channel.
The allocation of those charges also matters to competition. As this Court explained in Am. Express, prices on a two-sided platform reflect the interdependence of its users; higher merchant fees alone do not establish anticompetitive effects. 585 U.S. at 535–37. Commissions on successful transactions can help finance services available to developers who generate little or no revenue, supporting low initial charges that encourage entry and experimentation. Reducing those commissions may benefit the developers who pay them, but it can also shift charges to other participants or weaken incentives to maintain and improve the shared services.
The competitive inquiry must include Apple’s ability to continue competing. Developers benefit from improvements in the tools, distribution services, and customer experience that make a platform useful. Rules that reduce the return from supplying those services can weaken the incentive to improve them and shift their costs to other participants. A remedy aimed at competition must consider those consequences alongside the immediate benefit a developer receives from paying a lower fee.
B. Competition does not require every alternative to offer the same financial advantage
That inquiry also distinguishes permission to communicate from a guarantee about the commercial results of communication. Better information can help customers compare available offers and encourage sellers to improve them. The resulting competition may change prices, quality, and the division of revenue among businesses. Its value does not depend on ensuring that every purchasing channel attracts transactions or that the platform providing access to customers receives less compensation.
As noted, the Ninth Circuit’s commission analysis concentrated on the difference between Apple’s 30 percent in-app commission and its maximum 27 percent external-purchase commission, together with developers’ additional costs of completing a purchase elsewhere. It reasoned that those outside costs exceeded the difference and make the external alternative economically nonviable. Pet. App. 19a–21a.
That comparison focuses on only one source of commercial attraction (lower transaction charges), but developers may prefer a particular purchasing mechanism for other reasons. As the Ninth Circuit recognized in the initial appeal, it is incorrect to conclude that a developer would have no reason to choose another in-app payment processor if Apple charged a commission. Different refund practices, a closer customer relationship, or changes affecting fraud, for example, could make another arrangement attractive. JA478. Those features illustrate why alternatives can compete without offering an identical service at a lower price.
In fact, the district court also recognized that competition is multidimensional, and that part of the purpose of the injunction was to preserve competition through product differentiation—including by Apple. In its order rejecting Apple’s request for a stay of the injunction, the court made clear that it “did not enjoin [the availability of IAP,] but rather enjoined the prohibition to communicate [alternatives.].” Stay Order at 4. The intent was that Apple would have the incentive to compete by offering a better alternative. Thus the court acknowledged Apple’s assertion that IAP was more convenient for users and specifically noted that “if it can compete on pricing, developers may opt to capitalize on that convenience, including any reassur[ance] that Apple provides to consumers that it may provide a safer or better choice.” Id.
For similar reasons, it is a mistake to infer competitive effects from one developers’ idiosyncratic preferences. A channel unattractive to one developer may be useful to another. Some developers may value greater control over ongoing subscriptions, for example, while others—particularly less-established and smaller developers—may value the opportunity to establish and maintain a direct commercial relationship with users. Developers also differ in the costs of operating their own purchasing services and in the number of sales over which they can spread those costs. Requiring every alternative to produce a particular margin of savings would make one price comparison control the inquiry despite those differences.
Because of these differing developer characteristics and preferences (as well as the free-rider problem discussed above), whether a commission frustrates the conduct the injunction protects requires more than a superficial price comparison and recognition of one developer’s self-interested preference for a cheaper alternative. The commission’s effect cannot be understood simply by assuming that the competitive purpose of permitting information was to guarantee a cheaper transaction after accounting for outside payment costs. Rather, as the district court itself noted in its Stay Order, “[c]onsumer information, transparency, and consumer choice is in the interest of the public.” Stay Order at 4. Such considerations plainly implicate the full range of competitive attributes upon which a consumer might decide, not just price.
The original injunction removed specified restrictions on communications. Whether further limits on compensation would improve the competitive outcome requires an economic inquiry of its own. The breadth of the UCL supplies no answer to that inquiry.
IV. The spirit of an injunction cannot set commission rates
Nor can an appeal to the injunction’s “spirit” replace the requisite economic inquiry. In its 2025 decision, the Ninth Circuit recognized this Court’s instruction that “a decree will not be expanded by implication or intendment beyond the meaning of its terms,” and that those terms must be “‘read in the light of the issues and the purpose for which the suit was brought.’” Pet. App. 11a (quoting Terminal R.R. Ass’n of St. Louis v. United States, 266 U.S. 17, 29 (1924)). As noted above, those terms do not address commission rates at all. Nonetheless, citing one of its own earlier decisions, the Ninth Circuit argues that courts can “find a breach of the decree in a violation of the spirit of the injunction, even though its strict letter may not have been disregarded.” Pet. App. 11a-12a (citing Institute of Cetacean Research v. Sea Shepherd Conservation Society, 774 F.3d 935, 949 (9th Cir. 2014) (internal quotations omitted)).
