By Chenying Zhang
Internet applications are becoming increasingly widespread, Internet-based trading platforms and associated big data have gradually developed. There is worldwide concern about data security. This is increasingly becoming an issue in competition enforcement against big-tech companies. As a country with a large number of Internet users and a huge Internet economy market, China is also facing these problems. This article discusses the principles applicable to these concerns under China’s anti-monopoly law, and analyzes possible amendments to the law, which are under discussion.
By Chenying Zhang1
Internet technology has become increasingly widespread. Over more than 20 years, Internet-based platforms and their associated “big data” have gradually grown in economic importance. The use of big data, moreover, has enhanced cross-border Internet trading, which has aroused worldwide concern about data security. Apart from data protection concerns, an important issue is competition enforcement with respect to bigtech companies. As a country with an enormous amount of Internet users, and thus a huge Internet economy, China is also facing problems related to the supervision of digital markets. The State Administration for Market Regulation (“SAMR”), the top market regulator in China, has made a number of enactments, including its Interim Provisions on Prohibiting Abuse of Market Dominance (the “Interim Provisions”) on June 26, 2019.
The Interim Provisions set out nine considerations for determining market dominance on the Internet.2 On August 8, 2019, the Guiding Opinions of the General Office of the State Council on Promoting the Standardized and Healthy Development of the Platform Economy were released, designating the SAMR as the responsible agency for “formulating and promulgating relevant regulations on the supervision and administration of online transactions, investigating and imposing punishment for illegal acts such as abuse of market dominance to restrict transactions and unfair competition in the Internet field, prohibiting the platform from unilaterally signing exclusive service delivery contracts, to ensure fair participation of relevant market players of the platform economy in market competition.”
On January 2, 2020, the SAMR released a revised draft of the Anti-Monopoly Law (“AML”), adding considerations for determining market dominance in the new Internet industry,3 and it was different from the Interim Provisions noted above.
The concerns are well known. As far back as 2013, Qihoo 360 sued Tencent on the basis that the latter abused its market dominance, and recently, Tmall was sued by JD.com and Galanz, respectively, for the same reason. Further, competition in the so-called “digital economy” is not limited to alleged abuses of market dominance, but also raises challenges with respect to other rules, such as those on anticompetitive agreements and market concentration.
But the characteristics of the “digital economy” undermine the credibility of the traditional anti-monopoly tests. Is anti-monopoly enforcement needed in the field of big data? There are different opinions on different grounds. But what is certain is that the market competition unleashed through the catalytic effect of Internet platforms and big data has had a huge impact. The effectiveness of the current enforcement system in this field needs to be examined.
II. IMPLICATIONS OF THE “DIGITAL ECONOMY” FOR REGULATION
A. Basic Attributes of the “Digital Economy”
Due to the “digital revolution,” technologies such as “big data,” the Internet of Things (“IoT”), blockchain, and artificial intelligence (“AI”) have become more and more widespread. In addition, so-called “big-tech” companies have come into being in many sectors. In 2019, seven of the world’s ten largest companies were online “platform” operators,4 namely Microsoft, Apple, Amazon, Alphabet, Facebook, Alibaba, and Tencent.
The “new economy” is based on new business models based on innovative information technology. It has at least the following five distinct characteristics compared with the “traditional” economy:
- First, the main actors in the “digital economy” are platforms. Platforms, as a new means of cooperation based on new technology, can improve efficiency in resource allocation and reduce transaction costs. Operators of two-sided multi-sided platforms can enhance their overall profitability through cross-subsidies. Unlike the closed systems in the traditional industrial economy, platforms can be open ecosystems, which makes it necessary to re-examine the relationship between the scale and efficiency of a given platform, and the level of the so-called “optimal scale” needs to be reassessed.
- Second, cross-border platform competition in the “digital economy” makes it necessary to redefine relevant markets. The competitive relationship between players in the “digital economy” needs to be interpreted in a broader sense than in traditional industries. For example, Qihoo 360 (an Internet security service provider) versus Tencent QQ (an instant messaging service provider), and WeChat (an instant messaging service provider) versus toutiao.com (an online news service provider). Traditional demand elasticity analysis may be difficult to apply in online markets. This is due to the similar cost structures of different Internet platform companies and the relative ease of cross-border expansion.
