TY - JOUR
AU - Martínez, Alba Ribera
AB - I. Introduction On 22 June 2020, Apple announced the iOS 14.5 update.1 According to the firm, its aim was to protect its users’ personal data. Within the Apple ecosystem, we find multiple layers and sides of customers that interact amongst themselves. On one side we find users and on the other we find app developers, publishers, and advertisers. Apple, on behalf of its condition as ecosystem holder, brings both sides of the platform together. It creates net value for them by putting them together, insofar as they have interdependent demand functions.2 Users mainly exploit the device’s features through their interaction with apps downloaded on the App Store.3 These apps are developed by third-party app developers4 and by Apple through its own proprietary services. App developers, publishers, and advertisers generate their revenue mainly through advertising. The app publisher sells the space to the advertiser to display its ads. Since apps in the App Store are rarely paid, most third-party apps rely on the free ad-based business model, especially small and new app developers. In contrast, Apple is both the ecosystem holder and a competitor before third-party apps on the App Store. Its business model is different to that of these third-party apps: it relies on obtaining revenue through the sale of high-priced devices, such as the iPhone. Its proprietary apps depend on standalone subscriptions providing a wide array of services to users, for instance music streaming or the news service. In addition, it also provides services to business users, i.e., app developers and publishers, through advertising intermediation services on the App Store, namely Apple Search Ads. Different to that of third-party apps, Apple’s business model is closer to a paid business model rather than the free ad-based one. Both Apple and third-party apps use and process iOS users’ personal and non-personal data. These data are stored in the Identifier for Advertisers (IDFA onwards). The IDFA is an identifier from which app advertisers and publishers can infer data on user preferences and behaviours to personalise advertising and to custom user experience on the device. User consent was granted by default up until the iOS 14.5 update, and users could only opt-out through the device’s settings. The iOS 14.5 update changes the whole scenario for third-party apps. The main feature of the update is the App Tracking Transparency (ATT onwards) prompt. Apple mandates every single third-party app to display a pop-up asking the user for its permission to ‘tracking’ if it wishes to access the IDFA. Therefore, from now on, user consent will have to be rendered following this opt-in mechanism. However, Apple has not imposed this obligation on itself. The ATT prompt will not be displayed on its proprietary apps and Apple will continue to personalise advertising and content narrowly on account of its user’s personal and non-personal data. All things considered; we will analyse the potential anticompetitive harm that can be produced through the iOS 14.5 update. Along these lines, the behaviour will be considered based on the refusal to supply category as a starting point from the EU perspective. The requirements established by the EU case law will be reviewed to set out whether Apple’s conduct is anticompetitive. To this end, we will compare the evolution in case law around this legal concept from the Bronner judgment up to the Google Android case. By this same token, we will address the potential anticompetitive effects of the behaviour, compared to those considered by the French competition authority on its interim measures’ decision on the same topic.5 Due to the concerns voiced by competitors and competition authorities, Apple has been forced to backpaddle its own privacy policy: on the iOS 15 update it displayed a similar prompt for its own proprietary services, and it had to sit back on the strict interpretation of the iOS 14.5 privacy policy. We will not comment on the on-going proceedings within the German and Polish competition authorities since they have not issued any decision on the topic for the time being.6 We will discuss whether Apple is leveraging its position as the ecosystem’s holder by imposing unfair conditions to third-party apps through the ATT prompt. Likewise, we will assess the likeliness of foreclosure at the app-developing level of the value chain through the features introduced by the iOS 14.5 update. We will briefly consider the objective justification defence put forward by Apple in relation to the preservation of platform governance. In short, we will first make a brief explanation on the free ad-based business model as opposed to the fee-based business model (Section II). Then, we will outline the mechanisms used by third-party apps and Apple to store, collect, and process user’s personal and non-personal data through the IDFA (Section III). Later, we will illustrate the meaning of the ATT prompt and its implications for both sides of the platform (Section IV). Finally, we will make an analysis on the potential anticompetitive harm of the iOS 14.5 privacy update in relation to the refusal to supply category as a guide to establish anticompetitive effects (Section V). II. Apple’s closed ecosystem vis-à-vis third-party business models A. Apple’s current business model: highly priced hardware and 15–30 per cent fee Apple, Inc. manufactures its devices and runs its own operating system (iOS) within them. As opposed to Android, it does not license its iOS to third-party manufacturers, and this is one of the reasons why the Apple ecosystem is referred to as a ‘walled garden’.7 Apple has deployed its own brick and mortar chain of retail stores for its devices ranging from the iPhone, Mac, iPad to the Apple Watch. In addition, the firm is the ecosystem’s holder on iOS for all these devices. It maintains, runs, and sets out the rules, which will be applied on the devices. Apple has built and strengthened its business model based on the high-quality standards of its hardware, which is highly priced compared to that of competitors. Although its main source of revenue comes from hardware, it is increasingly moving towards a software and service-based model.8 Due to the perceived high quality of its devices, it has been increasingly amassing a user base—notwithstanding its limited number—that will be more prone to spending more on the device, both through advertising and apps.9 As means of drawing a comparison, the Google Play Store—which is well known for the openness of its standards—hosts 3,482,452 applications and generated $38,6 billion in 2020 whilst in the same period, the App Store produced $72,3 billion in sales.10 Enrolment on the App Store is not for free and is subject to the Apple Developer Program License Agreement, alongside with a membership fee.11 Hand in hand with the high-quality brand recognisance that Apple pursues, it has fenced its ecosystem with strict standards for third-party apps on their enrolment. According to Apple, each one of the applications, which can be disposed from the App Store, has been reviewed to guarantee the optimal user experience in terms of privacy, navigation, and functionality.12 These rules apply to third-party apps, i.e., its competitors in the downstream market, and to its own proprietary services. Apple is well known for being strict when non-compliance with its internal guidelines is detected.13 Furthermore, Apple’s services range across several applications, such as Apple Music, Apple TV+, Apple News+, or Apple Arcade. All of them are powered through the user’s subscription to the service, and they are not available elsewhere for free, except for the free trial month. Apple recently announced and launched Apple One: a standalone subscription service, which supports the services provided through Apple Music, TV+, Arcade, and iCloud across Apple-manufactured devices.14 Since December 2020 when the iOS 14.3 was rolled out, from the business-to-business perspective, the firm offers Apple Search Ads, its own paid attribution and personalised advertising via Apple Ads Attribution API.15 Although it is available for third-party apps on subscription, the whole set of Apple-owned applications are powered through the functionalities offered by Apple Search Ads.16 Moreover, Apple charges a standard 30 per cent fee for each in-app purchase (this is, purchases made on apps, for instance, to gain advantages on games), subscription, and paid app downloads made by users on third-party apps.17 Since 1 January 2021, small businesses and independent developers that earned up to $1 million in the previous year can qualify to a 15 per cent reduced commission.18 In fact, 20–40 per cent of Apple’s total services revenue comes from the App Store.19 Based on the foregoing, we can say that Apple’s business model is that of a financially independent platform from the rest of its competitors on the App Store on the downstream market.20 Ongoing proceedings of the European Commission and Bunderskartellamt are currently investigating the validity of Apple’s business model altogether.21 By this token, we gain some insight on the incentives, which have prevailed whilst designing the iOS 14.5 update in relation to privacy. B. App developers’ and advertisers’ business model: free ad-supported apps The Apple ecosystem relating to apps is comprised by several sides to it: on the demand side we find users that download apps to navigate through their iPhone, whereas on the supply side we find app developers. In addition, advertisers are willing to announce their offerings through in-app advertising or video in-app ads.22 First, app developers create applications and then contract with Apple to make them available in the App Store. Then, app publishers match advertisers with users to display advertising on their behalf.23 Due to the revenue produced through advertising, app developers are capable of subsidising users so they can access their services for free.24 This is all possible by virtue of the indirect network externalities produced by the platform.25 Besides, different sides of the platform are charged with different prices through skewed pricing; although users are charged nothing for the service on the app, advertisers are charged by developers with a fee when an action is performed.26 For instance, when a user buys one of the advertiser’s products. This mechanism ensures a sort of equilibrium amongst the different sides of the platform and that a critical mass of customers remain within the platform because they are, in appearance, ‘free-riding’ the third-party app’s service.27 The purchase of paid apps on the App Store is still rare, therefore app developers and advertisers rely on the revenue they can capture through advertising and make apps available for free.28 On average, ads account for 63 per cent of their total revenue.29 To enhance the possibilities of effective advertising, advertisers and publishers target ads and tailor content according to user preferences. Advertisers target those users that, in principle, should be the most interested in their offerings.30 Other business models that are possible within this ecosystem range over the App Store. For instance, Netflix is a clear example of the fee-supported model.31 When downloaded through the App Store, Netflix is available for free. Nonetheless, the service is rendered in exchange for a subscription. If the user decides to make the subscription through the iPhone, Netflix will be subject to the 30 per cent fee charge imposed by Apple. The user side of platform is not entirely subsidised by the app developer, i.e., Netflix. The media service produces its revenue from the subscription service and not through in-app or video in-app ads, which are not displayed on Netflix on principle. A different business model is that of Spotify, that depends upon the freemium model. On Spotify, the user can access the music streaming service for free on its ad-supported and basic version, or rather choose to pay for the subscription to access the richer app functionality version. Again, if the user decides to