nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
I. Introduction
Technology has transformed all kinds of interactions. In the economic sphere, it has multiplied commercial transactions and offers important benefits to both producers and consumers. Producers can expand their client base, diversify products and services, optimize productive processes, reduce costs, innovate, and differentiate themselves. On the other hand, consumers have access to a broader range of goods and services at better quality and price conditions, face lower transaction and search costs, have more information to take decisions and effective channels to settle disputes with providers.
In the social dimension, technology has a multifaceted role. It promotes interaction between peers as well as the expression of ideas; it also facilitates access to information and brings citizens closer to public life. At the political level, it has become a useful tool to accompany democratic processes and offers powerful accountability mechanisms that serve as a counterweight against public authorities.
However, this process does not come without risks. The characteristics surrounding some business models of the digital economy have given rise to players with dominant positions in activities that are becoming increasingly relevant. Evidence of this is the position held by the top five technology companies in the world: Google, Apple, Meta (Facebook), Amazon and Microsoft (hereinafter referred to as the “the Big Five”). Although these companies owe most of their success to sound investments, productive efforts, and constant innovation, they have also benefited from regulatory loopholes, as well as dynamics that lead to monopolization (tipping) and encourage anticompetitive practices that weaken competition on the merits.
The notion in competition policy that “size does not matter” and that rather the abuse of dominance is the problem, is not entirely true in the digital economy, since there are a number factors that do not apply in the same way in traditional markets, such as powerful network effects, organic growth at zero cost, massive data gathering and processing, the creation of ecosystems or clusters, platform envelopment strategies, and the constant acquisition of potential competitors or incipient innovations by incumbents.
This situation surpasses the capabilities of competition authorities and shows the systems shortcomings, including an inability to effectively combat abuses of dominance. It is also critical to limit the abnormal growth rate of the Big Five who monopolize activities, abuse their dominant position, block competitors and transfer costs to related markets.
On the other hand, there are also those who defend the role of big tech companies. For some scholars there is no evidence that the behavior of digital platforms-not even the much-criticized alleged self-preference tactics of Amazon or Google-are hurting consumers and hence consider there is no reason to deviate from current policies and practices.2 Other voices point to unprecedented productivity growth, constant innovation and entry, characteristics that would not typically be present in markets with no competition; in any case, there is “a group of diversified digital companies, coexisting and competing vigorously in an oligopoly situation, with new firms entering from time to time”.3
Hovenkamp contends that sustained competition is feasible in most of the business elements of digital platforms and advises against intrusive measures such as forced divestitures, breaking-up measures or generic regulation that limits supply, quality or innovation. Even so, he recognizes an underlying problem and proposes resorting to targeted measures such as restructuring decisionmaking processes within the firms or forcing interoperability or pooling in certain circumstances to increase network effects among all players and thus favor efficiency.4
It is fair to say that companies like the Big Five have gained market share due to their productive efforts, but also through practices that have helped them to maintain and increase their dominant positions. As in other industries, a serious deficit of competition invariably decreases consumer welfare. Certain digital business models entail risks that may increase costs on certain goods and services, as well as negative externalities that, if not controlled through public action, may erode the public interest.
II. The tipping phenomenon
In some traditional sectors there are usually economic conditions that favor market concentration, as it happens in television, telecommunications, financial services, transportation, oil, or food. These sectors have in common that entry costs are high, especially due to the infrastructure, technology and logistics required. Hence, the only way to produce efficiently and at low cost is by maintaining a large scale. Sometimes there are also economies of scope, where due to synergies and an efficient use of assets, it is cheaper to produce two or more services jointly rather than separately, which in turn influences the size of the company. That is why the existence of a certain degree of market concentration in some industries is understandable and even desirable.
In the digital platform world, we are dealing with infrastructures that allow interaction between different groups of users: for example, customers, buyers, or users of services, as well as advertisers, sellers, service providers and content creators. They are complex entities that are characterized by bringing together markets from several sides (mostly two) with interdependent demands, asymmetric price structures and strong network effects, both direct and indirect, so the more users the platform has the greater its value.
The literature has emphasized the presence of indirect network effects, that is, between different groups of users of the same platform.5 Rochet and Tirole have focused on the fact that the platform sets prices in a non-neutral way (where the prices of each side matter for the total volume of transactions)6. On these platforms, the price charged to a user reflects both the cost and the externalities generated by their participation, in such a way that pricing strategies can consider categories of users as an input, given the value they create for another category.7
The network effects which are inherent to multi-sided platforms encourage a massive accumulation of users and large-scale growth, which is supported by additional factors such as:
Organic growth capacity. Unlike traditional sectors, technology allows platforms to rapidly increase their number of users at no cost.
No distribution costs. In the digital world, borders are blurred-no need to pay freight or face trade barriers-and economies of scale foster market concentration globally.
Information. Platforms can collect and process huge volumes of information, make use of intelligent algorithms and exploit personal data to offer more refined and personalized services, which reinforces business value, facilitates growth into adjacent activities and attracts more users. A market participant that has access to significant amounts of information strengthens its market position and decreases the ability of others to compete.
Consumer cognitive biases. The platforms usually take advantage of the position-conscious or unconscious-and biases of the user, to induce their decisions so that they are more in line with the platform’s commercial interests.
III. The risks
As a rule of thumb, the presence of monopolistic (or quasi monopolistic) practices in a market translates into higher prices, lower quality, reduced supply, and loss of innovation. The characteristics of digital markets encourage companies to compete “for the market” and not “in the market,” leading to a “winner takes all” effect. Under these conditions, whoever comes first tends to monopolize, facing weak rivals and few prospects of future competition. As if that were not enough, it is common practice for dominant companies to buy startups at an early stage to prevent them from becoming competitors. As the Furman report points out, large incumbent firms in the digital market act in a way which, “at best, absorbs innovation to protect themselves from potential competition and, at worst, uses acquisitions to kill off or distort innovation, creating a ‘killzone’ around their positions”.8
The Big Five operate as full digital ecosystems, with one or more main services where they hold a dominant position, but with a much wider and diversified offering. Among the various goods and services offered there is a strong interdependence, given their complementary use, productive synergies, vertical relationships or joint processing of information. This attracts users while it also generates incentives to carry out anti-competitive practices, such as bundling, “lock-in,” predatory pricing or cross-subsidization. Platforms can also act as “gatekeepers” in some part of the value chain they control and simultaneously compete downstream with their own clients, giving way to practices such as refusal to deal, self-preference or discrimination.