A. Valuing Apple’s services and ensuring developers a margin of savings are different tasks
The injunction’s goal of promoting competition does not, by itself, tell Apple how much cheaper external purchasing must be. The contempt decision faulted Apple both for failing to establish that its commission reflected the value of its intellectual property and for choosing a rate that left insufficient savings after developers paid outside transaction costs. Pet. App. 135a–138a. Those criticisms ask different economic questions. One concerns what Apple contributes; the other concerns the financial attraction of a purchasing arrangement that depends on services supplied by other firms. The problem is that the two need not be the same.
Suppose a commission reasonably reflected the value of the software and distribution services a developer continued to use. That would not establish how much the developer must pay an outside processor, spend handling refunds, or devote to operating a website. Those costs could leave the developer with little reason to use the external channel even though the value of Apple’s contribution had been properly assessed. Conversely, a developer’s ability to operate an inexpensive purchasing service would not tell a court what Apple’s contribution was worth.
A court could require a commission to satisfy both conditions, but doing so would require it to decide which developers’ costs matter, what margin is sufficient, and which services Apple may charge for. That would be a substantive choice about the terms of trade. Neither a general objective of promoting competition nor an instruction to permit external purchasing communications determines those terms.
Similarly, a firm ordinarily prices by considering what customers will pay and how its proposed charge compares with alternatives. It need not begin every pricing decision by assigning a separate value to every component of its offering. If a particular valuation method is required, that is an additional constraint on the decision.
But the injunction prescribes neither that method nor a required relationship between the combined costs of the two purchasing channels. Pet. App. 166a–167a. Indeed, the injunction says nothing about commissions at all, let alone about how to calculate one. The Ninth Circuit recognized this and remanded the case to the district court with instructions to “amend the April 30 Order’s commission prohibition,” along with recommendations on how to do it. Pet.App.35a–36a. A later preference for one particular approach to pricing cannot establish that choosing a different approach was incompatible with the competitive spirit of the injunction.
B. The Ninth Circuit’s opinion would require continuing judicial decisions about prices and investment
This case illustrates how quickly a commission dispute becomes a rate-setting exercise “beyond the practical ability of a judicial tribunal to control.” Brooke Group, 509 U. S. at 223. The Ninth Circuit rejected the permanent ban on commissions and contemplated a prohibition lasting until the district court approved an appropriate fee. Its proposed approach would allow compensation for costs necessary to coordinate external links and for intellectual property directly used in external purchases, with reductions reflecting the non-use of intellectual property reading solely on Apple’s IAP. It also proposed accommodations for security and privacy features and possible assistance from experts or a technical committee. Pet. App. 34a–36a.
Those instructions require much more than merely identifying the expense of processing an external payment. The court would have to decide which parts of the platform contribute to an external purchase, what compensation they warrant, and how to divide costs among uses. Software can support an app whose customers buy through several channels. Distribution can bring a customer to the app long before a purchase. An expense can support both activities without being uniquely attributable to either transaction.
Describing an allocation as proportionate does not resolve the commercial question. The court would still need to choose a measure of use and explain why that measure should determine the allocation of costs or compensation. The number of transactions, the revenue they generate, and the value different developers receive can point toward different charges. A division based on current use can also change as developers respond to the resulting prices. The decision thus requires judgments about behavior and value, as well as accounting.
The problem becomes only more difficult over time. A new service may support several kinds of purchases; a change in outside processing costs may alter the relative attractiveness of different channels; an improvement may require expenditure before it generates revenue. If the permissible commission depends on maintaining a particular commercial advantage for external purchasing, those changes can require further judicial decisions. The court must then administer the continuing relationship among businesses whose products and costs keep changing.
California’s courts have recognized this difficulty in applying the UCL. In California Grocers Ass’n v. Bank of Am., a trial court limited a bank service fee through an injunction based on its assessment of costs and an allowable profit. 22 Cal. App. 4th 205, 211–12 (Ct. App. 1994). The appellate court held the ongoing limitations inappropriate, explaining the institutional problems with judicial supervision of a particular service fee. Id. at 218–19.
Further, the existence of broad remedial authority does not itself equip a court to choose and maintain commercial prices. See Willard v. AT&T Commc’ns of Cal., Inc., 138 Cal. Rptr. 3d 636, 641 (Ct. App. 2012) (collecting cases on the proposition that where a UCL action would drag a court of equity into a complex economic policy question like setting prices, abstention is appropriate because policy is “primarily a legislative and not judicial function”).