pay the subscription on his iPhone, the 30 per cent commission will be charged to Spotify. To counterbalance the fee, Spotify decided to pass it on to the consumer and charge a higher price on iOS devices.32 However, most third-party apps and brands might not be as differentiated and resourceful as Spotify or Netflix and may not be able to apply this business model due to their apps’ characteristics and specific functionalities. In addition, app publishers and advertisers can leave out personalised advertising from their business model altogether, whereas the return on their investment would be much lower.33 III. The Identifier for Advertisers: advertising and attribution purposes The prevailing business model amongst third-party apps is not all that different to other business models that predominate on different sectors of digital. For instance, in general search services, such as Google, they go hand in hand with online search advertising.34 To this end, digital markets have introduced identifiers: a unique record where the user’s preferences and behaviour manifested online are stored. For the Internet, website cookies are used for this purpose, whereas the user is identified through his IP. Not only the possibilities of advertising are enhanced, but also displaying targeted advertising and content is easier for them. Within mobile ecosystems, identifiers are also used based on the device’s ID or mobile advertising identification number (MAID).35 Rather than cookies, a unique encrypted identification number is assigned to the device by the operating system of the smartphone when it is first used.36 App advertisers and developers can access data on the use of apps easily because it is stored within a unique identifier and remains the same for all of them. For instance, Android uses the Android Advertising ID (AdID).37 Since 2012, Apple caters this feature through the Identifier for Advertisers (IDFA) and the Identifier for Vendors (IDFV). The IDFA is assigned to each iOS device by default when it is first set up. The user’s interactions and behaviour on the device are accumulated progressively and in real time for advertising and analytics purposes (third-party data), whereas the IDFV can be used by the app developer to monitor this same behaviour across the apps owned by this same vendor (first-party data).38 For instance, Meta can access the IDFV to monitor and combine the user’s behaviour across Facebook, Instagram, and WhatsApp. The IDFA’s lifespan ranges from the moment that the user switches on the iPhone for the first time throughout all the interactions performed on the device.39 Since it was first launched, the IDFA can be reset by the user. If this option is carried out, the data within it is erased and the device is allotted with a new and unique random identifier. From 2016 onwards, Apple includes a ‘Limit Ad Tracking’ option on its settings. Consequently, users are able not only to reset the IDFA but also to disable it altogether. Such being the case, through this option users can block third-party apps and Apple’s proprietary services to access it.40 At the moment of launching the IDFA in 2012, Apple did not create and develop its own proprietary apps, so the data stored within it was especially valuable exclusively for independent app developers. It still is. Through the use of the IDFA the possibility of offering high-quality targeted advertising and content on their apps is enhanced. Within it, they can find the strongest indicators of high-quality users such as the average session time of the user or the session frequency on apps.41 Opposed to regular advertising online, in-app ads can produce extraordinary sources of revenue for app developers and advertisers in terms of advertising and are most adequate when adjusted according to their context and time.42 For instance, an in-app ad can be the most effective when displayed after the user has spent a long amount of time on the app. In the same manner, they will also be the most effective when the same ad is not shown too frequently and repetitively to the same user.43 Moreover, these types of ads have been proven to work on the user’s impulsiveness.44 Thus, the in-app ad will be more likely to trigger an action on the user’s side if based on concrete and relevant preferences.45 Secondly, also through the IDFA, advertisers can personalise offerings to consumers more easily and increase the relevance of their content, with a similar effect to the learning by doing effect.46 Thirdly, advertisers, publishers, and developers can gain a better insight on how they are doing, in terms of advertising and personalised content.47 With the support and access to the IDFA and IDFV, they can perform attribution and digital measurement with ease. Put differently, they can draw out easily the causal link between a given event or ‘impression’—i.e., the in-app ad—and an action or ‘conversion’—i.e., a visit on a particular website or the purchase of the advertiser’s offering.48 Attribution and digital measurement not only come in handy for assessing performance but also for monetising their activity.49 The advertiser will only pay the app developer—other than for the space that is made available on the app—if the in-app ad renders effective and accurate in terms of user engagement, and only if it can establish the triggering event of the user’s action.50 Due to these reasons, the IDFA has been an attribution and digital measurement tool prime to most small and new app developers.51 Put simply, it matches supply and demand in relation to their advertising activity.52 Although large and resourceful firms can power their own attribution tools or contract with third-party companies that provide this service, for smaller players within the iOS ecosystem this is not always possible. In the face of this concern, Apple implemented in 2018 the SKAdNetwork and offered it free of charge for its developers. Although data at the user level are not available, it aims to facilitate attribution and model analytics of mobile app installation with ad campaigns.53 On Apple’s side, the IDFA is valuable for the same reasons as for third-party apps, given that in the last few years it has developed its proprietary apps and services. On top of that, Apple is also the ecosystem holder and can monitor the success of the advertising services offered through the IDFA and IDFV. Bearing in mind that it can oversee performance, it can feed its own services upon third-party app’s past experiences. In addition, it is capable of pre-empting competition in the guise of protecting platform governance and functioning.54 IV. The iOS 14.5 update: App Tracking Transparency A. The App Tracking Transparency prompt, explained Apple issued a statement in June 2020 communicating its intention to provide ‘enhanced privacy features for more transparency and control’ over the ecosystem through its newest update: the iOS 14.5 update.55 One of the features powered through the update is the ATT mechanism, a pop-up where users are asked to allow or disallow ‘tracking’ over apps as a binary choice.56 If we dig deeper into Apple’s statement, tracking refers to ‘the act of linking user or device data collected from (an) app with user or device data collected from other companies’ apps, websites, or offline properties for targeted advertising or advertising measurement purposes. Tracking also refers to sharing user or device data with data brokers’.57 However, is this what users think about when tracking is mentioned? It does not seem to be the case. Users tend to associate it with surveillance and monitoring and react poorly towards it.58 Nonetheless, app developers access the IDFA and IDFV to make attribution possible with advertisers as well as to personalise advertising and content. However, their activity is placed in a negative light before users.59 Apple’s definition on the same action is striking by comparison. The following message is displayed when the user opts-out from personalised advertising altogether, including Apple’s proprietary apps: ‘[t]urning off personalized ads will limit Apple’s ability to deliver relevant ads to you. It may not reduce the number of ads you receive’.60 Considering the definition applied for the ATT prompt, Apple ‘tracks’ its users. Apple does offer advertisers contextual information about users to facilitate personalised offerings.61 So, what is the difference between third-party apps and Apple’s proprietary apps? In terms of personalised advertising, nothing at all. Fundamentally, a great risk is associated when third-party apps ‘track’ users, but not when Apple does the same.62 Due to the concerns voiced out by competitors and regulators, on the last few months Apple has been forced to backpaddle its own privacy policy.63 Although on its subsequent iOS 15 update it displayed a similar prompt when its own proprietary services were powered by the user, the firm was also forced to loosen up its strict interpretation on the iOS 14.5 privacy requirements surrounding the ATT prompt and its enforcement on the App Store.64 B. Consequences on third-party apps From past experience, estimates show that any advertising identifier removal is bound to result in a 52 per cent decrease in advertising revenue.65 Therefore, the shift in Apple’s privacy policy can produce unfavourable market outcomes.66 The ATT prompt is mandatory for third-party apps and not for Apple’s proprietary services under Apple’s App Store Review Guidelines.67 Each one of the third-party apps that wish to ‘track’ users will have to display the ATT prompt. Consequently, the user, within the first few weeks to have installed the iOS 14.5 update, will be asked repeatedly to allow, or disallow ‘tracking’ on each app that is accessed. App developers can choose the time when they will display it. They can choose to disclose the feature when a user downloads and launches the app for the first time, or rather after the user has experienced the app’s service for a short period of time. The app developer can also choose to display an additional pop-up window to explain to the user the purpose of ‘tracking’.68 All the same, it will have to be displayed if the third-party app’s intention is to access the IDFA, whereas the IDFV’s use is not questioned.69 The user can enable or disable tracking as many times as necessary through the device’s settings, in the same way that opting out of the IDFA was optional through the previous iOS updates. De facto, app developer’s and advertiser’s chances to perform attribution have drastically reduced.70 When the ATT prompt is displayed, there are three possible outcomes71; first, the user renders consent to the advertiser, the app developer and publisher; second, the user renders consent to one but not the other; and finally, the user denies consent to perform ‘tracking’ whatsoever. If the user denies consent, all the players involved will be barred to access the IDFA. However, when consent is granted to one of the parties involved in the transaction, attribution cannot be performed. For instance, the user’s interactions will only be ‘visible’ for the app developer but not for the advertiser, so the causal link between the in-app ad and the user’s action cannot be drawn out using the IDFA.72 Thus, attribution can only be performed successfully when both parties involved are granted consent to ‘track’ the user. The user will have to opt-in twice so that the causal link between the in-app ad and the user’s action can be generated easily. We can only imagine the multiplicity of scenarios that will arise due to the ATT prompt. For instance, if the app developer and publisher want to display in-app ads, they will contract with several advertisers. The possibility to perform attribution will vary depending on user consent, which will be repeated by the millions for each third-party app available on the App Store. Consequently, the chances of third-party apps to generate revenue based on the ad-model are reduced.73 As a matter of fact, personalised advertising, and content, as well as attribution will not be possible. If both activities are unattainable, the free ad-model crumbles