Although Google stands out for dominating the market of search engines, emails, and mobile operating systems (Android), it also offers multiple other tools for consumers. Apple is a leader in the production of mobile devices and hardware, but it also offers a wide variety of products and services to its users. Amazon started as an online bookstore in 1994 and now operates the most important digital marketplace in the world. Microsoft’s software products are targeted to businesses and are designed to be used together. Meta generates most of its income through personalized ads on the world’s most successful social networks (Facebook & Instagram), but it also operates the most popular app for instant messaging (WhatsApp).
The success of these companies is mostly a result of their innovation and investment efforts. However, they have also engaged in behavior that has allowed them to artificially consolidate their power and close opportunities for new entrepreneurs. The indictments against Microsoft filed by the US Department of Justice in 1998 were a prelude to a much more recent wave of concerns and issues.9
Digital platforms may for instance deploy behavioral tactics that could hardly be explained by efficiency reasons. Vertical foreclosure may take place when a dominant platform owns an input and denies access (or offers access in discriminatory terms) to a third party that competes in a related market. This may include practices such as refusal to deal, constructive refusal, denial or degradation of interoperability and excessive price of the input.10 The incentives for abuse come into play when the gains from excluding a competitor are greater than the losses incurred from not granting access (such as loss of income).
Foreclosure through platform discrimination may also occur, mostly when consumers attach significant value to using a particular platform to access a secondary product, have incomplete information about the product in the secondary market and there are significant switching costs that leave consumers locked into a single platform. Such consumers are discouraged from attempting to access secondary products through a competing platform, due to the high cost of transitioning between platforms.11
In sum, abuse of an entrenched position may take the form of vertical or even horizontal foreclosure (such as tying or bundling), various forms of discrimination and contractual arrangements. In Europe and more recently in the United States, competition agencies have launched investigations and imposed sanctions for these and other reasons, involving Apple, Google, Amazon and Meta.12 In Latin America, there are currently open investigations regarding social networks (Chile and Argentina) and app stores (Chile and Mexico).13 After reviewing high profile cases, some of which are commented below, Motta argues that broadly speaking they belong to well-known and established categories of theories of harm (mostly on vertical foreclosure), albeit adapted or modified to fit the specificities of the digital sector.
The risks associated with the Big Five depend on each business model. Facebook and Google offer “zero price” services and earn revenue through targeted digital advertising due to the information they collect from their users on their own platforms and even from external sources. Caffarra sustains that this has multiple implications, since it makes it very difficult for an entrant to invest, enter and compete, while the platforms have strong incentives to use, collect and exploit the maximum amount of information with low levels of protection and security. They can also prevent third-party businesses in the value chain from monetizing the information they generate and focus on adjacent activities that could evolve and put competitive pressure on the primary market (such as specialized search sites).14 Google’s founders themselves promoted the impartiality of their search engine on the basis that it would not be subject to the harmful influence of advertising.15
Caffarra concludes that businesses that earn income differently, be it by charging for their services, selling a complement, or obtaining a percentage of a transaction, have different incentives. This does not mean that platforms that do not monetize their users’ information are free from risks. Apple’s focal business has been the sale of devices, so it could be assumed that it has the incentive to offer through these devices the widest possible variety of services (and therefore not obstruct or hinder the use of applications developed by third parties), in such a way that it is more attractive to acquire them. However, to the extent that the income obtained from the sale of services through devices decreases the incentives could be reversed.
In 2008, when the App Store was launched, Steve Jobs himself claimed that Apple had no intention of earning revenue this way, an idea that was abandoned when the service matured and showed its high profitability. Apple and Google became critical distribution channels for multiple services through mobile apps, allowing them to charge substantial fees or commissions (reaching 30% of revenue for the first year and 15% in subsequent years). In a study, Padilla and Perkins show that when demand for electronic devices is healthy, foreclosure in the aftermarket is unlikely; in contrast, when the demand for devices faces saturation, such exclusion is more likely when a service offered by the manufacturer of the device is not too inferior to that offered by a thirdparty competitor. Under that model, they claim, consumer welfare would be increased if the device is prohibited from using its own applications that compete with third parties.16
IV. Practices under scrutiny
1. The battle between the European Commission and Google
In 2017 Google was fined by the European Commission with €2.4 billion for having abused its dominant position in general search services to favor its own comparison-shopping service (Google Search).17 The authority found that Google Shopping was not subject to the normal results thrown by the algorithm, as the search engine systematically positioned Google’s service prominently over rival options. The evidence showed that even the best-ranked price comparison sites-other than Google- appeared on average down to the fourth page of the results, which in practice meant that consumers would hardly see them.
Google alleged, among other issues, that the Bronner criteria (related to refusal to deal and essential inputs) was not satisfied and that the Commission was unfairly imposing on Google a duty to promote competition by giving its own competitors greater visibility in general search results pages. Additionally, they claimed that there was no precedent to characterize this conduct as an abuse of a dominant position, as new abuse categories had to be consistent with the legal framework and be known in advance.
In 2021, the General Court of the European Union dismissed most of the claims made by Google and upheld the legality of the fine, pointing out three circumstances that weakened competition: (i) the importance of the traffic generated by Google’s general search engine for comparison shopping services, (ii) the behavior of users, who typically focus on the results that appear at first; and (iii) the large proportion of traffic diverted in comparison shopping services and the fact that it could not be effectively replaced by other means.18
In 2018, the Commission imposed another sanction on Google, this time for €4.34 billion, for having implemented anti-competitive restrictions on Android device manufacturers and mobile network operators to strengthen its dominant position in general internet search.19 In particular, Google: (i) required manufacturers to pre-install the Google Search app and browser app (Chrome) as a condition for licensing Google’s popular app store (Play Store); (ii) paid manufacturers and mobile network operators large sums of money in exchange for the pre-installation of Google Search and (iii) prevented manufacturers who had pre-installed Google apps (including Google Search, Play Store and Chrome) from selling devices if they allowed alternative versions of Android.20
This conduct is particularly harmful considering that Google is dominant not only in search engines, but also in mobile operating systems, since most mobile smart devices in the world use Android, the original version of which was bought by Google itself in 2005. The other highly popular system, Apple’s iOS, has less reach because it is used only by iPhone’s vertically integrated model. This scheme benefits the so-called status quo bias as empirical evidence shows that users tend to stick with pre-installed apps. The practice reduced the incentives of manufacturers to invest in search and browser apps, which in turn eliminated the possibility of rivals to emerge and compete against Google.