- Third, digital sectors are highly dynamic. Digital sectors are technological and creative. In other words, innovation is a key parameter of competition. This notion of “innovation” includes both technological and business model innovation. For example, Pinduoduo innovates in terms of supplying China’s vast “expanding markets” (the third-tier and below cities and rural areas in China). In as little as three years, it has become, in terms of size, a platform comparable to JD.com and also an important rival of Alibaba.5 Because of the cross-border attributes of the “digital economy,” basing antitrust analysis on market shares (as in traditional sectors) is controversial. On the one hand, it is difficult to define relevant markets in digital sectors. On the other hand, even if some companies have a higher market share in particular segments (and would therefore met the normal legal standard for “dominance”), competition has not necessarily been weakened in practice. In fact, it may have become much fiercer than in traditional industries. Under such circumstances, ease of entry becomes a crucial factor in assessing a given company’s market strength.
- Fourth, market structures in digital sectors tend to be in oligopolistic. Driven by network effects and the use of big data, oligopolistic market patterns often appear in certain segments. For instance, China has WeChat in the instant messaging field, Meituan and Eleme in the take-out service field, Alipay and WeChat Pay in the electronic payment market, and Baidu in the search engine market. Network effects are often significant in bilateral or multilateral markets, where consumers, merchants and other players are prone to being controlled by the leading platform operators, making these companies even stronger.
- Fifth, data has become one of the most important competitive resources in the “digital economy.” On the one hand, through the collection, research and application of big data, enterprises can analyze user preferences and make user portraits, so as to accurately promote their products and improve consumer satisfaction. In addition, with the help of big data, they can also expand their presence to neighboring markets and form a virtuous circle, thus consolidating and strengthening their presence in both the original and new markets. As companies continuously study and analyze a large amount of user data in depth, with the they can grasp information such as user preferences, consumption needs, consumption habits, and consumption levels, and launch, in a timely manner, customized products, place targeted advertisements and innovate in business models. This informational advantage will have a disruptive impact on business models and it can be expected that big data will become the core and most fundamental means of production for operators in the “digital economy.”
B. Impact of the “Digital Economy” on Competition
The essence of competition between players in the “digital economy” can be summed up in terms of two key considerations. One is to determine the target group, business model and resource allocation pattern (often based on big data) used by a given player. The other is how a given player realizes its cross-border trade and “network ecosystem.” With the continuous growth in scale of the of digital sector, there is a view that platforms are in essence monopolies, due to their use of data.
- Data Competition
Big data must be assessed in terms of the “6 Vs,” namely their volume, velocity, variety, value, veracity, and validation. The mainstream view is that data exhibit the attributes required to be subject to civil rights rules, and can also be classified as intangible property.6 Intangible assets are characterized by their replicability and low marginal cost. So, are data exclusive? This question is controversial in academic discussions.
Data have now become “resources” in competition analysis, i.e. a key input. In the background of increasingly fierce competition in data-driven industries, competition between operators for the acquisition and utilization of data is becoming more and more intensive. In view of this, some think that data should be considered to be an “essential facility.” Competition in and disputes over online data continuously emerge, in cases such as Sina Weibo versus maimai.cn in 2015, Dianping.com versus Baidu in 2016, coomix.com versus chelaile.net.cn in 2017, Taobao versus Meijing Group in 2017, Tencent versus Douyin and duoshanapp.com and Douyin versus shuabaola.cn in 2019.
- Ecosystem Competition
The modern Internet revolution has led to new economic and social developments. In the economic field, the “digital economy” has brought about new business models and organizational structures, and has created a large number of new industries, such as take-out, online ride-hailing, home stay, social networking, and e-commerce platforms. So-called “platforms” are designed to mediate the exchange of information between two or more parties in order to optimize the allocation of resources. As such, are platforms “neutral,” or do they have the characteristics of a “public good”? Should affiliated companies be protected?
From a practical point of view, competition among leading players is not limited to specific segments. Rather, platform-based companies compete with each other across segments, and display the characteristics of a “network ecology.” Examples include Taobao versus ebay, JD.com, Pinduoduo, Vipshop and Galanz; JD.com versus Tmall and dangdang.com; Didi versus Kuaidi and Uber. The services provided by digital platform operators are multifarious and address various user groups. Due to network effects, low marginal costs, and economies of scale, monopolies and oligopolies may be intensified and promoted. On the other hand, the centralization of data achieved through network effects and economies of scale may be to the advantage of users. Digital platform operators build their data-based business models through the accumulation and use of data. By further accelerating their accumulation and utilization of data, a cycle may be created to maintain and enhance their competitive advantages.