third-party app’s potential to produce revenue.74 We have in front of us the problem of the chicken and the egg. If third-party apps cannot access the IDFA, then they will not have the opportunity to display in-app ads tailored to user preferences. App developers and publishers cannot reassure advertisers when matching them with users through ads that the in-app ad will be successful and trigger an action. Therefore, advertisers will be less and less incentivised to display their offerings on third-party app’s in-app ads. Consequently, there will be less advertiser demand to display offerings on independent app developer’s apps, and the prices set by developers to display in-app ads will be lower.75 As a matter of fact, on average ad placement’s price without the IDFA functionality is 44 per cent lower compared with the price of ad placements with the IDFA embedded within them.76 All things considered; monetisation measured in costs per impression (CPMs) for advertisers is less probable for them when relying on third-party apps on the iOS.77 Hence, advertisers may shift to apps with enough background information about users to ensure a certain level of engagement with their offerings. That is, apps developed by large and resourceful contenders that can attribute advertising through alternative methods and/or that can personalise advertising through Apple’s proprietary attribution API. All in all, if third-party apps want to survive within the Apple ecosystem, they will have to fine-tune their business model to monetise their services in some other way.78 V. Exploitative abuse through the iOS 14.5 update From a competition law perspective, it is difficult to presume that the iOS 14.5 update is a simple update on Apple’s privacy policy, and not an abnormality on the platform, following the terms of the Google Android ruling.79 A. Foreground to the analysis Abusive conducts arise in exceptional circumstances where competition is substantially weakened.80 Apple’s conduct takes place in its closed ecosystem, and this is the factual, legal, and economic context of the practice.81 Therefore, when alternative methods are used as opposed to those of ‘normal’ competition, Apple is bound to maintain the degree of competition still existing in the market according to the general principal of equal treatment that must be preserved when an abuse of a dominant position is concerned.82 It goes without saying that an update such as the iOS 14.5 can be quite disruptive, especially on the neighbouring downstream market of third-party apps. In this regard, prior experience on antitrust, such as the Intel judgment, shows that some practices can be abusive by nature or due to their effects or, contrarily, can be the product of the legitimate efforts of the undertaking.83 Therefore, a case-by-case analysis is necessary to assess the impact of the practice considering its scenario. Along these lines, we will consider whether Apple’s conduct is deemed exclusionary on the grounds of the ‘no economic sense test’ since not every exclusionary effect is necessarily harmful for competition.84 That is to say, we will analyse if Apple’s practice to foreclose its competitors on the downstream market, i.e., third-party apps, can be justified on the basis of other economic reasons different to the anticompetitive practice, even when downstream operators can be marginalised or disappear from the market.85 It is important to bear in mind that the framework to the iOS 14.5 privacy update is the digital arena were zero-priced services are the rule, so that the analysis will not be centred on price parameters, but on the assessment around the potential diminishment of quality in advertising hand in hand with an increase on consumer choice.86 Experience shows that pre-existing categories of abusive behaviour are not always that comparable to anticompetitive conduct in the digital ecosystem.87 This is the case for the ATT prompt. We have chosen the refusal to supply category as a starting point for our analysis, considering that the IDFA is the input, which is needed to provide the service in the downstream market of retail, where leveraging is exerted.88 Apple conditions its access to third-party apps upon the acceptance of certain limitations sought out in the ATT prompt. Following these lines, first we will assess the relevant markets concerned by the practice as well as Apple’s dominance in them. Then, we will draw out the anticompetitive effects of the behaviour and the threshold of portability required to show them, in reference to the necessity to prove the indispensability of the IDFA.89 Even if the conduct is anticompetitive, we will overview if it can be justified by Apple’s role as the ecosystem holder to govern its own platform.90 B. Pre-condition on dominance Apple Inc’s activity enclose a wide array of products and services. Within the iPhone’s ecosystem, it is present at the upstream and the downstream market. We will highlight some of the conclusions that the French authority drew out from its analysis of its interim decision. 1. Upstream market At the upstream distribution level Apple provides for the App Store on the market of distribution of mobile apps on iOS.91Google Android has been the first EU wide case to analyse the distinctive environment of apps. It analysed the peculiarities of the Play Store in front of other app stores and the degree of substitutability between them. On one hand, from the supply viewpoint, the decision settled that app stores serve different purposes to regular apps.92 In fact, the user cannot download apps on the device without an app store. In line with this, the App Store is pre-installed on the device and there are no potential threats upon it to provide alternative app distribution channels. The App Store is the gateway to access the whole set of applications to users, either Apple’s or third-parties’. Similarly, it is the only store that allows app developers to reach iOS users.93 Furthermore, as opposed to Android, Apple does not license its smart mobile operating system to third-party device manufacturers. Therefore, the App Store is a separate market from Android app stores, regardless of their similar features and functionalities.94 When analysing the relevant markets involved on the iOS 14.5 update case, the French competition authority defined one existing market upstream. Considering the intricacies of Apple’s ecosystem, the App Store is the only distribution channel for apps on iOS that is available worldwide.95 Therefore, the worldwide market for the distribution of mobile apps on iOS was upheld in the decision. 2. Downstream market On the downstream market, Apple competes with third-party apps, where advertising and attribution is performed based on the Apple Developer Program License Agreement. Consequently, the French competition authority defined two separate markets according to the business activities and the existing relationships between app publishers and developers with advertisers.96 First, it defined the national market of mobile app install ads on iOS.97 Second, it defined the market of attribution services for mobile app installations on iOS.98 Other than that, it left open whether these markets could be narrowed even more depending on the type of service that is offered, as per the Google/Fitbit decision.99 For instance, Apple News+ would compete in a different market to that of Apple TV+ or Apple Music, but all of them compete at the same level of the value chain of the distribution of apps. 3. Dominance In digital and multi-sided platforms, market power can be assessed in the light of different elements, such as the number of developers and apps per app store or the share of multi-homing consumers and developers.100 However, in this case, establishing Apple’s dominance is more straightforward than that: Apple and Android app stores effectively form a duopoly on global markets.101 Apple is built on a closed ecosystem where no alternative and substitutable app stores are effectively available.102 On the market for iOS app stores, the Apple Store is near to 100 per cent in market share.103 It holds a quasi-monopolistic position on the market of distribution of mobile apps on iOS.104 Third-party app stores can be downloaded on the iPhone but do not pose a real threat to the App Store’s predominant position. For instance, the immediate competitor to the App Store is Builds.io, which holds more than 15,000 apps for the iPhone.105 For the sake of simplicity, we will analyse the interaction between the upstream market of distribution of mobile apps on iOS and the downstream market for advertising services for mobile apps on iOS and analyse whether Apple followed its legal obligation of non-discrimination notwithstanding its dominance.106 C. Causal link between dominance and the contested practice 1. The French competition authority’s decision on the ATT prompt The French competition authority in its interim measure decision issued on March 2021 decided in favour of the ATT prompt. In sum, it upheld that there was no causal link between the iOS 14.5 privacy update and the potential anticompetitive effects that could be caused by it. As per its decision, Apple is inherently free to govern its own platform by setting out the rules to the market of app distribution on iOS devices.107 In this same vein, the firm will only be limited on its action by the compliance with laws and regulations. The authority believes that Apple implemented its privacy policy under objective, transparent and non-discriminatory conditions.108 First, the iOS 14.5 update is coherent with Apple’s long-standing policy of enhancing privacy and protection of the user’s personal data. By doing so, Apple attends to the growing consumer demand on enhanced privacy.109 Second, the authority sustains that the measure is not excessive nor disproportionate to the interest of consumers. For example, there was due notice for app developers to adapt to the ATT prompt from June 2020 to April 2021. Finally, according to the opinion rendered by the National Commission on IT and Freedoms (CNIL onwards), the competition authority found that the ATT prompt adjusted to data protection principles. The CNIL compared the ‘Limit Ad Tracking’ option to the ATT prompt. By doing that, the data protection authority finds that the pop-up is, both in functionality and design, a simple way to render and withdraw user consent on ‘tracking’. Furthermore, the CNIL excludes the ATT prompt from the GDPR’s scope of application.110 The data protection authority corroborates that the ATT feature is designed as a complement to provide additional protection for the user about user ‘tracking’ on apps. Thus, the GDPR standards are not applicable to the iOS 14.5 privacy policy altogether.111 2. Connection between the iOS 14.5 privacy update and Apple’s dominance From our viewpoint, the French authority errs on its decision by making some assumptions on the operational features embedded within Apple’s mobile ecosystem. All in all, the iOS 14.5 privacy update hinders third-party apps’ access to the IDFA data stored through the ATT prompt. In doing so, two main parameters of competition are impaired: quality and consumer choice. The theory of harm must be built around these parameters to establish whether Apple’s practice generates anticompetitive effects. To this end, the functionalities of the ‘Limit Ad Tracking’ option on the devices settings and the ATT prompt are not to be conflated. Rather, they must be looked at through a different lens. On one side, the ‘Limit Ad Tracking’ option applies to third-party apps and to Apple’s proprietary services. When the user chooses to opt-out through it, personalised advertising is very difficult to perform altogether for both. On the other side, the ATT prompt is only directed at third-party apps.112 However, as opposed to the previous feature, if the user chooses to disallow ‘tracking’, not only personalised advertising is impossible but also personalised content and attribution. This is the point where the causal link between anticompetitive effects and the practice at issue come to mind.113 The