The “third round” of sanctions came in 2019, when the Commission imposed a €1.49 billion fine against the same firm, for abusing its dominant position in the online search advertising intermediation market. Many content sites, such as news, blogs, travel, or entertainment sites, have their own online search function that generates profits on advertising. Through the “AdSense for Search” service, Google acts as an intermediary between advertisers and the owners of these websites. Starting in 2006, Google included exclusivity clauses in its services contracts to prohibit sites from placing their competitors’ ads on their search results pages. Around three years later, Google began replacing its exclusivity clauses with “Premium Placement” clauses, whereby sites had to reserve the most profitable places in search results for Google’s own ads. Other clauses required publishers to obtain Google’s written approval before making any adjustments to the way any rival ads were displayed.21 All of this prevented Google’s rivals from competing on digital advertising by preventing them from effectively placing ads on third-party pages.
The constant revision of Google’s practices in recent years has not been limited to the European Commission’s watchful eye. Recently, the Competition and Markets Authority (CMA) launched a new investigation into Google’s advertising business over fears its practices may be unfairly freezing out competitors.22
2. The dual role platform: Amazon
When platforms offer a service as intermediaries, but are also competing downstream with their own clients, there is a risk that they could use some of the information they gather for the benefit of its own retail operations. Amazon falls into this scenario, since it operates a marketplace and, simultaneously, sells its own products through its platform. The main concern is that this agent could exploit information that is not otherwise available from other sources, such as the products searched by consumers (“consideration data”) or their purchasing decisions, which could be used to match or improve offers and displace products offered by third parties, unduly benefiting from the efforts undertaken by third parties.
In 2019, the European Commission launched a formal investigation into Amazon’s use of non-public data of its marketplace. In the corresponding Statement of Objection, the European Commission preliminary found that Amazon was dominant on the e-commerce French and German markets, the largest in the European Union. It also found that Amazon’s reliance on the use of non-public business data for its own benefit regarding retail decisions-for example new products offered through Amazon Basics-distorted fair competition on its platform.23
This investigation highlights that Amazon has access to non-public business data of third-party sellers such as the number of ordered and shipped units of products, revenue, sales, performance, claims and activated guarantees. Thus, large quantities of data may be available to employees of Amazon’s retail business and flow directly into its automated system, which could be used to modify Amazon’s offers and strategic decisions to the detriment of independent marketplace sellers,24 for example, by allowing Amazon to focus on the best-selling products by category and to adjust its offers accordingly.
Although Amazon does offer competitive prices and thus benefits consumers, a repeated and widespread practice of this sort could reduce innovation efforts and displace competitors, distorting the market in the medium term. Even though a retailer is free to choose the distribution channels that best suit its interests, it may not have a different choice given Amazon’s dominance in ecommerce in multiple countries. Against these claims, it has been argued that self-preferencing would not generate any consumer harm and, moreover, it could increase price competition between sellers and private brands offered by the marketplace.25
It is important to note that every company has the right to make individual decisions based on what its competitors are doing, and even carry out “reverse engineering” strategies to introduce new products without violating industrial property rights. The problem is that Amazon could be unfairly avoiding commercial risks and capturing efforts undertaken by its competitors due to its unique dual role position.
Amazon should not have an incentive to engage in these self-preference practices as they would affect the neutrality and reputation of the marketplace, its core business. Prioritizing Amazon’s role as a retailer at the cost of pushing or excluding sellers from the platform would not make much business sense. But its dominant position may give some space to do it if the benefit is greater than the cost; sometimes organizations just make bad decisions and in doing so break the law.
In parallel to the Statement of Objections issued on this investigation, in 2020 the European Commission opened a second investigation against Amazon to assess whether the criteria that Amazon sets to select the winner of the Buy Box and to enable sellers to offer products under its Prime Programme26 lead to preferential treatment. In this investigation, the Commission preliminary concluded that Amazon abused its dominant position in the French, German and Spanish markets for the provision of online marketplace services to third-party sellers and determined that the criteria used by the company unduly favored its own retail business, as well as marketplace sellers that use Amazon’s logistics and delivery services.27
In response, Amazon offered commitments, and in December 2022, the Commission made the amended version of the commitments legally binding under EU antitrust rules.28 In general terms, Amazon committed: i) not to use non-public seller data by Amazon Retail, ii) to establish an unbiased selection of sellers for the selection of the Buy Box winner and display of two Buy Boxes, iii) to guarantee equal treatment of marketplace sellers and offers on Prime, as well as free choice of carriers and the improvement of the communication channel used between independent carriers and Amazon customers, among others.
3. Abusive Exploitation: Germany against Facebook
Facebook’s business model has been a common topic of discussion, since it is a platform that charges a zero-price on one side (social media) and monetizes on the other side (digital ads). It is therefore misleading to consider this service as “free” since consumers are paying with their personal data. The dominant provider can then exploit consumers in an open or subtle way, extracting an excessive amount of information-which necessarily implies a loss of privacy-when the price should be negative, that is, the company should be paying the user because the value of its data outweighs the value of the service itself. Are consumers being overcharged by Facebook? What value do we give to our data?29
These strategies take advantage of a peculiar behavioral bias that has been referred to as the “privacy paradox”, which consists in the fact that consumers tend to express great concern for their privacy but do not act accordingly, since they usually offer their personal data in exchange for little or even nothing.30 Many consumers are unaware of the costs associated with the loss of privacy and the income that their information represents for companies.