III. THE PATH FOR ANTITRUST REGULATION IN THE “DIGITAL ECONOMY”
The “digital economy” has the potential to greatly improve efficient of resource allocation, but it also carries the risk of restricting competition. Specifically, it not only changes the way in which companies compete, but also poses challenges to legal rules and their implementation. The influence of the digital economy on the AML is reflective of a worldwide problem, and also a common concern in major jurisdictions throughout world. Competition law enforcement agencies in the EU,7 the United States,8 Germany, Japan and other jurisdictions, as well as the OECD, the ICN and other organizations and research institutions, have launched investigations and published a series of research reports on the digital economy in recent years. The EU, the United States, Germany, etc. have raised their concerns about big-tech companies, such as Google, Facebook, Amazon, and Apple.
A. Chinese Rules for the Regulation of the Digital Economy
In China, the digital economy is supported by two fulcra, namely big data and online platforms. Competition problems arise in the following main forms: (1) Exclusivity conduct in search engines: (i.e. search companies ask content providers for exclusive supply of information); (2) Exclusive transactions by e-commerce platforms (i.e. e-commerce platforms require merchants not to trade with specific operators); (3) MFN (Most Favored Nation) clauses in platform transactions; (4) “swindling” practices relating to big data (i.e. applying different trading conditions to specific counterparties, often in the form of price discrimination); and (5) The price leadership of operators in certain market segments.
Under the current legal framework, there are three laws that regulate these questions.
- The E-commerce Law and the AML
The E-commerce Law (“ECL”) was formally implemented on January 1, 2019, with the purpose of regulating and protecting the rights and interests of parties involved in e-commerce. Article 35 stipulates the rules for platform operators: “E-commerce platform operators shall not use service agreements, trading rules, technology and other means to unreasonably restrict or attach unreasonable conditions to the operators’ transactions within the platform, the trading prices and their transactions with other operators, or charge unreasonable fees to the operators within the platform.” The players involved in an Internet platform are generally placed in three categories: P2P (platform to platform), P2B (platform to business), and P2C (platform to consumer).
The law’s main aim is to prevent the improper use of a platform’s advantages to harm the interests of merchants. But it does not concern the relationship between platforms themselves. As a matter of fact, of the above five issues, the first three involve the content of agreements between platforms and the merchants. The fundamental concern is that platforms may use their dominant positions to weaken the competitiveness of rival platforms through exclusive transactions with downstream merchants. Therefore, this clause is not applicable.
- Unfair Competition Rules and the AML
Cases concerning disputes over data competition have so far been tried under the framework of the Anti-Unfair Competition Law (“AUCL”), without the application of the AML. Regulation of unfair competition is carried out in two ways, namely applying the rule against infringing on trade secrets9 and the general clause.10 One of the major controversies is whether data are in fact “trade secrets.” According to Article 9 of the Interpretation of the Supreme People’s Court on Several Issues Concerning the Application of Law in the Trial of Civil Cases of Unfair Competition, trade secrets are defined as “trade information not known to the public.” The dispute in academic discussions is whether, if given data exist in the public domain, whether a dataset formed by their can be regarded as a “secret.” Relevant cases include Sina Weibo versus maimai.cn,11 Tencent versus Douyin and duoshanapp.com,12 and Dianping.com versus Baidu.13 User data existed in the public domain, and even though the platforms did not take any technical measures to block web crawling and other data acquisition, third parties still needed to spend time and money to acquiring “big data” by technical means, and to classify and screen that data to produce usable datasets. Therefore, some big datasets can be considered to be “secret” for the purposes of the AUCL.
The right to data is confirmed in all of the above cases. That is, given the efforts required to collect and process information, big data can be considered to be confidential and of economic value. Therefore, legal relief can be available for improper use of data. The AUCL and the AML pursue different goals, apply to different scenarios, and employ different means of protection. On the premise that such rights exist, is behavior prohibiting use of data by other operators in compliance with the law? To take another example, “swindling” through the use of “big data” is a commonly alleged abuse, but such harm to consumers cannot be solved through the AUCL.
B. Basic Principles of Antitrust Regulation in the Digital Economy
- Modest and Restrained use of the AML
In digital sectors, the key trading partners are typically the “triangle” of platforms, consumers and merchants. Only when the three parties have balanced market strength can this triangle be stable, and can the market develop in an orderly and healthy manner. Each of the three parties has different roles in market, different relative strengths, and different means of relief open to them. The legal rules are subject to a hierarchy. As a “constitution-like” law governing economic regulation, the AML is the last line of defense to solve any alleged market failure. Therefore, priority should be given to more targeted rules.
Specifically, in P2C and B2C relationships, consumers are in a weak position in the absence of collective action. For the moment, even if a platform or merchant is in a dominant position, priority must be given to the application of the Contract Law and the Law on the Protection of Consumer Rights and Interests. In a P2B relationship, if an alleged illegal act ultimately affects the rights and obligations of the two parties, the ECL should in principle apply.