disparity between both features show that Apple is financially incentivised to hinder third-party app’s access to the IDFA. Mainly, because it takes Apple’s ecosystem one step further to achieve a total transformation to a fee business model as opposed to an ad business model. D. The indispensability requirement through the Google Shopping lens Since 1998, the ECJ set out the requirements to the refusal to supply practice through the Bronner judgment.114 On the basis of the general principle of equal treatment contained in the Google Shopping ruling, the General Court indicated in what conditions the Bronner requirements are applicable.115 In Bronner the Court outlined the three cumulative requirements that must be met so that a refusal to supply practice can be established. First, the conduct should eliminate all effective competition in a different market. Second, there should not prevail any objective justification to explain the refusal. Last, the input should be indispensable to carrying out the service in the downstream market.116 The economic rationale behind the high threshold placed on the refusal to supply is twofold. First and foremost, the imposition to grant access to an input collides with the undertaking’s fundamental freedom to contract and the right to property.117 Second, although on a short-term basis competition on innovation could be fostered, forcing access can lead to decreased innovation on the long run regarding the dominant player’s incentives on building such infrastructure.118 These incentives will disappear for the dominant undertaking to develop the input and for the party seeking access. Taking this reasoning at face value, the iOS 14.5 privacy update cannot be interpreted as a refusal to supply scenario if the IDFA is not indispensable. Third-party apps can perform attribution by virtue of other datasets different to the IDFA. For example, data brokers and companies can perform attribution for them in exchange for a fee. However, this may not be possible on the downstream market in an economically viable manner for smaller players since traffic within iOS devices can only be produced through the App Store.119 Therefore, different to Bronner, we are not in front of an outright refusal to supply since the conditions set out by the Google Shopping ruling are not met.120 First, there is not a clear request from the competitor followed by a consequential refusal from the dominant player. Apple instrumentalises users and the overreaching power of consumer choice to deny access to the IDFA for third-party apps. Second, the effects triggered by the practice do not rely principally on the refusal. The effects are not directly attributable to the conduct on itself, but on the disparities caused by the decisions rendered by the users on the device when the iOS 14.5 update is downloaded on the device. The practice is, therefore, an implicit refusal of access given that the conditions and terms for access have been altered so that they result in unreasonable and uneconomic circumstances.121 Under the ‘no economic sense test’, the denial can only be explained to limit the degree of competition on the downstream market. Following Google Shopping, the practice constitutes an independent infringement of article 102 TFEU other than an outright refusal to supply.122 If indispensability is not required, then what is the standard that Apple’s conduct should be measured against? The Google Shopping ruling gives us some insight on the rule to apply since the implicit refusal to access is an independent manifestation of article 102 TFEU. To this end, it is enough to demonstrate that there has been an active behaviour from the dominant player in the form of positive acts of discrimination. Unequal treatment is not the same as an abuse, but the factors contributing to prove an exclusionary effect are part of the legal test applicable.123 That is, anticompetitive effects must be shown in relation to the existence of discrimination stemming from the difference in treatment of the access rendered to the IDFA for third-party apps in front of the expected and relevant principle of equal treatment.124 E. Anticompetitive effects through the elimination or restriction of competition The exclusionary effects causing the elimination or restriction of competition must be established according to its likelihood since discrimination is not enough to find an abuse, and other elements should be considered, such as user behaviour stemming from the practice or Apple’s position as an ecosystem holder.125 To this end, in the Microsoft judgment, the Court highlighted that the Commission is not required to wait until a competitor is effectively eliminated from the market. On the contrary, this requirement would hinder the maintenance of undistorted competition in the common market.126 Once likeliness is established to be the rule, we will analyse the parameters of competition affected by the iOS 14.5 privacy update. The ATT prompt does not involve a monetary transaction between the user and the app developer. Thus, the traditional price-related analysis is not applicable. However, economic implications can be drawn out from the perspective of non-price related parameters of competition, for instance, the quality of advertising and consumer choice. 1. Impact on advertising and attribution According to the constructive refusal orchestrated by Apple, the access to the IDFA is effectively hindered for third-party apps.127 Therefore, the anticompetitive harm of the behaviour manifests on their capacity to perform personalised advertising and content. First, third-party apps are hampered from performing both personalised advertising and content in the detriment of users.128 By this token, the quality of these two services is diminished within third-party apps. Although advertising through in-app ads is still possible, without the access to the IDFA, targeted advertising will no longer be as effective as it was.129 This will have a pronounced impact on the short run concerning their possibility to produce up to 50 per cent of their revenue through advertising, especially in comparison with Apple’s proprietary apps.130 Although non-personalised advertising will be the rule for independent app developers, Apple has its way paved out in the opposite direction as the ecosystem holder. Also, Apple can favour itself through its economies of scale and scope.131 As long as the user does not choose to ‘Limit Ad Tracking’ through the device’s settings, Apple will have access to both the IDFA and IDFV, as well as to the additional data produced by all of its divisions and units, i.e., usage of its iPad, Apple Music, iCloud, etc. On this account, as opposed to third-party apps, Apple will be able to target and re-target users across the iPhone device.132 Besides, it will have the capacity to repurpose, recombine and multi-purpose its datasets at a low marginal cost.133 This could explain the calls for future expansion on Apple Arcade, Apple News+, and Apple Search Ads.134 As a matter of fact, this may not be so prejudicial for independent app developers if advertising can be performed. The fact that consumers prefer targeted ads to non-target ads is disputed.135 There is little evidence to acknowledge that personalised content and advertising is, in fact, an improvement in quality for consumers. In addition, the rising awareness of consumers concerning the protection of their personal data has shown that non-targeted advertising can be more appealing to them. On this front, Apple can trip over its own strategy to accumulate large datasets for targeted advertising. On top of that, the effects on scale and scope can also lead to an overexposure of consumers, which may not be desirable in terms of user experience.136 However, the truth is that independent app developers, especially small and medium-sized ones, monetise their free content through in-app personalised ads and may not have other sources of revenue.137 The collection, storage and processing of datasets requires large resources to face upfront investment.138 This may not be possible for independent app developers, so that replicating a dataset like the IDFA is not conceivable for them, as it is for the Big Tech companies.139 For instance, Google and Facebook have integrated their SDKs140 for user ID cookie matching on Apple devices since the iOS 14.5 was rolled out.141 The economic value of the IDFA does not rely on the fact that only Apple can access it and, therefore, cannot be replicated—as seems to be the benchmark established by the case law on exclusionary practices not related to pricing,142—but instead by the existence of economic barriers for data collection.143 These are difficult to overcome for third-party apps and require an upfront investment that cannot be matched with Apple’s expenditure on data collection. Apple leverages its dominance to the downstream neighbouring market where it competes with third-party apps, so that only the fee-based business model is economically viable within the ecosystem.144 Henceforth, revenue on iOS devices will be produced through the upfront payment to download an app as well as through in-app purchases and subscription services.145 It goes without saying that the possible transformation of the business models that are harvested within Apple’s ecosystem, although lucrative, may be risky on many fronts.146 Up until now, the App Store has obtained a critical mass of users and independent app developers on both sides of the platform. Users can obtain free apps, and app developers have remained interested in the platform to provide their services.147 In addition, advertisers have exploited in-app ads to display their offerings on apps. At present, these externalities maintain the multi-sided platform in balance, so that all the consumers concerned remain connected to the platform. The change on the prevailing business model on the App Store will imply that, as a rule, consumers are no longer fully subsidised on their side of their platform.148 Instead, they will have to internalise some of the costs that have been assumed up until now by independent app developers. Thus, the nature of the indirect network effects sustaining the Apple Store will be overturned completely.149 For instance, it may drive away users, since they are not willing to internalise these costs.150 Having said that, switching costs for users of iOS devices are high at present. Interoperability with Android devices is low. In 2017, more than 90 per cent of users with Apple-branded devices chose again the brand when they renewed their smartphone.151 If consumers are increasingly bound by standalone subscriptions and in-app purchases to the App Store, these switching costs may increase.152 A user will not switch to another device, i.e., an Android device, if the features and level of functionality achieved through payments is not interoperable with it. Thus, users are locked-in into the Apple ecosystem through their past purchases insofar as they amount to sunk costs performed on the device.153 In the event that the user wants to accomplish the same enhanced user experience obtained through purchases, he will have to pay them again on the Android device.154 That is, pay twice for the same service. That being said, some features and applications obtained through subscriptions or in-app purchases on the App Store are not available in other devices other than iOS devices. On top of the lock-in effect of consumers within the Apple ecosystem, cross-platform competition will be effectively hampered between Android and iOS devices.155 Second, the ATT prompt means that quality decreases for the performance of attribution on third-party apps, and in some cases, it is hindered entirely. For effective attribution to take place between advertiser and developer the user will have to opt-in twice. De facto, attribution will no longer be possible for independent app developers through the IDFA, given that advertisers will no longer be able to assign the user’s action to the specific app that triggered it.156 This will create a free-riding effect in favour of advertisers and to