In March 2016 the German Federal Cartel Office (Bundeskartellamt) announced that it was investigating Facebook for the alleged abuse of its dominant position in social networks. According to the Bundeskartellamt the company collected data from its users without their consent, using both its own platforms-such as Instagram and WhatsApp-and third-party platforms, in the latter case through social plugins (“like” or “share” buttons) that allow to follow the activity of users. The competition authority established that Facebook’s data policy which allowed them to collect user and device-related data from sources outside Facebook and to merge it with data collected from Facebook, constitutes an abuse of a dominant position on the social network market in the form of exploitative business terms.31
Facebook claimed that the data collected was necessary to provide a better service and part of its legitimate business model. The company argued that data aggregation from various sources was efficient to the extent that it improved the product itself, as well as the quality of the targeted advertising. Nevertheless, the company was found guilty in 2019 for abusing its dominant position by inappropriately collecting, using, and merging its users’ data.32
The competition authority considered that the terms and conditions set by Facebook and the way in which it collected and used the data of its users violated the General Data Protection Regulation.33 In the opinion of the Bundeskartellamt, the consent requested by Facebook was illegal and ineffective, since the benefits obtained by Facebook outweighed the interests and benefits of users. In other words, in the absence of alternatives, users had no choice but to give their consent that their data be collected from sources outside of Facebookrelated activities.
Facebook appealed against this decision to the Düsseldorf Higher Regional Court (DHRC). The DHRC considered possible that at least part of the data collected could be based on a legitimate interest of Facebook. Shortly after, the German Federal Court of Justice overruled the DHRC’s decision and rejected Facebook’s request for suspension. On March 24, 2021, the DHRC decided to refer General Data Protection Regulation (GDPR) compliance questions to the European Court of Justice, which ruled that the Bundeskartellamt may take data protection rules into consideration when weighing interests in decisions under competition law.34
4. The New US Momentum
In December 2020, the Federal Trade Commission (FTC), together with prosecutors from 46 states, filed lawsuits against Facebook35 for monopolizing social networks and preserving its position through a series of anticompetitive conducts, such as the acquisition of rivals Instagram in 2012 and WhatsApp in 2014. The lawsuit also describes the imposition of anticompetitive conditions on software developers. For example, Facebook ensured that the applications that interconnect with Facebook do not compete with the platform in any of its functions-such as messaging-and do not export data or promotions to social networks different than Facebook.36 According to the lawsuit, these actions, individually and collectively, removed the ability and incentive of other apps to become competitive threats to Facebook.
On June 2021, the federal judge of the District of Columbia dismissed the lawsuit arguing that although the FTC had made out a plausible market definition for personal social network services, such authority failed to provide an estimated actual figure or range for Facebooks’ market share at any point over the past ten years.37 On January 2022, the judge admitted the FTC’s amended complaint which detailed the claim that Facebook holds monopoly power in the market of personal social networking services.38
There is still a hard road ahead to prove FTC’s allegations, especially considering US’s prevailing judicial standards and the fact that Facebook’s acquisitions of Instagram and WhatsApp were consummated more than a decade ago and were not investigated or otherwise objected at that time. And just recently, in April 2024, Meta filed a motion asking the court to award them summary judgment and dismiss the FTC’s lawsuit, arguing that they face fierce competition from a range of platforms (such as TikTok, X and YouTube) and that the FTC had failed to prove its claims of alleged harm to competition and consumers.39 In fact, they contend that the acquisition has provided benefits both for consumers and for the apps.
5. Department of Justice v. Google
In October 2020, the Department of Justice (DoJ) sued Google for unlawfully maintaining monopolies in the markets for general search services, search advertising, and general search text advertising in the United States through anticompetitive and exclusionary practices. The plaintiffs argued, among other claims, that Google licenses its apps and interfaces only if device manufacturers agree to bundle other apps and prevent their removal. These practices are also used to secure default status for its general search engine through the payment of a revenue share based on online queries; actually, Google has paid billions of dollars each year to device manufacturers, wireless carriers and browser developers.40 The lawsuit has many similarities with the findings of the European Commission in the Google Android case.
The authority claims that these practices allow the monopolization of the activities in question, since Google represents 82% of computer search queries and 94% of queries on mobile devices in the United States.41 Among other defenses, Google argued that consumer loyalty to its search engine is due to the quality of the results, that the agreements in question are no different from those used to distribute software, and that consumers can switch search engines on mobile devices at any time. Additionally, they stated that search engines are not their only competition, considering that they also compete with other sources of information.
In January 2023, the DOJ filed a second antitrust lawsuit against Google for monopolizing digital advertising technologies. The complaint alleges that Google has monopolized what is known as the “ad tech stack,” the website publishers depend on to sell ads and that advertisers rely on to buy ads and reach potential customers.42 Similar to the case filed against Facebook, the lawsuit alleges anticompetitive and exclusionary conducts in the form of neutralizing or eliminating ad tech competitors through acquisitions. The one thing that is crystal clear from these lawsuits is that the US authorities have no intention of backing down from the scrutiny that has been placed recently on these tech giants.
V. Digital mergers: efficient or risky?
The Big Five share a common feature: they have a voracious appetite to acquire any business that facilitates their growth, complements their ecosystem or strengthens their position. The Furman report states that, in the last decade, they made over 400 acquisitions globally.43 In 2017 alone, they spent $31.6 billion to acquire start-ups. From 2001 to 2018, Google bought an average of one firm per month, so the success of these companies could hardly be explained without this non-organic growth strategy.44 Although some of these transactions may have had the deliberate aim of killing innovation projects that could represent some kind of competition in the future, the reality is that most of these had the purpose of complementing goods or services already offered by the acquirer.45
The so-called killer acquisitions are more common in different markets, such as the pharmaceutical industry where incumbents have acquired new firms in order not to adopt the acquired product, but rather to stop their development.46 Using a sample of 35,000 projects from more than 6,700 pharmaceutical companies over the past 25 years, Cunningham found evidence that there is a 36.6% probability that companies will fail to develop the acquired projects when they overlap with their own portfolio, and out of those projects, 6.4% would be killer acquisitions in the full sense. In the digital sector, what firms want instead is to incorporate an innovation or functionality into their ecosystem. Between 2008 and 2018, Amazon, Facebook, and Google concluded 299 operations, 60% of which were not horizontal but aimed to acquire young companies-no older than 4 years-that offered something else.47
Many of these acquisitions generate synergies and efficiencies as the buyer develops the acquired innovation, diversifies its portfolio, and offers better services to its users. Even the possibility of selling can stimulate entry and facilitate financing the project in the first instance. However, the acquisition affects the competitive environment when (i) the acquirer is a dominant platform and (ii) the acquired company had the resources and capabilities to grow on its own merits, scale up, diversify, or become a viable competitor. In this scenario, it is very likely that the acquisition is anticompetitive, since it would reinforce leadership, eliminate potential competition and provoke a loss of innovation.