This Court has made the same economic concern central to its competition decisions. Trinko cautioned against requiring courts to administer terms of dealing, including prices, in ways that can impair investment and competition. 540 U.S. at 407–08, 414–15. And National Collegiate Athletic Ass’n v. Alston emphasized the difficulties of continuing judicial supervision and the importance of leaving room for business decisions in implementing relief. 594 U.S. 69, 102–06 (2021). These concerns apply with particular force when a court derives detailed compensation terms from an injunction that supplied no pricing standard.
The practical effect is the district court is put in the position of managing Apple’s business relationships with developers, right down to the cent of what is charged as a commission. This type of “micromanaging is beyond the scope of court oversight contemplated by injunction law.” LabMD, Inc. v. FTC, 894 F. 3d 1221, 1237 (11th Cir. 2018). See also Stay Order at 3 (noting the court is not “here to micromanage”).
C. Expanding an injunction by reference to its spirit can discourage beneficial conduct
A business subject to an injunction must decide which activities it can continue, how it should change its arrangements, and which investments remain worthwhile. If complying with the order’s stated terms leaves the firm exposed to contempt because a court later prefers a more expansive understanding of its purpose, the firm may decide to abandon commercially valuable choices the order left available.
And that pressure operates in only one direction. An appeal to an injunction’s spirit can be used to demand restrictions beyond those stated in the order; the firm cannot invoke that spirit to disregard an express command.
Adding such an enforcement rule therefore encourages firms to do more than the order requires whenever the court might later demand an additional restriction. In a competition case, that can mean foregoing prices, products, or investments that serve customers but expose the firm to a court’s disagreement about how best to promote competition.
The resulting costs can be difficult for courts to observe. Cf. Frederic Bastiat, That Which Is Seen, and That Which Is Not Seen (1850).[4] A complaint about continuing restrictions creates a dispute that can be investigated and reviewed. A service improvement abandoned because its permissible compensation is uncertain may produce no comparable record. Neither will a pricing arrangement discarded to avoid another round of litigation. An enforcement approach aimed at improving the competitive process must account for both these “unseen” lost opportunities as well as visible conduct that fails to comport with an injunction. That is the practical importance of the error-cost framework in competition law. See Frank H. Easterbrook, The Limits of Antitrust, 63 Tex. L. Rev. 1, 2–5 (1984); Geoffrey A. Manne & Joshua D. Wright, Innovation and the Limits of Antitrust, 6 J. Competition L. & Econ. 153, 157–59 (2010).
Apple supplies services that retain value regardless of where payment occurs. Commissions can support their continued provision, and the terms of compensation involve choices that the injunction does not resolve. Treating those choices as settled by an appeal to the injunction’s spirit of facilitating competition risks suppressing the conduct that competition law seeks to encourage. Those economic consequences leave a “fair ground of doubt as to the wrongfulness of the defendant’s conduct” under the injunction. Taggart, 587 U.S. at 561 (quoting California Artificial Stone Paving Co. v. Molitor, 113 U.S. 609, 618 (1885)).
Conclusion
Applying this Court’s contempt standard requires reversing the judgment below. The commission-related contempt determination and the remedial directions dependent on it should be set aside.
[1] No counsel for a party authored this brief in whole or in part, and no party or counsel for a party made a monetary contribution intended to fund its preparation or submission. No person other than Amicus and their counsel made a monetary contribution to the preparation or submission of this brief.
[2] Apple’s link-out commission generally applied only to purchases made within seven days after a user followed an in-app External Purchase Link, but renewals of subscriptions initially purchased within that period could remain subject to commissions after the seven days expired. JA518–19. Apple estimated that this seven-day limitation brought the effective rate closer to 18%, rather than the 27% figure the decisions below treated as controlling. Pet. App. 77a n. 25.
[3] Epic challenged some of the assumptions underlying Apple’s estimate, and the district court questioned the extent of its evidentiary support. 6-ER-1140–42; Pet. App. 77a n.25. But even without accepting Apple’s specific estimate of an 18% effective commission, the record supplies a firm basis to expect a rate below 27%. Apple’s customer-spending data and experience with direct purchasing channels supported the expectation that some purchases would occur on that basis. JA518–20; Feb. 25, 2025 Tr. 1588:14–19, 1590:15–1591:18 (Oliver), Dkt. 1307. The presence of any such purchases would reduce the average commission below the headline rate.
[4] Available at https://mises.org/articles-interest/which-seen-and-which-not-seen.