the detriment of third-party apps. Regardless of the app that triggers the action, the advertiser will obtain engagement for its offerings. Even if the advertiser would want to reward the app, it would be extremely difficult to do so without the access to the IDFA. As an alternative when users choose to disallow ‘tracking’, Apple enables its own attribution tool for free: SKAdNetwork.157 However, this source of mobile ad analytics is not quite the same as the access to the IDFA. The ability to provide with ad attribution in real time and in a granular way is limited.158 App developers have voiced out their concerns on SKAdNetwork’s performance in relation to attribution, which is only possible through aggregate data and operating with a delay between 24 and 48 hours after the user performs an action.159 Although SKAdNetwork is powered by Apple, it does not respond to the Apple Ads Attribution API. Apple’s proprietary attribution service functions provide greater and more granular ad attribution. For instance, it includes the date of the conversion and notifies on attribution in real time.160 As a matter of fact, Apple Search Ads is now responsible for over a half of the total downloads resulting from an ad, as opposed to its share in 2020, which only amounted to 17 per cent.161 On top of the effects caused by the ATT prompt on advertising, the iOS 14.5 update drives third-party apps away from the free ad-based business model. Both effects to the ATT prompt concerning the reduction in quality in advertising and attribution, give way to the decreased rivalry on the downstream market competition. On one hand, the feasibility of third-party app’s business model is put into question. On the other hand, even if third-party apps adapt to the paid model, Apple’s proprietary apps will be two steps ahead of the start line in terms of competition regarding attribution and engagement with users. 2. Consumer choice v. privacy The iOS 14.5 privacy update gives leeway to the user to benefit from an additional layer of protection before data storage, collection, and processing. As opposed to price, consumer choice is another parameter of competition to be considered in front of an anticompetitive behaviour. However, privacy enhancing features have been conflicting with consumer choice.162 This may not always be true. First, let us keep in mind that protection of personal data, as drawn out by article 8 of the Charter of Fundamental Rights,163 is a fundamental right.164 As such, it must be protected, and its exercise must be ensured. Moreover, privacy has even been considered as an additional competition parameter in itself.165 User’s willingness to allow firms to collect their data can, in fact, affect the dynamics of the market.166 In the case of Apple’s ecosystem, market outcomes can vary depending on the data accessible to third-party apps.167 In a nutshell, there is still some ambivalence on the relevance and implications of an economic theory of harm based on privacy.168 The main problem is that privacy protection can produce ambiguous effects on competition. Several intertemporal trade-offs are to be considered when assessing user preferences on privacy.169 For instance, users of iOS devices trade-off the control to their personal and non-personal data to third parties in exchange for personalised services, such as advertising and increased accuracy on user experience and content.170 On top of this, users do not always act on their privacy preferences whilst online. According to the privacy paradox rule, although users wish to protect their personal data, they do not always act on it.171 When a user is more prone to protecting his privacy, he may only render consent to those apps considered, in his view, more trustworthy, i.e., those apps that he uses the most. The user enables access to his data from an individual perspective, that is, in reference to the protection of his personal and non-personal data.172 One of the reasons of this is that the sensitivity towards privacy is, usually, subjective, and idiosyncratic.173 The same goes for the concept of ‘consumer well-being’, which is purely an element of inherent subjectivity.174 Taking into consideration the mechanisms of the ATT prompt, user experience can be as different as the choices that each user makes across different apps when accessing them.175 These choices can be driven by several behavioural biases at a time.176 For instance, the repetitive nature of the pop-up can cause a shift in the initial decision on allowing ‘tracking’ as a rule.177 That is to say, the repetitive use and need for consent can result in a dissuasive effect making the disallowing of ‘tracking’ the rule, and not the exception. In addition, when faced with consent-based data collection repeatedly, users tend to favour and accept the terms of larger and more diversified rather than small and medium-sized firms.178 3. Effects on privacy and consumer choice Given that the trade-off resulting from the ATT opt-in mechanism can produce several outcomes, we will look into both scenarios: when the user opts-in and when he does not. First, we will address the concerns when the user does not enable ‘tracking’. In principle, the user will remain more protected due to the iOS 14.5 privacy update. Users wish to protect their personal data so that this could be considered welfare enhancing, under the terms of the purposes that justify the intervention of competition law.179 However, the user incurs in major opportunity costs if a decision is made to disallow ‘tracking’ on third-party apps. These apps are hindered from offering tailored and context-specific advertising and content with detriment to user experience.180 The fact that this happens, however, may not always be negative for users, as the value of targeted advertising from an economical perspective is also contested.181 This may not always be true if independent app developers can access alternative datasets to the IDFA if they decide to invest in it. However, smaller, and less resourceful firms may find it difficult to match Apple’s investment on datasets. Under these circumstances, an enhanced privacy protection results in deprived consumer choice provided that advertising and content is not narrowed down to user preferences.182 The opposite may also be true. For instance, if app developers choose to perform non-targeted advertising through in-app ads, consumer choice can be more diverse considering that content is not captured exclusively by user preferences.183 Second, we will address the scenario where the user renders consent on ‘tracking’. As opposed to the other decision, this will not provide for the privacy enhancing feature introduced by the iOS 14.5 update. Users that choose to opt-in are more prone to be biased by the expectations of receiving an immediate gratification.184 They will expect to be provided a better service, both on advertising and content, instantly and free of charge.185 However, user experience will only improve increasingly overtime.186 The association between personalised content and advertising and increased consumer choice is also contested.187 When access is rendered to the IDFA, users can be easily narrowed down in relation to their preferences. They can also be discriminated as easily. Third-party apps could display ads of those products that are most profitable for them, but not those, most suited to their interests.188 When too much data are in their hands, free choice can be easily undermined.189 Moreover, the non-visible costs of enabling ‘tracking’ through the ATT prompt are unpredictable.190 Once access is rendered to the IDFA, data are long gone from the user’s hands. The data accessed through ‘tracking’ can be shared and brokered with additional third parties (i.e., data brokers) different from third-party apps on multiple markets.191 The ATT prompt poses particular and nuanced trade-offs in terms of privacy.192 Users, app developers, publishers, and advertisers can gain and miss out on the beneficial outcomes produced by the interaction with the IDFA. Therefore, the assumption that privacy is always and unambiguously welfare enhancing cannot be said to be true.193 Users can miss out on opportunity costs by refusing to give access to the IDFA. By the same token, not prioritising privacy for users can mean that discrimination and manipulation can be easily conducted through advertising. F. Objective justification and anticompetitive effects The last step to take once we have settled that Apple’s conduct produces anticompetitive effects, at least on the advertising and attribution front, is to assess whether the behaviour is tolerable for competition law. Apple has built an ecosystem where third-party apps and its own services coexist for the benefit of the consumer, and the firm is now promoting its image as a privacy enabler.194 Apple acts as a private regulator, setting out the guidelines and rules, which app developers must abide to maintain the ecosystem’s governance.195 In 2020, over 215 thousand app submissions were rejected by Apple for violating privacy guidelines.196 Based on the role of Apple as the ecosystem holder, anticompetitive effects can be balanced out or objectively justified.197 The question is whether Apple can justify its behaviour on the basis of enhancing privacy because there was a risk for users in terms of their privacy before the iOS 14.5 update was introduced. Therefore, the validity of the practice’s justification must be assessed against its proportionality and effect in practice. Thus, we will be able to identify if the underlying motive to the iOS 14.5 privacy update is that of vertical foreclosure against third-party apps’ free ad-based model.198 According to the Hilti case, the General Court discarded as a valid justification the dangerous nature or inferior quality of competitor products in front of their foreclosure. Hilti argued that the practice sook to eliminate competing products from the market because they were dangerous to the public. The Court did not accept the incumbent’s justification because there were laws in place on product liability and misleading advertising, which already solved this problem.199 Hence, the objective justification to an anticompetitive behaviour cannot be built on overriding problems that have already been addressed by existing laws. The same must be true for Apple.200 Since the GDPR came into force, it has become the benchmark to determine whether users can be identified and are in control of their personal data before data holders and controllers. Moreover, it is also the legal regime applicable to establish whether data are processed for a legitimate purpose, in a lawful, fair, and transparent way. Therefore, Apple cannot use data protection as an excuse to foreclose the activity of its competitors on the downstream market, even if the ATT prompt is thought as an additional layer to protect privacy on top of the ‘Limit Ad Tracking’ option, in the terms of the French competition authority decision.201 The complainants on the Interim measures on iOS 14.5 decision rendered by the French authority voiced their concerns on the redundancy and illegitimacy of the ATT prompt. As a matter of fact, the changes on the privacy policy only apply to Apple’s competitors in the downstream market, because of a decision produced by the ecosystem holder. This does not seem to be proportionate to the aim that was—allegedly—pursued by Apple, privacy protection. In fact, users are overexposed in front of Apple’s overpowering reach on the data stored within the iPhone, and it does nothing to protect them from it. However, based on its interpretation of the policy, third-party apps pose a great risk to privacy by using the IDFA and supply should be cut off as a rule. In sum, there is no proportionate or effective reasoning that justifies Apple’s anticompetitive behaviour other than the foreclosure of third-party apps from the downstream market. VI. Conclusions In 2020 Apple announced its iOS 14.5 update. From June 2020 to April 2021, third-party apps have been given time to adapt to the changes that the update would entail. However, one