The concept of “reverse” killer acquisition can also take place. This occurs when the buyer stops innovating organically-having enormous capacities to do so-to instead buy something already developed by a third party: “the acquisition may effectively extinguish the standalone effort of the buyer to expand in a particular space because the target immediately provides it with those capabilities”.48
Among the acquisitions that have been questioned are targets that may have evolved to become competitors of the acquirer (Facebook/Instagram), others that have given the acquirer a strong position in a related market (Google/DoubleClick49) or that have involved significant amounts of data which may have consolidated the position of the buyer in both their focal business and the acquired one (Google/YouTube, Google/Fitbit and Facebook/WhatsApp). In a recent paper, Padilla and Condorelli argue how a dominant firm in a data-intensive primary market may enter a data-rich secondary market in which it could set below-cost prices and acquire customers; this, in turn, would entrench its position in such a primary market, making entry therein less desirable for potential competitors.50
In the acquisition of Facebook/Instagram, there were clear signs that this innovative company had an exciting future. While Instagram was small in assets, sales, employees, and users, charged a subscription fee, and offered few features compared to what it is now, the $1 billion purchase price was a clear sign of the value Facebook placed on the business. When a company-incipient or less mature- represents a potential competitor, the price reflects not only the value of the acquired assets but also the loss of future income of the acquired company and the protection of the acquirer’s leadership position.
According to the Lear report, the merger strengthened Facebook’s position as an advertising platform in three ways: (i) it allowed more information to be used and combined for advertising purposes, (ii) it avoided the competitive pressure that Instagram might have exerted in the future, and (iii) it increased its user base.51 This report criticizes that the authority paid too much attention to the specific functionalities offered by the apps of the merging parties, instead of analyzing with a broader perspective how they could become competitors in the business of harvesting consumer attention and selling it to advertisers.
It is of course easier to criticize the Facebook/Instagram merger now that we see the huge success of Instagram, but still this merger involved a market leader who was taking over a platform that had a strong chance of becoming an important participant. A fundamental feature of digital platforms was overlooked, that is their inherent ability to diversify and evolve. Instagram was, like Facebook, a platform that allowed social interaction between peers, attractive to consumers and mature enough to start monetizing through advertising. Instagram did develop under the umbrella of Facebook, but the relevant question was whether this company could reasonably have grown on its own or even with a different buyer.
The acquisition of WhatsApp by Facebook also raised several issues.52 At the time, the European Commission identified that, out of the services offered by Facebook, there was an overlap only in communication services, discarding effects on social networking services and digital advertising. Although the communication services of both companies offered the possibility of sending text, voice and content, the means used were different (WhatsApp via mobile app and, for Facebook Messenger, through the platform).
Hence, Facebook and WhatsApp were considered, if anything, “distant” competitors. Then the Commission did analyze the possibility that Facebook increased its power in social networks by introducing digital advertising in WhatsApp or using this service as a data source to feed Facebook services. But these risks were ruled out because the parties stated that it was not technically possible for WhatsApp and Facebook to interact between themselves, and that both systems would continue to function independently. A few years later, WhatsApp announced changes to its privacy policy, including the possibility of linking its users’ telephone numbers with Facebook accounts. The Commission fined Facebook with €110 million for providing misleading information regarding the Whatsapp takeover,53 although it assured that the original conclusion would not have been different.
It is striking to see that almost no transaction has been objected by authorities worldwide when digital mergers pose particular risks. This can be partly explained due to the notification thresholds that in many jurisdictions are based on the number of sales of the firms involved. Therefore, the acquisition of start-ups or incipient firms that do not generate much income could fall out of the radar, even if the amount of the operation is high. In response, the UK, Austrian and German competition agencies implemented changes to the criteria under which a merger is to be reviewed. But this does not explain the full story, as there are a good number of mergers of digital businesses that were either reviewed and approved, without in-depth examination, or not questioned or investigated afterwards.
In any industry, whether traditional or digital, dominance is a concern, as it may disrupt supply or price conditions to the detriment of consumers or facilitate anti-competitive conducts, whether exploitative or exclusionary. Prior merger control seeks to prevent a transaction from creating an anticompetitive situation or structure that did not exist before. Further, there are special considerations regarding concentration of information, something which has been referred to as “data-opolies”.54
On the other hand, it is likely that competition authorities have made some “type II” errors or false negatives, by authorizing concentrations which could reasonably have an anticompetitive impact of some sort. This is mainly due to three reasons:
Digital mergers do not normally involve a horizontal overlap, but the purchase of a differentiated product, so it can strengthen its ecosystem or expand into adjacent segments.
Digital markets are dynamic, and what appears to be a service at one point may evolve, add functionalities, and mutate into a different service later. Complements can become substitutes, and portfolio effects can evolve into horizontal overlaps.
Where there is an overlap between a strong acquirer and a small firm, a conclusion of potential loss of competition is difficult to reach as it must be based on robust analysis to avoid mere speculation.
Consequently, authorities must acknowledge that the digital economy sometimes requires a departure from traditional analysis. This does not necessarily mean altering the traditional objectives of protecting efficiency and consumer welfare, but rather using more appropriate analysis tools and reducing the strong risk aversion of generating false positives. In certain cases, the level of scrutiny could be raised and an anti-competitive result may be assumed prima facie, especially when the acquirer has a dominant position and the transaction raises barriers to entry through network effects or the accumulation of data, which makes it difficult for users to move -generating a lock-in effect- or it reduces the levels of competitive pressure in the present or near future.
