This note discusses recent developments at the Brazilian Antitrust Authority – CADE – on handling digital markets and platform cases. It provides a brief view of official reports and jurisprudence, also bringing the perspective for the coming years. Despite the absence of institutional changes directed to digital markets, CADE undertook extensive efforts to properly assess such mergers and abuse of dominance cases. The results are visible and important given Brazil’s position as a developing economy. Considering the BRICS countries, CADE is an early adopter of the concepts and evolution is visible. On the other hand, abuse of dominance case analysis is still in earlier steps and divided in its legal standard. When discussing digital platform antitrust there are many global challenges with no simple answers. CADE is up to the task, with transparency and engagement with the antitrust community.
By Eduardo Pontual Ribeiro, Svetlana Golovanova, Camila Pires-Alves & Marcos Puccioni de Oliveira Lyra1
I. INTRODUCTION
The investigation of business practices of digital platforms in merger and antitrust investigations requires the adoption of new economic analysis tools. They should consider multi-sidedness, cross-platform network effects, the role of relative prices, information-based technologies effects, dynamic efficiency, and other considerations. These platform features have been discussed in economic literature2 while the application of these theories in antitrust investigations has picked up only in the recent years. The multi-sided platform (“MSP”) concept is actively discussed in mature jurisdictions. Detailed reports on platforms and digital markets have been presented in the U.S., Europe, Germany, and other OECD countries.3 Younger jurisdictions have also started incorporating two-sided logic in their analysis of merger and abuse of dominance cases. In particular, the report of the BRICS4 Working Group on the Digital Economy5 presented at the fourth BRICS Competition Conference in Moscow, 2019, marked the authorities’ view on the challenges competition agencies face due to the digitalization of the world economy.
This note discusses recent developments at the Brazilian Antitrust Authority – CADE – on handling digital markets and platform cases. It provides a brief overview of the Authority’s opinions based on available documents and jurisprudence, and offers a perspective on specific topics of interest as they are likely to develop in coming years.
II. CADE’S INSTITUTIONAL FRAMEWORK AND EFFORTS ON DIGITAL MARKET ANALYSIS
The current Brazilian competition law – Law 12,529/2011, which came into force in May 2012 – is the main legal framework for Competition Policy in Brazil. The competition law and policy it sets out is considered to be modern and compatible with international standards.6 Resolutions and best practice guidelines, such as the recent “Guide for Horizontal Merger Review,”7 support law enforcement and increase the transparency of the analysis. However, so far, no guidelines have been issued or modified to meet the peculiarities of digital platforms.
Law 12,529/2011 created the Department of Economic Studies (CADE’s Chief Economist Office), which is responsible for providing economic advice to both the investigative authority (“General Superintendence”) and the decision-making body (“Tribunal”). The DEE has had a special role in the debate and knowledge capacity building at CADE, contributing to improve the analysis of new and challenging topics, such as digital markets.
Institutional and legal issues have been considered in the recent debate about digital markets and platforms across the globe. Particularly, in merger review, notification thresholds based solely on turnover or market shares have been called into question. This is due to the fear that relevant acquisitions are not being scrutinized by authorities even in major jurisdictions, because firms may have low or absent turnover. Austria and Germany made amendments to their competition laws to include transaction-value thresholds, and are leading this discussion worldwide. In Brazil, relevant cases such as Facebook/Whatsapp and Google/Waze were not notified to CADE due to low revenues from one of the parties.8
There is no clear or acknowledged intention of amendments in the Brazilian notification thresholds which would require changes on the law. As stated in the BRICS report and still valid as of 2021: “In Brazil, no particular formal changes in the legislation are under consideration to specifically address the digital economy. The same applies for changes in notification thresholds […].” This may be in part a consequence of the possibility given by the Law 12,529/2011 to CADE to request the submission of any non-notifiable merger in the country within one year of its consummation. This possibility is rarely used by the authority.9 However, CADE has suggested to be aware of the issue and may be monitoring acquisitions made by digital platforms in recent years.10
Although there have not been recent institutional changes directed at digital markets and platforms issues, CADE has been actively debating digital markets in the last few years, making efforts to better address the challenges presented by the digital era. Considering that the assessment of competition is by nature a global challenge (given the worldwide reach of some digital platforms), these efforts included participation by international actors.