of the core ‘improvements’ of the update, i.e., the privacy policy update, forces a shift on their business model. Apple holds, designs, and governs the ecosystem on iOS devices. Through its App Store, apps can be downloaded by users, both developed by Apple’s proprietary services and independent app developers. Its position at the upstream market of the distribution of iOS apps is quasi-monopolistic, given that third-party app stores do not pose a realistic threat to it. Apple is increasingly expanding and financing its services on its proprietary apps, through Apple News+, Apple Arcade, etc. Thus, Apple is also present on the downstream market of retail, where apps are downloaded. For third-party apps, the playing field is not levelled. Their main source of revenue depends on advertising. They offer advertisers their in-app space to display an ad or a video in-app ad to attract the user to the offering. This was possible only due to attribution: advertiser and app developer/publisher matched up the interaction produced by the user to the advertisement that had been displayed. The IDFA was the main mechanism used for this purpose, especially by those developers with scarce economic and financial resources. However, this is no longer possible for them since the iOS 14.5 update effectively hinders their possibility of access to the IDFA. Only if the user opts-in through the ATT prompt to enable ‘tracking’ they will have access to it. Even assuming that user consent is granted, the possibility of performing attribution is challenged, considering that not all third-party apps will be granted consent, and each user will adjust the rendering of consent to his own preferences on privacy. The wide array of problems that have been outlined throughout do not apply to Apple or its proprietary services. Instead, Apple is authorised to personalise advertising, by default, when the iPhone is first set up. In addition, its own Apple Search Ads feeds off attribution through alternative methods to the IDFA and IDFV. Considering this scenario, the French competition authority found that Apple implemented its privacy policy under objective, transparent, and non-discriminatory conditions. It ruled in favour of the ATT prompt based on its complementarity to other data protection methods that were already embedded within the iPhone. In my view, the French competition authority errs in its judgment since the privacy operational features available for third-party app developers and Apple are conflated. They have different mechanisms at hand within their toolkit to enable the user to disallow ‘tracking’, thus, the operational advantages in Apple’s hands in terms of personalised advertising should not be disregarded. The iOS 14.5 update amounts to an implicit refusal of access to the IDFA to third-party apps. That is, it forecloses its competition on the downstream market through an update, which is intended to be justified because of platform governance. The cornerstone to this line of thought is the element of indispensability, which has been redefined recently through the Google Shopping judgment. The economic rationale behind the high threshold placed on the refusal to supply is not applicable in this case, given that we are in front of a constructive, and not of an outright refusal to supply, as well as due to the fact that anticompetitive effects do not derive directly from Apple’s refusal but from consumer choice repeated by the millions on the device when the update is downloaded. Bearing in mind the abovementioned, we assessed the anticompetitive effects of Apple’s conduct: the elimination of competition. Stemming from the Microsoft judgment, and different to the Bronner case, these effects should be considered on the basis of their likeliness, and not in relation to the elimination of all effective competition. On one hand, third-party apps are placed in a position where personalised advertising and content as well as attribution are hindered effectively. The business model they rely on—the free ad-based business model—is no longer viable and they will have to produce their revenue elsewhere. On the other hand, from the user perspective, the privacy policy update results in a decrease on the quality of personalised advertising on in-app ads. Although this reasoning is commonly accepted, it is still contested that an increase on privacy protection will always be beneficial for consumer welfare. In sum, Apple’s iOS 14.5 privacy update infringes article 102 TFEU, absent any objective justification. This is a peculiar scenario where privacy is used in the guise of platform governance. The aim of the update through the ATT prompt is to incentivise a shift on the business model of third-party apps towards a fee-based model, which will be the most beneficial for Apple and its proprietary services. Footnotes 1 Apple, ‘Apple Reimagines the iPhone Experience with iOS 14’ (Apple NewsRoom, 22 June 2020), available at https://www.apple.com/newsroom/2020/06/apple-reimagines-the-iphone-experience-with-ios-14/ accessed 20 July 2021. 2 David S. Evans, ‘Governing Bad Behavior by Users of Multi-Sided Platforms’ (2012) University of Chicago Institute for Law & Economics Olin Research Paper No. 582, available at https://ssrn.com/abstract=1950474 accessed 20 July 2021; Anja Lambrecht, Alessandro Bonatti, Avi Goldfarb. Anindya Ghose, Daniel G. Goldstein, Randall Lewis, Anita Rao, Navdeep Shani and Song Yao ‘How Do Firms Make Money Selling Digital Goods Online?’ (9th Triennial Choice Symposium, Noordwijk, 2014) https://www.researchgate.net/publication/278393875_How_do_firms_make_money_selling_digital_goods_online accessed 29 July 2021. 3 Yoram Wurmser, ‘US Time Spent with Mobile 2019’ (Insider Intelligence, 30 May 2019), available at https://www.emarketer.com/content/us-time-spent-with-mobile-2019 accessed 5 August 2021. When referring to the App Store (upper case), it is in reference to Apple’s App Store, and when referred as to in lowercase it is to the general distribution system of apps (including Android). 4 The term ‘third-party app’ is used in reference to the set of apps which can be downloaded through the App Store developed by independent developers, different to Apple’s proprietary apps. 5 Autorité de la Concurrence Decision of 17 March 2021 in request for interim measures submitted by the associations Interactive Advertising Bureau France, Mobile Marketing Association France, Union Des Entreprises de Conseil et Achat Media, and Syndicat des Régies Internet in the sector of advertising on mobile apps on iOS (21-D-07). Will be referenced further as Interim measures on iOS 14.5. 6 Javier Espinoza, ‘German Groups File Apple Antitrust Complaint as It Makes Privacy Changes’ Financial Times (Brussels, 26 April 2021); Office of Competition and Consumer Protection, ‘Apple—the President of UOKiK Initiates an Investigation’ (Warsaw, 13 December 2021). 7 Competition and Markets Authority, ‘Mobile Ecosystems: Market Study Interim Report’ 14 December 2021, para 6.23; Dieter Bohn, ‘Apple Isn’t Just a Walled Garden, It’s a Carrier’ (The Verge, 7 June 2021), https://www.theverge.com/2021/6/7/22521476/apple-walled-garden-carrier-app-store-innovation accessed 9 February 2022. 8 Competition and Markets Authority ‘Appendix D: Financial Analysis of Apple’s and Google’s Mobile Ecosystems’ 14 December 2021, para 8. 9 Commission Decision of 18 July 2018 in Google Android (Case COMP/AT.40099), para 287. 10 Stephanie Chan, ‘Global Consumer Spending in Mobile Apps Reached a Record $111 Billion in 2020, Up 30% from 2019’ (Sensor Tower Blog, 4 January 2021), available at https://sensortower.com/blog/app-revenue-and-downloads-2020 accessed 10 August 2021. 11 Apple, ‘Enrollment’ (Apple Developer), https://developer.apple.com/support/enrollment/ accessed 5 August 2021. 12 Statista Research Department, ‘Number of Available Apps in the Apple App Store from 2008 to 2020’ (Statista, 6 July 2021) https://www.statista.com/statistics/268251/number-of-apps-in-the-itunes-app-store-since-2008/ accessed 30 July 2021. 13 Friso Bostoen and Daniel Mandrescu, ‘Assessing Abuse of Dominance in the Platform Economy: A Case Study of App Stores’ (2020) 16 European Competition Journal 2020 455. 14 Apple, ‘Apple One Makes Enjoying Apple Subscription Services Easier Than Ever’ (Apple Newsroom, 15 September 2020) https://www.apple.com/newsroom/2020/09/apple-one-makes-enjoying-apple-subscription-services-easier-than-ever/ accessed 4 August 2021. 15 Apple, ‘Apple Ads Attribution API’ (Apple Search Ads) https://searchads.apple.com/help/reporting/0028-apple-ads-attribution-api accessed 8 January 2022. An API is a piece of software that facilitates communications between different applications. 16 Bostoen and Mandrescu (n 13) 446. 17 Bostoen and Mandrescu (n 13) 450–453. 18 Apple, ‘Apple Announces App Store Small Business Program’ Newsroom (Cupertino, 18 November 2020). 19 Competition and Markets Authority ‘Appendix D: Financial Analysis of Apple’s and Google’s Mobile Ecosystems’ 14 December 2021, para 21. 20 Frederic Jenny, ‘Competition Law Enforcement and Regulation for Digital Ecosystems: Understanding the Issues, Facing the Challenges and Moving Forward’ (2021), available at https://ssrn.com/abstract=3857507 accessed 3 January 2022. 21 European Commission, ‘Antitrust: Commission Opens Investigations into Apple’s App Store rules’ IP/20/1073 (Brussels, 16 June 2020); Bundeskartellamt, ‘Proceeding against Apple based on new rules for large digital companies (Section 19a(1) GWB)—Bundeskartellamt examines Apple’s significance for competition across markets (Bonn, 21 June 2021). 22 AdColony, ‘The AdColony Mobile Publishing Survey’ (AdColony, 2020), https://www.adcolony.com/blog/2020/02/12/top-publishers-rely-on-video-ads-for-mobile-monetization/ accessed 29 July 2021. 23 John Thorne, ‘Reviving Competition, Part 1: Proposals to Address Gatekeeper Power and Lower Barriers to Entry Online’ (U.S. House of Representatives Committee on the Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Washington D.C., 25 February 2021), https://docs.house.gov/meetings/JU/JU05/20210225/111247/HHRG-117-JU05-Wstate-ThorneJ-20210225.pdf accessed 30 July 2021; Christophe Carugati, ‘The Antitrust Privacy Dilemma’ (2021), available at https://ssrn.com/abstract=3968829 accessed 28 December 2021. 24 D. Daniel Sokol and Feng Zhu, ‘Harming Competition and Consumers under the Guise of Protection Privacy: An Analysis of Apple’s iOS 14 Policy Updates’ (2021) USC CLASS Research Paper 21–27, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3852744 accessed 30 July 2021; Wilko Bolt and Alexander F. Tieman, ‘Heavily Skewed Pricing in Two-Sided Markets’ (2008) 26 International Journal of Industrial Organization 1251. 25 David S. Evans and Richard Schmanlensee, ‘The Antitrust Analysis of Multi-sided Platform Businesses’ (2013) National Bureau of Economic Research 18783, available at https://www.nber.org/papers/w18783 accessed 5 August 2021; Bundeskartellamt, ‘The Market Power of Platforms and Networks’ (2016) Working Paper B6–113/15, available at https://www.bundeskartellamt.de/SharedDocs/Publikation/EN/Berichte/Think-Tank-Bericht-Zusammenfassung.html accessed 5 August 2021. 26 Bolt and Tieman (n 24) 1251. 27 Inge Graef, Sih Yuliana Wahyuningtyas and Peggy Valcke, ‘Assessing Data Access Issues in Online Platforms’ (2015) 39 Telecommunications Policy 2015 379; Evans and Schmanlensee (n 21). 28 Statista Research Department, ‘Distribution of Free and Paid iOS Apps 2019–2021’ (Statista, 6 July 2021) https://www.statista.com/statistics/1020996/distribution-of-free-and-paid-ios-apps/ accessed 10 July 2021; Kyle Andeer ‘Questions for the Record from the Honorable David N. Cicilline, Part 2: Innovation and Entrepreneurship’ (U.S. House of Representatives Committee on the Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, 16 July 2019), https://docs.house.gov/meetings/JU/JU05/20190716/109793/HHRG-116-JU05-20190716-SD036.pdf accessed 6 August 2021. 29 AdColony (n 22). 