For this strategy to be possible, procedural and substantive approaches would have to be explored. Many voices -including the Stigler and European Commission reports, as well as Motta and Peitz- propose reversing the burden of proof and presumptions, so that when a dominant digital platform acquires assets, it should be assumed that the operation is anti-competitive, unless the parties show evidence that it does not generate risks or that the efficiencies are strong enough. This proposal is not without problems since each jurisdiction would have to define how to administer these exceptions. Also, alleviating the burden of proof of competition agencies to show at least a prima facie theory of harm could clash with basic legal principles. Also, it could be difficult to prove a negative situation -that there are no risks- or even the positive in this case -the presence of efficiencies that are not well known yet-.55
A viable alternative is to change the type of analysis and lower the threshold that is typically used to object a merger. As Shapiro argues, competition policy must tolerate certain false positives to avoid false negatives and not allow transactions that eliminate competitors that could become competitors of incumbents.56 This is consistent with the proposal of the Furman report, which suggests moving from a “balance of probabilities” test to a “balance of damages” test,57 which would involve looking not only at the probability of harm, but also at the probable scale of that damage, in such a way that the intervention of the agency could take place if the scale of the damage is high, even though its probability of occurrence is low.
In this regard, Salop comments that rational decision-making under scenarios of imperfect information should give greater importance to costs rather than risks, as the courts and lawyers claim.58 This approach is useful for acquisitions involving small or start-up firms where the probability of harm is much lower since it is difficult to establish or predict their future development and success.59 It could be applied as a tool that strengthens merger analysis in a particularly risky situation, without making distinctions between sectors or companies, nor relieving the authority of its obligation to show a reasonable theory of harm. To mitigate information asymmetries that play in favor of involved parties, authorities should have sufficient power to require any relevant information regarding the transaction, including efficiencies.
Regarding the substantive analysis, flexibility is much needed. The standard should not only consider the typical concerns of a horizontal overlap -the creation of a dominant company or the empowerment of an already dominant company- a vertical implication -the denial or restriction of access to an essential input- or coordinated effects -where the new structure of the market would be more prone to collusion-. Greater attention should be given to a possible conglomerate effect that we rarely see materialize in traditional sectors, as the diversification and growth of digital ecosystems can bring together complementary or interconnected services that have or could have common users. Thus, the combination of products in related markets may give the merged entity the ability and incentive to leverage a strong position in some activity through foreclosure practices such as tying or bundling. This of course is particularly risky when one of the companies holds a dominant position, the product involved is relevant to consumers, and there are only a few or no alternatives available.
The rapid evolution of digital ecosystems also requires the injection of horizontal elements into the theories of harm. Aspects such as data accumulation and its use, growing network effects, mobility of users and the modus operandi of the ecosystem involved must be a priority in the analysis. The European Commission report suggests that the Commission must revisit substantive theories of harm under the “Significant Impact on Effective Competition Test” (SIEC test), especially for those cases in which there is a dominant platform or ecosystem which benefits from strong network effect and data access, both of which are significant barriers to entry.60
For which, the following questions should be answered: (i) does the acquirer benefit from barriers to entry linked to network effects or the use of data? (ii) is the target a potential or actual competitive constraint within the technological/ users space or ecosystem? (iii) does removing this force increase power within this space through higher barriers to entry, and (iv) if so, is the merger justified by efficiencies?
With regard to the loss of potential competition analysis, although any conclusion in this regard must be based on facts and evidence, it is also not desirable for agencies to set a high threshold for themselves that prevents them from making reasonable predictions. As the Merger Assessment Guidelines of the Competition and Markets Authority (CMA) points out, “[s]ince merger assessments are prospective, an element of judgment is necessary in deciding whether any loss of competition is substantial rather than any exact quantitative measurement”.
Unlike traditional analysis, which compares the situation before and after the transaction, in these cases one would have to ask whether the target, had it not been acquired, could have been developed by itself and become a source of significant competitive pressure. This would imply accepting that the counterfactual cannot be bulletproof, but the conclusion may be duly motivated based on the company’s profile, service attributes, innovation capacity, growth prospects and viability, among other elements.
The possible loss of dynamic competition is highly relevant. Certain mergers can eliminate efforts such as the development of new products or the improvement of existing ones, introducing more efficient or disruptive business models, or sacrificing short-run margins in order to attract users to their platform and benefit from network efficiencies. In line with this discussion, in 2021 the Competition and Markets Authority ordered Meta to sell Giphy, the largest provider of animated figures on social networks. The Competition and Markets Authority found that the deal removed Giphy as a potential challenger in the UK display advertising market, preventing UK businesses from benefiting from innovation.61 Especially considering its innovative service called “Paid Alignment”,62 which had already been offered in the US.
Another central concern was the increase of Meta’s already significant market power by denying access or offering less favorable terms to rival social networks,63 as there were practically no alternatives at the time of the acquisition; Giphy offered special features and functionalities that made it very attractive for social networks, including the quality of the content, the superiority of the search algorithm and its reach.64
In short, it is essential to consider price and non-price aspects, which can include the levels of quality and service, privacy conditions offered to the user and the way in which they can enjoy a platform without interference.
1. The Cornershop case (Mexico)
In Mexico, the Cornershop acquisition drew much attention since the Federal Competition Commission (Comisión Federal de Competencia) (COFECE) originally blocked Walmart´s attempt to acquire the platform in 2018 and then cleared Uber to do so a few months later. Cornershop was a digital platform for the purchase and delivery of products offered by supermarkets and other retail businesses. It allowed users to interact with a “shopper” in real time and view the delivery route. It also offered its commercial users a service called “Cornershop Merchant Center,” through which they had access to disaggregated information about the performance of their own stores, including the number of orders, tickets, sales by category, most popular products, shortages, and heat maps.