In July/August 2019, CADE hosted the international event Designing Antitrust for the Digital Era, inviting local and international guests from competition authorities, academics, and private players. September 2019 saw the publication of the report BRICS in the Digital Economy – Competition Policy in Practice (2019). This put forth the views of BRICS countries competition authorities on a number of challenges is presented, including CADE’s.11
First, when it comes to relevant market definition and market power assessment, CADE uses traditional tools to assess market power in digital markets, but there is an exception for multi-sided markets. Second, regarding innovation and dynamic aspects, the report briefly mentions that CADE considers such issues on a case-by-case basis, recognizing the difficulties of assessing the long run effects of competition policy interventions, but stating that considering post-merger incentives to innovate need to be considered. Third, acquisitions of nascent and innovative firms (often referred as startups) frequently do not trigger CADE’s thresholds for merger submission, although, as mentioned, the Brazilian Competition Law allows CADE to review any merger. However, CADE notes the difficult balancing act that this type of acquisition raises. According to the report, while startup acquisition by large incumbents risk eliminating potential competition, a restrictive approach may discourage innovation. Furthermore, one needs to consider that these acquisitions could also lead to know-how and technology transfer to the acquired firm. Fourth, the Brazilian authority mentions that it monitors attempts to prevent entry by dominant platforms, specifically pointing to firms that leverage their user base and data concentration.
In August 2020, a working paper published by its Economic Studies Department12 reviewed twenty-one documents produced by competitions authorities and specialized centers on some topics about competition in digital markets, such as the key features of digital markets, relevant market definition, assessment of price and non-price effects, specificities of different business models and possible solutions for the identified challenges. The report is an effort by CADE to study how international authorities and specialists are facing digital market challenges in order to improve its assessment of such cases. The emphasis in the paper is on online advertising markets. Many businesses in digital markets and platforms depend on advertising revenues for service provision, and there are indications that CADE will pay particular attention to the effects of increasing market concentration and the role of digital platforms on the advertising side of platforms in the coming years.
CADE’s latest working paper on digital markets was published in August 2021,13 and focuses on features related to digital platforms, along with an extensive overview of its case law. To begin with, the report lists the characteristics of digital markets: (i) positive direct network effects, (ii) positive indirect network effects, (iii) cross subsidies, (iv) scale without mass,14 (v) low marginal costs, (vi) possibility of attracting consumers from all over the world, (vii) scale and scope economies, (viii) data generation and use, (ix) disruptive innovation, (x) switching costs, (xi) winner-take-all or winner-take-most markets. It is important to mention that not all these features are present in all digital platforms.
CADE lists three groups of digital platforms: subscription-based, ad-based, and open access, adapting the usual classification of transaction, non-transaction, and matching platforms in the academic literature.15 In the first, the user pays a flat fee to access the service (e.g. video and music streaming services). In ad-based platforms, the user usually does not pay a monetary fee, and the platform generates its revenue through advertising and selling data which improves advertising effectiveness (e.g. non-subscription online news portals). Finally, open access platforms connect suppliers and users. They may charge the user and/or the supplier for selling or buying goods and services (e.g. app stores).
Most of the report reviews the case law. Features such as how CADE defines relevant markets, barriers to entry, rivalry and efficiency were discussed in several cases. Discussing the specificities of all these sectors is beyond the scope of this paper, but we can discuss a few interesting points from the report. CADE identified 143 merger cases related to digital platforms between 1995 and 2020 (around 35 percent related to online advertising and 20 percent to e-commerce). Most of them (around 86 percent) were assessed under the simplified procedure, with 14 percent under the ordinary, regular, procedure.16 Almost all of them (140) were approved without restrictions, as only two were approved subject to remedies and one was abandoned by the parties. Cases handled by CADE included the following platform types: (i) online music services, (ii) e-commerce, (iii) online tourism, (iv) food delivery platforms, (v) digital mapping platforms, (vi) social networks, (vii) on-demand video services, (viii) ride hailing apps, (ix) online ticket sales, (x) online financial investments, (xi) online classified ads, (xii) platforms for searching, ads, and price comparison, (xiii) physical exercises apps, (xiv) apps for intermediating domestic services.
The report also finds that 16 alleged anticompetitive practices in digital markets were assessed by CADE between 1995-2020: six of them are still under investigation, nine were closed with no penalties and one was subject to conditions. Most of the cases relate to alleged abuses of dominant positions and exclusive dealing. The most relevant sectors were search, price comparison, and online advertising (37.5 percent) and ride hailing apps (31.25 percent). The latter represents half of the cases that have been subject to a final judgement (five out of ten).