30 Sigurd Næss-Schmidt, Bruno Basalisco, Signe Rølmer, Katrine Poulsgaard, Benjamin Barnet Christiansen and Lærke Kilsdal, ‘Empowering the European Business Ecosystem: An Impact Study of Businesses Using Facebook Apps and Technologies’ (2020) https://www.copenhageneconomics.com/dyn/resources/Publication/publicationPDF/2/522/1579535391/empowering-the-european-business-ecosystem_copenhagen-economics.pdf accessed 29 July 2021. 31 Howard Beales and Jeffrey A. Eisenach, ‘An Empirical Analysis of the Value of Information Sharing in the Market for Online Content’ (2014), available at https://ssrn.com/abstract=2421405. 32 Bostoen and Mandrescu (n 13) 461. 33 Stephen Dnes, ‘Browser Tying and Data Privacy Innovation’ (2021), available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3867625 accessed 19 July 2021; Beales and Eisenach (n 34). 34 Google Android (n 9), para 328. 35 Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 19), para 20. 36 Lambrecht, Bonatti, Goldfarb, Ghose, Goldstein, Lewis, Rao, Shani and Yao (n 2). 37 Google Android (n 9), para 110. 38 Tami Kim, Kate Barasz and Leslie K. John, ‘Why Am I Seeing This Ad? The Effect of Ad Transparency on Ad Effectiveness’ (2019) 45 Journal of Consumer Research 907; Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 19), para 33. 39 Owen Ray, ‘What is IDFA and Why Apple Killed it: Everything Marketers Need to Know’ (Invoca, 14 September 2020), available at https://www.invoca.com/blog/what-is-idfa-and-why-apple-killed-it-everything-marketers-need-to-know accessed 10 August 2021. 40 Garrett Johnson, Scott Shriver and Shaoyin Du, ‘Consumer Privacy Choice in Online Advertising: Who Opts Out and at What Cost to Industry?’ (2019) Simon Business School Working Paper No. 17–19, available at https://ssrn.com/abstract=3020503 accessed 11 August 2021. 41 AdColony (n 22). 42 European Parliamentary Research Service, ‘Digital Service Act: European Added Value Assessment’ (2020); Nico Neumann, Catherine E. Tucker and Timothy Whitfield, ‘How Effective Is Third-Party Consumer Profiling and Audience Delivery? Evidence from Field Studies’ (2019) Marketing 38(6) Marketing Science 918. accessed 29 July 2021. Kim, Barasz and John (n 38) 906. 43 Thorne (n 23). 44 Interim Measures on iOS 14.5 (n 5) para 124. 45 ibid., paras 119–121. 46 Jenny (n 20); Hal Varian, ‘Use and Abuse of Network Effects’ (2017), available at https://ssrn.com/abstract=3215488 accessed 20 August 2021. 47 Johnson, Shriver and Du (n 40). 48 Competition and Markets Authority, ‘Online Platform and Digital Advertising Market Study’ Appendix O: measurement issues in digital advertising. 49 Avi Goldfarb and Catherine Tucker, ‘Online Display Advertising: Targeting and Obstrusiveness’ (2010) 30 Marketing Science 389. 50 AdColony (n 22). 51 Deloitte LLP, ‘Digital Tools in Crisis and Recovery: Small and Medium Business Report’ (Deloitte, October 2020), https://about.fb.com/wp-content/uploads/2020/10/Deloitte-Digital-Tools-in-Crisis-and-Recovery-SMB-Report-Oct-2020.pdf accessed 3 August 2021; Næss-Schmidt, Basalisco, Rølmer, Poulsgaard, Barnet and Kilsdal (n 30). 52 Thomas Hoppner and Philipp Westerhoff, ‘Privacy by Default, Abuse by Design: EU Competition Concerns About Apple’s New App Tracking Policy’ (2021), available at https://ssrn.com/abstract=3853981 accessed 10 October 2021. 53 Meta for Developers, ‘Facebook APIs, SDKs & Guides’ (Facebook Developer Docs) https://developers.facebook.com/docs/ accessed 8 January 2022; Developer Android, ‘Developer Guides’ (Documentation) https://developer.android.com/guide accessed 8 January 2022. 54 Jenny (n 20). 55 Apple (n 1). 56 Sokol and Zhu (n 24); Hoppner and Westerhoff (n 52); Maya Bar-Hillel, Eyal Peer and Alessandro Acquisiti, ‘“Heads or Tails?”—A Reachability Bias in Binary Choice’ (2014). April 2014 Journal of Experimental Psychology: Learning, Memory and Cognition http://dx.doi.org/10.1037/xlm0000005 accessed 8 January 2022. 57 Apple, ‘User Privacy and Data Use’ (Apple Developer, 2021), https://developer.apple.com/app-store/user-privacy-and-data-use/ accessed 30 July 2021. 58 Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 38), para 118; Leslie K. John, Tami Kim and Kate Barasz, ‘Ads That Don’t Overstep’ (Harvard Business Review, January–February 2018), available at https://hbr.org/2018/01/ads-that-dont-overstep accessed 12 January 2022. 59 Hoppner and Westerhoff (n 52). 60 Thorne (n 23). 61 Apple, ‘Apple Advertising & Privacy’ (Support Apple, 2021), https://support.apple.com/en-us/HT205223 accessed 8 January 2022. 62 Hoppner and Westerhoff (n 52); Punam Anand Keller, Bari Harlam, George Loewenstein and Kevin G. Volpp, ‘Enhanced Active Choice: A New Method to Motivate Behavior Change’ (2011) 21 Journal of Consumer Psychology 377 376. 63 Patrick McGee, ‘Apple Reaches Quiet Truce over iPhone Privacy Changes’ Financial Times (San Francisco, 8 December 2021). 64 Kendra Clark, ‘Apple’s iOS 15: ‘Enhanced User Privacy to the Detriment of Marketers & Publishers’ (The Drum, 17 September 2021), https://www.thedrum.com/news/2021/09/17/apple-s-ios-15-enhanced-user-privacy-the-detriment-marketers-publishers accessed 9 February 2022. 65 Deepak Ravichandran and Nitish Korula, ‘Effect of Disabling Third-Party Cookies on Publisher Revenue’ (Google, 27 August 2019), available at https://services.google.com/fh/files/misc/disabling_third-party_cookies_publisher_revenue.pdf accessed 30 August 2021. 66 Goldfarb and Tucker (n 49) 393. 67 Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 19), para 57. 68 Interim Measures on iOS 14.5 (n 5), paras 27–28. 69 Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 19), para 60. 70 Interim Measures on iOS 14.5 (n 5), para 30; Hoppner and Westerhoff (n 52). 71 Keller, Harlam, Loewenstein and Volpp (n 62) 377. 72 Sokol and Zhu (n 24). 73 Sokol and Zhu (n 24). 74 Hoppner and Westerhoff (n 52). 75 Thorne (n 23). 76 Pan Katsukis, Sergio Serra and Umair Malik, ‘ID Or No ID? Tracking the Post-IDFA Changes in the World of Programmatic Mobile Advertising on iOS’ (Remerge, 24 October 2021), https://post-idfa-dashboard.remerge.io/, accessed 25 October 2021. 77 Sokol and Zhu (n 24); Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 19), para 76. 78 Thorne (n 23); Dnes (n 33); Competition and Markets Authority, ‘Mobile Ecosystems Market Study’ (n 19), paras 77–79. 79 Case T-612/17 Google LLC, Formerly Google Inc. and Alphabet, Inc. v European Commission, EU:T:2021:763, para 179 (Google Shopping onwards). 80 Pablo Ibáñez Colomo, ‘What is an Abuse of a Dominant Position? Deconstructing the Prohibition and Categorizing Practices’ Pinar Akman, Or Brook and Konstantinos Stylianou (eds), Research Handbook on Abuse of Dominance and Monopolization (Edward Elgar 2022, forthcoming). 81 Opinion of AG Rantos in Case C-377/20 Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante Della Concorrenza e del Mercato and Others, para 56. 82 Case C-307/18 Generics (UK) Ltd, GlaxoSmithKlineplc, Xellia Pharmaceuticals ApS, Alpharma LLC, Actavis UK Ltd, Merck KGaA v. Competition and Markets Authority, EU:C:2020:52, para 148; Opinion of AG Rantos in Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante Della Concorrenza e del Mercato and Others (n 81), para 49; Google Shopping (n 79), para 155. 83 Case T-286/09 Intel Corp. v. European Commission, EU:T:2014:547; Colomo (n 80); Google Shopping (n 79), para 162. 84 Case C-62/86 AKZO Chemie BV v. Commission, EU:C:1991:286, para 71; Google Shopping (n 79), para 157. 85 Colomo (n 80); Opinion of AG Rantos in Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante Della Concorrenza e del Mercato and Others (n 81), paras 61 and 62. 86 OECD, Quality Considerations in Digital Zero-Price Markets’ DAF/COMP(2018)14, October 2018. 87 Colomo (n 80). 88 Guidance on the Commission’s Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings [2009] OJ C45/7, para 76; Inge Graef, ‘EU Competition Law, Data Protection and Online Platforms. Data as Essential Facility’ (PhD, KU Leuven 2016); Vikas Kathuria and Jure Globocnik, ‘Exclusionary Conduct in Data-Driven Markets: Limitations of Data Sharing Remedy’ (2020) 8 Journal of Antitrust Enforcement 511; Google Shopping (n 79), para 163. 89 Wen Wen and Feng Zhu, ‘Threat of Platform-Owner Entry and Complementor Responses: Evidence from the Mobile App Market’ (2016) NET Institute Working Paper No. 16–10, available at https://ssrn.com/abstract=2848533 accessed 29 August 2021. 90 Bostoen and Mandrescu (n 13) 468–475. 91 Commission Decision of 17 December 2020 in Google/Fitbit (Case COMP/M. 9660), para 48; Google Android (n 9), paras 306–322. 92 Google Android (n 9), para 270. 93 Interim measures on iOS 14.5 (n 5), paras 110 and 112. 94 Google Android (n 9), paras 284–305 and 318. 95 Interim measures on iOS 14.5 (n 5), para 117. 96 Interim measures on iOS 14.5 (n 5), paras 102–129. 97 Interim measures on iOS 14.5 (n 5), para 125. 98 Interim measures on iOS 14.5 (n 5), para 129. 99 Google/Fitbit (n 91), para 177. 100 Bundeskartellamt (n 25); Evans and Schmanlensee (n 25); Varian (n 46); Google Android (n 9), paras 591–614. 101 European Parliamentary Research Service, ‘European App Economy: State of Play, Challenges and EU policy’ PE 621.894 (May 2018); Justus Haucap and Ulirch Hemisehoff, ‘Google, Facebook, Amazon, eBays: Is the Internet Driving Competition or Market Monopolization?’ (2014) 11 International Economics and Economic Policy 49; Bundeskartellamt (n 25); Competition and Markets Authority, ‘Mobile Ecosystems: Market Study Interim Report’ (n 9), para 8.20. 102 Hoppner and Westerhoff (n 52). 103 Bostoen and Mandrascu (n 13) 442. 104 Interim measures on iOS 14.5 (n 5), para 130; European App Economy: State of Play, Challenges and EU policy (n 101). 105 Pallavi, ‘A Descriptive List of Alternative App Stores for 2021’ (Mobile App Daily, 24 May 2021), available at https://www.mobileappdaily.com/app-stores-list accessed 12 August 2021. 106 Google Shopping (n 79), para 180. 107 Interim measures on iOS 14.5 (n 5), para 134. 108 Interim measures on iOS 14.5 (n 5), para 138. 109 Interim measures on iOS 14.5 (n 5), paras 144–147, 56, and 64. 110 European Parliament and Council Regulation (EU) No 2016/679 of 27 April 2016 on the Protection of Natural Persons With Regard to the Processing of Personal Data and on the Free Movement of Such Data, and Repealing Directive 95/46/EC [2016] OJ L 119. 111 Interim measures on iOS 14.5 (n 5), paras 169 and 155. 112 Kim, Barasz and John (n 38) 918. 113 Opinion of AG Rantos in Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante della Concorrenza e del Mercato and others (n 81), para 76. 114 Case C-7/97 Oscar Bronner GmbH & Co. KG and Mediaprint Zeitungs-und Zeitschriftenverlag GmbH & Co. KG, Mediaprint Zeitungsvertriebsgessellschaft mbH & Co. KG, Mediaprint Anzeigengesellschaft mbH & Co. KG, EU:C:1998:569. 115 Google Shopping (n 79), para 155. 116 Bronner (n 114), para 41. 117 Case 238/87 AB Volvo and Erik Veng (UK Ltd), EU:C:1988:477 para 8. 118 Opinion of AG Saugmandsgaard Øe in Cases C-152/19 P and C-165/19 P, Deutsche Telekom AG and Slovak Telekom a.s. v. European Commision, EU:C:2020:678, paras 76–79; Case C-165/19 P Slovak Telekom a.s., v. European Commission and Slovanet, a.s., EU:C:2021:239, para 47; Google Shopping (n 79), para 217. 119 Google Shopping (n 79), para 224. 120 Google Shopping (n 79), para 232. 121 Jose Rivas, ‘How Indispensable is the Indispensability Criterion in Cases of Refusal to Supply Competitors by Dominant Companies? (Slovak Telekom, C-165/19 P)’ (Kluwer Competition Law Blog, 1 April 2021), available at http://competitionlawblog.kluwercompetitionlaw.com/2021/04/01/how-indispensable-is-the-indispensability-criterion-in-cases-of-refusal-to-supply-competitors-by-dominant-companies-slovak-telekom-c-165-19-p/ accessed 11 August 2021. 122 Google Shopping (n 79), para 233. 123 Friso Bostoen, ‘The General Court’s Google Shopping judgment. Finetuning the legal qualifications and test for platform abuse’ 13(2) Journal of European Competition Law & Practice 75. 124 Case C-52/09 Konkurrensverket v TeliaSonera Sverige AB, EU:C:2011:93, paras 55 and 56. 125 Lena Hornkohl, ‘Article 102 TFEU, Equal Treatment and Discrimination after Google Shopping’ 13(2) Journal of European Competition Law & Practice 99. 126 Google Shopping (n 79), paras 560–562. 127 Bostoen (n 123). 