COFECE identified an existing vertical relationship between the parties to the transaction, as Walmart used Cornershop as a one of its distribution channels. It then detected a number of competitive risks, due to the fact that Walmart -as the would-be owner of Cornershop- could have access to strategic commercial information of its own competitors and deploy self-preferential tactics (for example, by displaying its own products in a more favorably way). The possible departure of commercial users from the platform (mainly competing supermarkets), most likely due to a loss of trust, would in any case imply the elimination of a distribution channel and fewer options for consumers. Due to these and other reasons the operation was objected.65
A few months later, a new proposal to acquire Cornershop came from Uber. Unlike the previous case, the acquirer did not have a commercial relationship with the target company since its core business was the provision of passenger transportation services and the delivery of prepared food through mobile applications. Even though there were no horizontal overlaps in the main services provided by the notifying parties,66 nor vertical relationships, the operation warranted an in-depth analysis since it would combine two very popular digital platforms, although of very different size and scope. In the digital economy certain transactions must consider the possible loss of potential competition, the effect of data aggregation and the strengthening of ecosystems. COFECE’s approach showed a flexible and practical manner to perform this analytical exercise.
Uber’s objective was to expand its portfolio of services through its application, in order to allow its users to buy products from supermarkets and other businesses already integrated into Cornershop (complementing Uber Eats). The relevant questions were: would the integration of Cornershop into Uber´s ecosystem strengthen the latter’s position in passenger transportation? Or, viewed the other way around, would the integration strengthen Cornershop’s position in such a way that it could raise prices without competitive constraints? Could Uber’s diversification generate anticompetitive conglomerate effects? The answers, to a large extent, relied on the nature of the client’s portfolios and information on the companies involved, and the synergies that could be generated by their integration.
These risks were ruled out mainly due to the following: (i) Cornershop faced significant competitive pressures not only from similar digital platforms, but also from the physical supermarkets and stores themselves, including their online sales, (ii) the groceries market is quite dynamic and contestable, with existing companies with brand recognition, technology and resources able to enter successfully, (iii) users do “multihoming” and are not tied to a single platform; (iv) from a demand perspective, the main services offered by Uber and Cornershop were not necessarily complementary, (v) likewise, the information provided by Uber users was not complementary to that provided by Cornershop’s users; (vi) Cornershop’s user base was very small compared to Uber’s; (vii) bundling Uber Rides and Uber Eats with Cornershop under one membership could ultimately benefit users; (viii) after the transaction it could be expected that Cornershop (directly or through Uber Eats) would offer its services in new cities, as eventually did happen.
Finally, although there was evidence that Uber had plans and made certain efforts to participate in the delivery of groceries market, it was hard to confirm whether this would happen and when.
VI. Competition Policy and Regulation: Challenges Ahead
Some digital models pose significant challenges, not only because they tend to monopolize markets, but also because of their inherent dynamism and complexity, as well as the incentives that incumbents have to strengthen their position through artificial strategies or anti-competitive practices driven by technology. It is key to distinguish what falls within the scope of competition policy and what is a matter for regulatory intervention. The conditions of entry, the degree of contestability and competitive rivalry depend on multiple variables ranging from economic to institutional. Therefore, action on multiple fronts is required.
The digital environment needs regulation and public policy like any other activity that causes negative externalities, affects consumers, or generates social risks. Competition law is not supposed to set the rules of the game for specific industries, but rather to correct deviations that affect the efficient functioning of markets. Therefore, it is necessary to distinguish between strengthening competition systems and necessary regulatory approaches. Both could complement each other towards similar or complementary ends.
There is no need to redesign the fundamental principles and concepts of competition law. Most of the procedures and tools still have a solid ground and may work quite well even in the digital context. The system in most jurisdictions is flexible enough and apt to resort to different methodologies, analytical approaches and risk scenarios depending on the circumstances.
The consumer welfare standard normally requires determining an overprice created by anti-competitive practices, while in the digital economy it is common to observe “zero prices”. Hence, different approaches are needed such as considering counterfactual “negative price” scenarios or analyzing variables other than price, e.g. quality, availability, variety, access or loss of innovation. In the digital economy, aspects such as displacement through the increase in switching costs, lock-in effects, misuse of information or the loss of potential competition become critical.
Another discussion has to do with market definition. In two-sided or multiple-sided markets, which is common to see in the digital economy, the question arises as to whether they should be defined jointly or separately. This only implies one more step in the analysis, not an obstacle. Thus it becomes necessary to analyze the dynamics of interdependence between each side, so they may belong to the same market when transactions occur between each other (a marketplace or transport services through mobile apps), or one side may be a market in and of itself when the network effects only operate within that group (as in social networks, where one component would be the user side and the other the advertiser side). Multi-sided markets are not exclusive to the digital world, as they have always existed to the extent that two categories of users with interdependent demands interact with each other; a shopping center attracts both tenants and buyers while a newspaper attracts both readers and advertisers.
This certainly comes with complexities. The definitions become more complex when the platform operates as an ecosystem or a conglomerate, with multiple functionalities and a great capacity for versatility and development. Likewise, the price of each side should not be analyzed independently since the increase in price of one side can derive from the increase in the cost or in services offered by the other side. The interdependence of demands could require more information and complex analytical methods or even measure values such as changes in quality.
On the other hand, market shares may or may not be indicative of dominance, depending on the duration of innovation cycles, entry conditions and the evolutionary nature of services that offer one thing one day and another the next. In this context, it is essential to consider the impact of data and the mobility of users. In the digital world it would make more sense to determine the relevant market through observed dominance and not the other way around. Basing dominance conclusions on a previous market definition under a rigid approach may distort the results, as it could ignore the competitive pressure that two seemingly different activities exert between each other. Unlike traditional sectors, competitive pressure in the digital world can come not only from the same products but also from differentiated versions. In addition, services are often offered in a complementary manner through dynamic and evolving ecosystems.
Finally, certain digital concentrations must be analyzed with a greater degree of scrutiny and the scale of the possible damage should be considered rather than the likelihood of it materializing, especially when a digital platform with market power acquires a service that has the possibility of becoming a competitor, or when the merger raises the barriers to entry through network effects, restricts the mobility of users or triggers data accumulation.
Therefore, it is not necessary to modify the competition regime, but to make its application more flexible. There are of course concrete adjustments that would strengthen the system, for example modifying the thresholds or criteria to notify a merger, expanding the catalogue of anti-competitive practices, increasing fines or empowering competition authorities when it comes to their investigation tools and resources.