- Application of the AML
Different laws protect different interests, and the legislative purpose of the AML is to maintain unrestricted competition. If there is a competition dispute between operators, it is necessary to look at the individual case to determine which rule to apply. Article 22 of the ECL stipulates that “e-commerce operators who are with market dominance because of their technological advantages, the number of users, their power to control related industries and the degree of dependence of other operators on them in transactions shall not abuse their market dominance to exclude or restrict competition.” For e-commerce operators that intend to abuse their market dominance, transactions between equal market players (which is beyond the scope of the ECL), should be governed by the AML.
In addition, when it is necessary to restore competition in specific cases, the AML prevails and cannot be replaced by the AUCL. For example, in 2018, the Haiyan County Administration for Industry and Commerce in Jiaxing, Zhejiang Province investigated and dealt with the first case of unfair competition using technological means in China.14 Specifically, in order to maintain their market share, Jiaxing Dongdongguai Network Technology Co. and Shanghai Youang Investment Co. resorted to improper technical means to narrow the delivery area of merchants on their platforms or set the merchants’ delivery area to “no man’s land,” thus forcing them to close down or stop operation on the “Shandianxiaoge” platform. Doubts about the case lie in that the two companies involved are technically supported by Meituan and Eleme and have no competitive relationship with “Shandianxiaoge” as a take-out platform. As such, how can one characterize their use of improper technical measures as a means “to maintain their market share”? The response is that, as instructed by Meituan and Eleme respectively, or in order to maintain the market share of Meituan and Eleme in Haiyan, and so as to maintain their own market share, Dongdongguai and Youang modified the delivery settings of specific merchants to prohibit them from trading with “Shandianxiaoge.” In the former case, it is a typical exclusive transaction and it is improper to apply the AUCL.
IV. AMENDMENT OF THE AML IN RESPONSE TO DIGITAL COMPETITON
When China adopted the AML in 2008, its Internet economy was not as developed as it is today. China’s Internet platform economy has entered a new stage of development now, with fundamentally different technologies, a different social basis, and different modes of development. The development of the Internet is characterized by rapidness, iteration and high-frequency innovation. Specifically, such progress includes the development of cloud services, big data, improvements in computing power, and the evolution of algorithms. In this context, regulatory concepts, ideas and technical means need to keep abreast of the times.
A. Digital Competition and Anticompetitive Agreements
- Horizontal Collusion
In the online world, collusion between operators, such as algorithmic collusion, is more technology-based, secretive, and therefore more difficult to prove than in traditional fields. Although once considered as a way to solve information asymmetry and increase social transparency, big data may also facilitate the development and running of anti-competitive algorithms. As such, it may become the prime tool used by data owners to cause consumer harm. If companies in a competitive relationship agree to use a specific pricing algorithm, this collusive behavior is more difficult to identify than traditional pricing agreements between operators.
- Collaborative Acts
Collaborative acts refer to acts of coordination between operators, in the absence of a clear agreement or decision.15 The identification of collaborative acts is difficult. In American Law there are concepts of “conscious parallel acts” and “implied collusion.” Article 13 of China’s AML 2008 sets out rules concerning “collaborative acts.” However, in the past 11 years, this article has never been enforced publicly or in private litigation due to issues of proof. However, in the case of abnormal price increases of chlorpheniramine16 and isoniazid,17 the penalty decision mentioned an abuse of “collective dominance” in an attempt to identify collusion between operators in the absence of sufficient evidence. In transparent digital markets, where low-cost automatic monitoring is possible, along with high speed responses, collaboration between operators is more feasible, and may be more difficult to be detected and proved. From a rational perspective, the standard of proof in administrative law enforcement should therefore be lowered in this field.
- Hub-and-Spoke Conspiracies
The case of insurance companies in Loudi, Hunan18 involves a typical hub-and-spoke conspiracy. In that case, 11 insurance companies in Loudi signed agreements with Ruite Insurance Brokerage Company to fix prices, boycott transactions, and divide regional markets. The case was filed by a local law enforcement authority and six insurance companies and trade associations were punished. No punishment was imposed on Ruite Company (which was in a pivotal position and regulated and supervised the implementation the agreements). According to the provisions of the AML on anticompetitive agreements, horizontal and vertical agreements should be distinguished from each other. In this case, Ruite Company and the insurance companies were not in a competitive relationship with each other, but rather in a vertical relationship. As such, this did not meet the main requirements for the finding of a horizontal monopoly agreement under the AML. However, this case was not an RPM (resale price maintenance) case by nature. This case illustrates the difficulty of applying the rules to a hub-and-spoke conspiracy.