128 Digital Advertising Alliance, ‘Consumer Value Ad Supported Services’ (Digital Advertising Alliance, September 2020), https://digitaladvertisingalliance.org/sites/aboutads/files/DAA_files/Consumer-Value-Ad-Supported-Services-2020Update.pdf, accessed 3 August 2021; Network Advertising Initiative, ‘Consumer Survey on Privacy and Digital Advertising) (Digital Advertising Alliance, 22 October 2019), available at https://www.networkadvertising.org/sites/default/files/final_nai_consumer_survey_paper_22oct2019.pdf accessed 3 August 2021. 129 Kim, Barasz and John (n 38) 909. 130 Savanta ‘What Would an Internet Without Targeted Ads Look Like? (IAB Europe, December 2020), https://iabeurope.eu/knowledge-hub/iab-research-what-would-an-internet-without-targeted-ads-look-like/ accessed 29 July 2021; Erin Egan and Steve Satterfield, ‘A Path Forward for Privacy and Online Advertising’ (Facebook Newsroom, 2 October 2020), https://about.fb.com/news/2020/10/a-path-forward-for-privacy-and-online-advertising/ accessed 25 October 2021. 131 Michal S. Gal and Oshirt Aviv, ‘The Competitive Effects of the GDPR’ (2020) 16 Journal of Competition Law & Economics 349, 384. 132 Samson Esayas, ‘Privacy-As-A-Quality Parameter: Some Reflections on the Scepticism’ (2017) Faculty of Law, Stockholm University Research Paper No. 43, available at https://ssrn.com/abstract=3075239 accessed 17 August 2021. 133 Joseph Farrell, ‘Can privacy be just another good?’ (2012) 10 Journal on Telecommunication & High Technology Law 251; Lambrecht, Bonatti, Goldfarb, Ghose, Goldstein, Lewis, Rao, Shani and Yao (n 2). 134 Mike Peterson, ‘Apple News Could Reach 19M Subscribers, $2.2B in Revenue by 2023, Analyst Says’ (appleinsider, 6 April 2021), available at https://appleinsider.com/articles/21/04/06/apple-news-could-reach-19m-subscribers-22b-in-revenue-by-2023-analyst-says accessed 25 July 2021; Apple, ‘Apple Arcade Launches Its Biggest Expansion Yet, Growing Its Award-Winning Catalog to More Than 180 Games’ (Apple Newsroom, 2 April 2021), available at https://www.apple.com/newsroom/2021/04/apple-arcade-expands-its-award-winning-catalog-to-more-than-180-games/ accessed 25 July 2021. 135 Catherine Tucker and Avi Goldfarb, ‘Privacy and Innovation’ (2012) 12 Innovation Policy and the Economy 65; Digital Advertising Alliance (n 128); Savanta (n 130). 136 Esayas (n 132). 137 Dan Levy, ‘Speaking Up for Small Businesses’ (Facebook for Business, 16 December 2020) available at https://www.facebook.com/business/news/ios-14-apple-privacy-update-impacts-small-business-ads accessed 28 August 2021; Network Advertising Initiative (n 128). 138 Daniel Rubinfeld and Michal S. Gal, ‘Access Barriers to Big Data’ (2017) 59 Arizona Law Review 353–356. 139 Competition & Markets Authority, The Commercial Use of Consumer Data: Report on the CMA’s Call for Information (CMA38, June 2015); Næss-Schmidt, Basalisco, Rølmer, Poulsgaard, Barnet and Kilsdal (n 30). 140 An SDK is a set of tools for third-party developers to use in producing applications using a particular framework or platform. 141 Google Digital Advertising Antitrust Litigation, 5:20-cv-03556 (2020). 142 Opinion of AG Rantos in Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante della Concorrenza e del Mercato and Others (n 81), para 71. 143 Rubinfeld and Gal (n 138) 350–367. 144 Sokol and Zhu (n 24); Lambrecht, Bonatti, Goldfarb, Ghose, Goldstein, Lewis, Rao, Shani and Yao (n 2). 145 Thorne (n 23). 146 Evans and Schmanlensee (n 25); Interim Measures on iOS 14.5 (n 5) para 87. 147 Graef, Wahyuningtyas and Valcke (n 27) 381. 148 Evans (n 2). 149 Hoppner and Westerhoff (n 52). 150 Marshall W. Van Alstyne, Geoffrey G. Parker and Sangeet Paul Choudary, ‘Pipelines, platforms, and the New Rules of Strategy’ (2016) April 2016 Issue Harvard Business Review. 151 Dennis Sellers, ‘92% of iPhone Users Likely to Upgrade in the Next 12 Months’ (Apple World Today, 17 May 2017), available at https://appleworld.today/2017/05/17/92-of-iphone-users-likely-to-upgrade-in-the-next-12-months/ accessed 4 September 2021. 152 Interim measures on iOS 14.5 (n 5), para 111; Lambrecht, Bonatti, Goldfarb, Ghose, Goldstein, Lewis, Rao, Shani and Yao (n 2); Network Advertising Initiative (n 128). 153 Hoppner and Westerhoff (n 52). 154 JR Raphael, ‘iPhone to Android: The Ultimate Switching Guide’ (Computerworld, 7 February 2020), available at https://www.computerworld.com/article/3218067/how-to-switch-from-iphone-to-android-ultimate-guide.html accessed 12 August 2021. 155 Documents from the Hearing on Online Platforms and Market Power: Examining the Dominance of Amazon, Apple, Facebook, and Google (U.S. House of Representatives Committee on the Judiciary Subcommittee on Antitrust, Commercial and Administrative Law) https://judiciary.house.gov/uploadedfiles/014701.pdf accessed 14 August 2021; Interim measures on iOS 14.5 (n 5), para 115. 156 Interim measures on iOS 14.5 (n 5) para 86; Keller, Harlam, Loewenstein and Volpp (n 62) 382; Sokol and Zhu (n 24). 157 Interim measures on iOS 14.5 (n 5) para 38. 158 Sokol and Zhu (n 24). 159 Interim measures on iOS 14.5 (n 5) para 39; Patrick McGee, ‘Apple’s Privacy Changes Create Windfall for Its Own Advertising Business’ Financial Times (San Francisco, 17 October 2021); Competition and Markets Authority, ‘Mobile Ecosystems Market Study’, para 95. 160 Eric Benjamin Seufert, ‘ATT Advantages Apple’s Ad Network. Here’s How to Fix That’ (Mobile Dev Memo, 1 November 2021) https://mobiledevmemo.com/att-advantages-apples-ad-network-heres-how-to-fix-that/ accessed 8 January 2022. 161 McGee (n 159). 162 Graef (n 88). 163 Charter of Fundamental Rights of the European Union [2012] OJ C 326. 164 Francisco Costa-Cabral and Orla Lynskey, ‘Family Ties: the Intersection between Data Protection and Competition in EU law’ (2017) 54 Common Market Law Review 11. 165 Esayas (n 132); Costa-Cabral and Lynskey (n 164) 15–16. 166 Gal and Aviv (n 131) 353. 167 Kim, Barasz and John (n 38) 918. 168 Alessandro Acquisiti, Curtis Taylor and Liad Wagman, ‘The Economics of Privacy’ (2016) 52 Journal of Economic Literature 442; Farrell (n 133). 169 Acquisiti, Taylor and Wagman (n 168) 448; Kim, Barasz and John (n 38) 920. 170 Esayas (n 132). 171 Patricia A. Norberg, Daniel R. Horne and David A. Horne, ‘The Privacy Paradox: Personal Information Disclosure Intentions versus Behaviors’ (2007) 41 Journal of Consumer Affairs 100; Susanne Barth and Menno D.T. de Jong, ‘The Privacy Paradox—Investigating Discrepancies between Expressed Privacy Concerns and Actual Online Behavior—A Systematic Literature Review’ (2017) 34 Telematics and Informatics 1038; Allen P. Grunes and Maurice E. Stucke, ‘No Mistake About It: The Important Role of Antitrust in the Era of Big Data’ (2015) University of Tennessee Legal Studies Research Paper No. 269, available at https://ssrn.com/abstract=2600051 accessed 12 January 2022. 172 Andrew F. Daughety and Jennifer F. Reinganum, ‘Public Goods, Social Pressure, and the Choice between Privacy and Publicity’ (2010) 2 American Economic Journal: Microeconomics 191. 173 Acquisiti, Taylor and Wagman (n 168) 447. 174 Opinion of AG Rantos in Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante della Concorrenza e del Mercato and others (n 81), para 107. 175 Barth and de Jong (n 171) 1046–1047; Competition and Data Protection in Digital Markets: A Joint Statement between the CMA and the ICO, 19 May 2021, paras 50–59. 176 Patricia Lorenzo, Jorge Padilla and Alejandro Requejo, ‘Consumer Preferences for Personal Data Protection in Social Networks: A Choice Modelling Exercise’ (2020), available at: https://ssrn.com/abstract=3716206 accessed 4 January 2022. 177 Dnes (n 33); Daniel. R. Feenberg, Ina Ganguli, Patrick Gaule, Jonathan Gruber ‘It’s Good to be First: Order Bias in Reading and Citing’ (2015) NBER Working Paper 21141 http://www.nber.org/papers/w21141 accessed 8 January 2022; Lorenzo, Padilla and Requejo (n 176). 178 Pinar Akman, ‘A Web of Paradoxes: Empirical Evidence on Online Platform Users and Implications for Competition and Regulation in Digital Markets’ (2022) 16(2) Virginia Law and Business Review 217. 179 Cisco, ‘Consumer Privacy Survey: The Growing Imperative of Getting Data Privacy Right’ (Cisco Cybersecurity Series 2019, November 2019)https://www.cisco.com/c/dam/global/en_uk/products/collateral/security/cybersecurity-series-2019-cps.pdf accessed 17 July 2021; Geoffrey A. Manne and Ben Sperry, ‘The Problems and Perils of Bootstrapping Privacy and Data into an Antitrust Framework’ (CPI Antitrust Chronicle, 29 May 2015), available at https://ssrn.com/abstract=2617685 accessed 28 July 2021; Evans and Schmanlensee (n 25); Adam S. Chilton and Omri Ben-Shahar, ‘Simplification of Privacy Disclosures: An Experimental Test’ (2016) Coase-Sandor Working Paper Series in Law and Economics, No. 737, available at https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?article=2443&context=law_and_economics accessed 20 August 2021; Jane Bambauer, Jonathan Loe and D. Alex Winkelman, ‘A Bad Education’ (2017) 2017 University of Illinois Law Review 109; Opinion of AG Rantos in Servizio Elettrico Nazionale SpA, Enel SpA, Enel Energia SpA v Autorità Garante della Concorrenza e del Mercato and others (n 81), paras 96 and 97; Lorenzo, Padilla and Requejo (n 176). 180 Interim measures on iOS 14.5 update (n 5) para 84; Barth and de Jong (n 171) 1051. 181 Justin P. Johnson, ‘Target advertising and advertising avoidance’ (2013) The RAND Journal of Economics 44 (1), 128. 182 Opinion of the European Data Protection Supervisor on the Communication from the Commission to the European Parliament and the Council on the follow-up of the Work Programme for better implementation of the Data Protection Directive (Brussels, 25 July 2007); Lambrecht, Bonatti, Goldfarb, Ghose, Goldstein, Lewis, Rao, Shani and Yao (n 2). 183 Bambauer, Loe and Winkelman (n 179) 162–164. 184 Leslie K. John, Alessandro Acquisti and George Loewenstein, ‘Strangers on a Plane: Context-Dependent Willingness to Divulge Sensitive Information’ (2011) 37 Journal of Consumer Research 858; Barth and de Jong (n 171) 1047. 185 Akman (n 178). 186 Gal and Aviv (n 131) 373. 187 Akman (n 178). 188 Simon Board and Jay Lu, ‘Competitive Information Disclosure in Search Markets’ (2018) 126 Journal of Political Economy 1965; Lorenzo, Padilla and Requejo (n 176); L. Pepall and Joseph T. Reiff, ‘The “Veblen” Effect, Targeted Advertising and Consumer Welfare’ (2016) Economics Letters 145 218; Veronica Marotta, Kaifu Zhang, Yue Wu and Alessandro Acquisiti, ‘The Welfare Impact of Targeted Advertising Technologies’ (2017) Information Systems Research, available at https://ssrn.com/abstract=2951322 accessed 3 August 2021. 189 Esayas (n 132). 190 Acquisiti, Taylor and Wagman (n 168). 191 Julia Lane, Victoria Stodden, Stefan Bender and Helen Nissenbaum, ‘Privacy, Big Data, and the Public Good: Frameworks for Engagement’ (2016) 70 Cambridge University Press 119. 192 Hoppner and Westerhoff (n 52). 193 Gal and Aviv (n 131) 383; John Deighton and Leora Kornfeld ‘The Socioeconomic Impact of Internet Tracking’ (IAB, February 2020), https://www.iab.com/wp-content/uploads/2020/02/The-Socio-Economic-Impact-of-Internet-Tracking.pdf accessed 29 July 2021; Lorenzo, Padilla and Requejo (n 176). 194 Google Shopping (n 79), para 188. 195 Evans (n 2). 196 Apple, ‘The Apps You Love. From a Place You Can Trust’ (App Store, June 2021), https://www.apple.com/app-store/ accessed 25 October 2021. 197 Evans and Schmanlensee (n 25). 198 Jan Blockx, ‘Revaluing the Role of Intent Evidence in EU Antitrust Law’ (2019) 3 European Competition and Regulatory Law Review 354. 199 Case T-30/89 Hilti v. Commision, EU:T:1991:70, para 68. 200 Hoppner and Westerhoff (n 52). 201 Interim measures on iOS 14.5 (n 5), paras 73–82. © The Author(s) 2022. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Trading Off the Orchard for an Apple: the iOS 14.5 Privacy Update
JF - Journal of European Competition Law & Practice
DO - 10.1093/jeclap/lpac023
DA - 2022-05-03
UR - https://www.deepdyve.com/lp/oxford-university-press/trading-off-the-orchard-for-an-apple-the-ios-14-5-privacy-update-gMJx7u4Uf0
SP - 200
EP - 216
VL - 13
IS - 3
DP - DeepDyve
ER -