III. RELEVANT CASES AND EVOLUTION
From the many cases listed at the recent survey by CADE’s DEE and the review by Golovanova and Pontual Ribeiro (2021) mentioned above, we select a few merger or abuse of dominance cases that illustrate CADE’s analysis of platform and/or digital cases.
From a multi-country perspective, BRICS countries comparison by Golovanova, Pontual Ribeiro (2021)17 shows that CADE is an early user of platform antitrust concepts. It is also a leader between the jurisdictions in terms of the number of cases related to platform businesses and variety of markets investigated.
CADE demonstrates familiarity with the MSP concept starting from at least 2004. The first explicit mentioning of multi-sidedness of the markets and cross-side network effects were on credit/debit cards (Visa Vale, 2004; Visanet 2009-2010). More than a decade ago CADE correctly identified sides of the platforms (issuers and acquires; cardholders and merchants) and characterized the role of relative prices.
Case decision documents on abuse of dominance and merger cases also reveal the evolution of platform business practices analysis. For example, in the Google/DoubleClick merger (2008), CADE did not pay attention to multi-sidedness in defining the market as “ad serving services on the Internet.” One year later, considering the merger of Microsoft and Yahoo, CADE stressed the need to match audience and advertisers to ensure a successful business. In 2013, in the Google/Vevo case, CADE recognized two interrelated relevant markets, as would be expected for non-transaction audience-providing platforms (an advertising market and a content market).
The Itaú/XP merger is another example of platform antitrust. The second largest bank, Itaú, acquired a portion of XP, an investment fund platform. The market for investment fund distribution through independent brokers was seen as a two-sided. The merger was cleared with remedies that explicitly addressed the two sides of the market: on one side, XP is not allowed to impose exclusive agreements in distributing the investment products of a specific bank or fund; on the other side, XP may not impose exclusivity agreements with the independent brokers. Interestingly the Central Bank of Brazil, which has also oversight on financial markets mergers, issued a decision with more strict restrictions on the merger, voiding the planned majority acquisition by Itaú.18
The growth of the platform delivery business model forced CADE to look closely to this market. The Naspers/Rocket (or their better-known brands iFood/PedidosJá) decision from March 2018, has extensive use of platform concepts and recognition of its economic forces. The case documents refer often to the Just Eat/Hungyhouse decision in the UK by the CMA. The case documents discussed relevant markets in detail, concluding on a platform-orders only market. It concluded that concentration levels were low to generate an antitrust harm presumption. More interestingly, econometric evidence used at the case suggested that PedidosJá was not a close competitor to iFood. Nevertheless, the case decision document noted the possible anticompetitive effects of exclusive dealing with restaurants by a dominant online order platform. In an abuse of dominance case brought to CADE in late 2020, the Authority did place an injunction against iFood, blocking exclusive dealings with restaurants, given their continuing expansion and possible anticompetitive harm19 in March 2021. The case has not been decided as of September 2021.
The above cases sample CADE’s development on handling digital platform mergers. Its experience with abuse of dominance cases is more limited. Three types can be highlighted. First, cases on payment methods (credit and debit cards), where the Brazilian Antitrust Investigative body at the time (“SDE”), jointly with the Central Bank issued a detailed report on the industry and denouncing the vertical integration between scheme owners and acquirers. Since 2017, several cases or abuse of dominance in the acquiring industry have been carried out against Visa and Rede (formerly the acquirer associated with MasterCard), the two dominant acquirers (controlled in part or wholly by issuing banks). The earlier cases echoed the international investigations in markets that led to the regulation of interchange fees in Australia, United States, and some countries of Europe.20
Second, online travel agencies, namely, Booking, Expedia and a regional competitor Decolar most favored nation price agreements with hotels have also been investigated by CADE. The case was settled with the decision to allow narrow price parity clauses but banning wide price parity clauses. The decision referred ostensibly to similar cases in Europe.
Last, but not least, Google has been accused of abuse its dominance in the online sponsored search market. The cases were decided in 2019-2020 but refer to practices from the first part of the 2010 decade. One of them was under scrutiny in other jurisdictions regarding the algorithm implemented by Google on its sponsored search. While the decision in Europe led to the application of a multibillion-euro fee on Google, in the case of Brazil the Tribunal was markedly divided with equal votes to convict of an abuse of dominance and to close the case. It was a heavily debated case with written opinions of all six voting members of the Tribunal at the time of the ruling.21
The casting vote by CADE´s president (that had already voted to close the case in the individual vote) let to the closing of the case without conviction. The debate rested on whether there was actual damage or not on price comparison websites once Google implemented its policy not to allow these price comparison websites to advertise on product listing ads (“PLAs”). Part of the Commissioners believed that the fast growth of Google Shopping was caused by the business practice at the expense of comparison websites, while the other half of the Tribunal believed that such harm, if present, would not have exceeded the possible benefits of an alleged improved more effective algorithm by Google.