Regarding regulation, the outlook is different, as some digital giants have flourished in the absence of government controls. The public reaction has been notoriously slow. The phenomenon is new, and even to this day it is not clear how we should deal with the risks and challenges. There is also a natural tension between restrictions that may serve legitimate purposes but they could be counterproductive and lessen innovation and entrepreneurship.
In any case, social objectives related to privacy, security, and the flow of information in general must be protected. The same is true when minimum conditions of competition and rivalry are needed. At the same time, it is essential that regulation does not inhibit innovation efforts and investment, as this could bring greater costs than benefits. Unlike competition procedures, where it is valid to expect false positives, regulation is applicable across the board and its consequences have a great impact.
It is also preferable to only resort to measures such as structural remedies and divestments in exceptional situations, since the evidence indicates that this type of intrusive intervention is usually counterproductive and does not produce benefits for consumers.67 This is so because they increase costs and reduce economies of scale and scope, while well-designed regulatory schemes could allow economic agents to be larger and reach more users.
In a structure with several participants of similar size, for example, it would be convenient for all of them to operate with a total user base instead of having each their own, or to contribute information to a common database that allows everyone to operate more efficiently with less risk, as it happens in the financial sector. The problem in the digital economy is not dissimilar to what has been seen in the telecommunications sector, where the largest operator has incentives to deny interconnection to its user base to block rivals.68
The regulatory measures that have the potential to address different challenges depending on the underlying problem and the type of business model include data portability, access to data that is in the hands of a dominant participant, pooling, interoperability, and marketplace regulation. After an intense public debate, in March 2022 the European Parliament and the Council- representing the 27 member countries of the European Union-agreed on the new regulation named the Digital Markets Act (DMA).69 The objective of the regulation is “complementary to, but different from that of protecting undistorted competition, which is to ensure that markets where gatekeepers are present are and remain contestable and fair, independently from the actual, potential or presumed effects of the conduct of a given gatekeeper covered by this Regulation on competition on a given market”.70 In other words, it emphasizes that it is a regulation that seeks to influence ex ante the behavior of the regulated subjects, without having to investigate or determine the existence of an anticompetitive practice.
This is the first regulation specifically aimed at the large digital platforms that offer services to users and that are the gateway for commercial users. The DMA is only applicable to companies designated as “gatekeepers,”71 which are those that play a particularly important role in the internal market because of their size and their position as gateways for business users to reach their customers.72
The DMA entered into force in November 2022; in September 2024 the European Commission designated for the first time six gatekeepers-Alphabet, Amazon, Apple, ByteDance, Meta and Microsoft.73 The DMA has been subject to debate, and it will take years for its usefulness to be seen, but it offers a good approximation of the way in which specific digital models could be regulated. In Latin America, although there have been important case law and current investigations, up to date, there are no specific regulations in force for competition in digital markets such as the ones enshrined in the European Union.74
VII. Conclusions
Certain digital platforms have created ecosystems of excessive dominance, reduced the options available to consumers, deployed aggressive practices that reduce competition on merit and limited a healthy competitive rivalry, thus justifying public action on various fronts.
Regulation is pertinent when there is market failure or a legitimate public objective is pursued, which is visible in the digital environment regarding consumer rights and security, protection of personal data and contestability conditions. In economic activities with constant innovation, interventions must be rational, focused, proportional and should not limit legitimate business decisions, since there is a risk of reducing innovation and lessening social welfare.
Radical measures such as the structural separation of businesses or the divestment of assets, should be the exception, considering that their effects are not predictable. Instead, measures that encourage entry and create scale for smaller businesses should be preferred, such as interoperability, user mobility, multihoming, information portability and pooling, even taking some regulatory experiences from traditional sectors such as finance or telecommunications. Specific prohibitions or mandates, such as those contained in the DMA, can also come into play. Since the main digital platforms operate globally, some solutions will have to be transnational in nature, which complicates (but does not prevent) the adoption of solutions. Therefore, the reaction of the main jurisdictions and the work of international organizations are essential.
The competition system can continue to function around its fundamental premises, without losing sight of the fact that it was never designed or intended to formulate specific rules for industries. It has the ability to face the challenges posed by the digital economy, particularly to monitor and forcefully penalize abuses of dominance, as well as to prevent anti-competitive mergers.
But this implies strengthening the material and technical capabilities of the agencies, revisit analytical approaches, favor flexibility, reduce risk aversion and understand that digital platforms behave differently from traditional sectors. Furthermore, the consumer welfare standard must be viewed broadly, so that not only short-term effects on prices are considered, but also dynamic aspects such as the loss of innovation and the impact on privacy, user mobility and the characteristics of the use of the services involved.
There are multiple cases around the world, which have shed light on the type of anticompetitive behavior where it is worth paying more attention and devoting resources. The response to these challenges should not punish size or success. Above all, it is essential to maintain economic and institutional conditions that stimulate entrepreneurship and constant innovation, since historically this is what has eliminated dominances that were believed to be indestructible.
Creative destruction and legitimate state intervention should not be seen as mutually exclusive. The digital economy must involve complementary actions from different authorities, at least in a triple dimension: (i) an economic aim to “protect” the process of free competition and rivalry, mainly repressing behavioral deviations and preventing mergers that lessen competition; (ii) economic regulation that seeks minimum conditions of fairness and contestability, with ex ante interventions of a preventive nature; and (iii) social regulation that protects the privacy conditions and the basic interests of consumers, in an environment where personal information is increasingly valuable to companies. Digital markets are not especial and should be subject to justified controls as many others. It is the obligation of governments to achieve objectives based on public interest grounds without depriving society of the enormous benefits that innovation and technology have brought to society.
All this discussion should also remind us that technology will keep posing new public challenges and we cannot stay at a standstill. Artificial intelligence has made its way into people’s life in manners nobody expected. AI tools are now capable of generating content, ideas and influencing human behavior. Disclosure requirements, pre-check systems and basic prohibitions must be put in place. Regulatory loopholes cannot prompt other digital models to flourish again in detriment of the public interest.