Interestingly, the case illustrates CADE’s handling of abuse of dominance cases trying to move beyond per se analysis. In the general perspective of rule of reason decision making at CADE, it is often the case that the analysis includes structural conditions to gauge the possibility of generating anticompetitive hard. Many cases are explicit in the possible efficiencies of the business practice, but few, if ever, effectively provide a balancing of the actual harm (at least in the quantitative sense) and procompetitive effects of the conduct under evaluation.
IV. CONCLUDING COMMENTS, GOALS AND CHALLENGES AHEAD
When confronted with the challenges of the digital era, Brazil’s antitrust authority undertook intensive efforts to properly assess mergers and abuse of dominance cases, considering international developments and case specifics. This effort can be seen in the events it held and participated, the documents it produced and, most importantly, in the case law. As mentioned, a considerable number of mergers in digital sectors were assessed, while some anticompetitive practices were investigated. Considering the BRICS countries, CADE is an early adopter of the concepts and tools used for these cases.
On one hand, the evolution of the assessment is clear. From the first use of the multisided-platform concept in 2004 to the recent Google and iFood cases, we can see a refinement in the use of economic tools. CADE’s efforts in merger control are extensive, covering many sectors in digital services. On the other hand, CADE has still some ground to cover. The assessment of abuse of dominance is still timid and divided in its legal standard even in other industries.
Furthermore, the decision of assessing on a case-by-case basis not changing its merger notification thresholds as a response to the growing challenge of nascent and innovative firms’ acquisition may be considered a conservative stand, especially given that other jurisdictions are applying more aggressive solutions. However, the legal possibility for CADE to require the submission of non-notifiable transactions may partially compensates this conservative position, when compared to other countries where there is no such prerogative. The OECD peer review released in 2019, taking into consideration the challenges posed by the digital economy, advised that Brazil should “[i]ntroduce a new notification threshold based on the value of the assets involved in the transaction” following other countries initiatives. It also broadly recommends that: “Brazil should regularly review its merger notification thresholds […] to consider a reform of the merger thresholds to reduce the number of non-problematic filings;” and “[e]xtend the deadline that CADE has to open an investigation against non-notifiable transactions from 12 to 24 months.”
Some of the challenges to competition policy practice on digital platforms are global. CADE’s DEE has shortlisted some of them in the Digital Platform Study cited above, namely: How to intervene in markets so dynamic? How to estimate long-term effects of competition policy? How to adapt measures to platform specificities? How to deal with and identify exclusionary practices by a dominant firm due do data concentration? How to deal with discrimination based on user data and profiling technology? How to deal with algorithmic collusion and vertical relations in retail? How to coordinate actions between privacy and competition policy?22
CADE is catching up. The results of its efforts are visible and especially important given Brazil’s position as a developing economy and the lack of definitive answers by other national antitrust authorities on a number of issues. There are many challenges for which the answers are far from simple but facing them is on the agenda and CADE is up to the task, with transparency and engagement with the antitrust community.
1 Respectively: Professor at the Institute of Economics – Federal University of Rio de Janeiro (UFRJ), researcher at GDEC (Research Group on Law, Economics and Competition) eribeiro@ie.ufrj.br. Professor, National Research University Higher School of Economics. Nizhny Novgorod, Russia, svgolovanova@hse.ru. Professor at Institute of Economics – Federal University of Rio de Janeiro (UFRJ), coordinator and researcher at Gdec (Research Group on Law, Economics and Competition), camila.alves@ie.ufrj.br. PhD student in Economics – Federal University of Rio de Janeiro (UFRJ), Teaching Faculty – Fluminense Federal University (UFF) and Researcher at Gdec (Research Group on Law, Economics and Competition), marcosplyra@gmail.com.
2 Rochet J.C., Tirole, J. 2003. Platform Competition in Two-Sided Markets, Journal of the European Economic Association 1(4), 990-1029 is often cited as a canonical paper, but the modelling can be traced to Blair, R. and Romano, R. 1993. Pricing decisions of the newspaper monopolist. Southern Economic Journal 59 (4) 721-732.
3 A collection of reports is available at https://research.chicagobooth.edu/stigler/events/single-events/antitrust-competition-conference/world-reports-digital-markets. A survey on the reports is found in Lancieri, F., Sakowski, P. Competition in Digital Markets: A Review of Expert Reports, Stanford Journal of Law, Business & Finance 26, 65-170 http://dx.doi.org/10.2139/ssrn.3681322.
4 BRICS is an acronym which refers to Brazil, Russia, India, China, and South Africa.
5 The Group was set up in the 2017 BRICS Competition conference in Brasilia.
6 E.g. the 2019 OECD peer review https://www.oecd.org/daf/competition/oecd-peer-reviews-of-competition-law-and-policy-brazil-ENG-web.pdf.
7 https://cdn.cade.gov.br/Portal/centrais-de-conteudo/publicacoes/guias-do-cade/Guide-for-Horizontal-Merger-Review.pdf.
8 According to Brazilian law, a merger must be submitted to CADE when one party revenues exceed R$ 750 million (USD 132.84 million) and the other party turnover exceeds R$ 75 million (USD 13,3 million). Law 12.529/2011 article 88 and “Portaria Interministerial 994”, 05/30/2012. See PIRES-ALVES, C. C.; GONZALO, M.; LYRA, M. P. O. (2019). Startups and young innovative firms mergers & acquisitions: an antitrust debate? Lessons from the ICT. Revista de Economia Contemporânea V.23, n.2, may/aug, 2019.
9 Law 12,529/11, Article 88, §7.
10 See the recent Digital Markets Report, discussed below, and Market Monitoring n° 08700.002785/2020-21 “intended to monitor the history of acquisitions […] in the last 10 years” carried out by several players in digital markets.
11 A summary may be seen in Sakowsky and Park (2019). https://www.competitionpolicyinternational.com/cade-and-the-challenges-of-the-digital-economy/.
12 A version of the report has been published as Lancieri, F., Sakowski, P. Competition in Digital Markets: A Review of Expert Reports, Stanford Journal of Law, Business & Finance 26, 65-170.
13 https://cdn.cade.gov.br/Portal/centrais-de-conteudo/publicacoes/estudos-economicos/cadernos-do-cade/plataformas-digitais.pdf
14 Refers to the situations in which platforms grow faster and cheaper than physical goods markets as they present scale economies without a tangible good, being referred as scale without mass. The concept is taken from BRYNJOLFSSON, E. et al. (2008), “Scale without mass: Business process replication and industry dynamics”, Harvard Business School Technology & Operations Mgt. Unit Research Paper No. 07-016, Retrieved from http://ebusiness.mit.edu/research/papers/2008.09_.
15 For example, Golovanova, S., Pontual Ribeiro, E. 2020. A Unified Presentation of Competition Analysis in Two-Sided Markets. Journal of Economic Surveys. doi: 10.1111/joes.12362
16 Simplified procedures refer to cases where small concentration in relevant markets does not generate the probability of anticompetitive effects. Ordinary cases refer to those cases where concentration levels (or a sequence of mergers in the same market) may raise the possibility of anticompetitive effects and require detailed analysis of the competitive forces affected by the merger.
17 Ref. footnote 16.
18 Up to 2018 CADE and the Central Bank argued in courts on the sole mandate to analyze mergers, possible creating a de fact sector exclusion from Law 12,529/2011 to the financial sector. This was solved by a joint agreement between the entities. The agreement indicates that CADE and BCB have simultaneous mandate over financial market mergers, with a unilateral mandatory decision by BCB if it understand there is a systemic risk to the financial sector without the merger; at the same time BCB recognizes CADE´s sole role in judging financial sector abuse of dominance cases. A law has been passed in the Senate PLS 350/15 and is pending a decision in the lower house of Congress.
19 Case number 08700.004588/2020-47.
20 The Central Bank of Brazil started regulating the interchange fee from 2019, in part influenced by the number of abuse of dominance cases handled by CADE.
21 In general, there are at most two detailed written decisions, by the designated Commissioner and the leading dissent vote, if present.
22 CADE’s chief economist Guilherme Resende discusses algorithmic collusion on a CPI Column. Check https://www.competitionpolicyinternational.com/algorithmic-collusion-competition-implications-and-anticompetitive-evidence-in-brazil/.