In the digital economy, if operators share sensitive information, they can similarly engage in a conspiracy, with the platform as the “hub.” More importantly, the platform may assist or even organize operators to reach agreements, by relying on its special position in transaction. Examples includes Apple’s manipulation of the price of e-books in the European Union and the United States. In order to fix loopholes in the original legal rules, especially in response to illegal acts that are prone to occur on platforms, a proposed amendment to the AML would have prohibited operators from “organizing or helping other operators to reach monopoly agreements.”
B. Determination of Dominance in the Digital Economy
- The Three-Tier Determination of Market Dominance
The determination of market dominance is a very challenging technical issue under the AML. The proposed amendment adopts a progressive model: Paragraph 2 of Article 20 defines a position of market dominance as one held by an operator that has such a position in the relevant market that it can control prices, quantities or other trading conditions, or can hinder or affect other operators’ ability to enter the relevant market. However, evaluating whether an operator has market dominance is mainly based on the factors set out in Article 21, and the market share presumption set out in Article 22. Therefore, the system of determining market dominance consists of the following three tiers, namely, (1) the substantive standard (Paragraph 2, Article 20), (3) the factors set out in Article 21, and (3) the presumptions set out in Article 22. Market share is an important indicator for the determination of market dominance. As for the determination of market shares, the Interim Provisions specify that in addition to the amount and volume of sales, there are also other indicators such as the amount of data gathered, which provides a basis for more scientific and effective determination of the market share of online operators.
- Special Factors for Determination of Dominance Online
Article 21 sets out the factors for determining dominance, and its second paragraph sets out special considerations for determining the dominance of Internet operators. Compared with the provisions of Article 11 of the Interim Provisions,19 only three factors, i.e. network effects, lock-in effects, and ability to gather and process relevant data, remain (and economies of scale are added). This rule pays specific attention to the characteristics of the digital economy, namely the effects of platform models and big data.
The differences between the two texts concerning the elements to be used to determine dominance does not mean that there has been a significant conceptual change. Under Chinese law, the AML takes priority over the Interim Provisions. The latter are administrative rules and regulations, namely documents for their administrative management of public bodies that apply within their scope of authority. For concepts such as “market dominance,” the AML enforcement authority needs a specific, operable standard to make decisions. Such factors should be as detailed as possible, to help limit the discretion of law enforcement authorities and improve the transparency and predictability of law enforcement.
- Dispute over the Supplementary Determination Standards
Experts have different opinions on whether it is necessary to set out specific considerations to determine the market power of Internet operators. Some hold that the existing analytical framework is sufficient to deal with problems in all fields, including the digital economy, and that the AML is not a special case. Others hold a completely opposite view, stating that if the above-mentioned considerations are specified, the importance, complexity and uniqueness of this field must be highlighted, and special attention must be paid to the need for enforcement in the digital economy.
In fact, some uncertainty in the interpretation of the AML is inevitable. First of all, the AML protects diverse interests. How are benefits to consumers to be measured objectively? Second, from a technical point of view, whether a certain act should be governed by a per se rule or under a rule of reason is a century-long debate. Third, effects-based analysis is difficult, and conclusions based on mere economic hypotheses are not always reliable. Specifically, in terms of legal liability, sometimes illegal gains cannot be accurately measured.
Specifically, with regard to the source of laws, China only has statutory rules, and does not recognize case law, which means that statutory law must necessarily be abstract and principle-oriented in order to be of universal application. Under such basic principles, it is unrealistic to find a “legal” means that would account for the complexity and variability of the digital economy. A viable option is to set out practical and targeted documents formulated by administrative law enforcement authorities and courts, such as guidelines and judicial interpretations.
C. Market Concentration
As a preventive means of market intervention, examination of market concentration is subject to specific special rules in the digital economy.
- Increasing the Threshold for Examination
Due to the characteristics of the digital economy, big data are valuable, yet difficult to measure accurately. Sometimes when transaction data are rejected, there is no transaction volume for the relevant part of assets. Therefore, the expected income is often taken as the standard for security issuance and pricing. In the digital economy, the risk of “false negatives” is higher. In the examination of economic concentration, special regulations might also be needed, with the transaction value rather than the sales of the previous year as the threshold standard.
- Data Blocking and Stifling Acquisitions
Generally speaking, the anti-competitive effects of vertical integration are not as obvious as those of horizontal concentration. However, in the digital economy, special attention should be paid to vertical mergers where there would be a risk that other downstream participants would suffer from lack of access to key data. Although the United States and the European Union hold different views on vertical theories of harm, both raise concerns about possible data monopolization. In addition, stifling acquisition could also damage innovation. As such, behavioral remedies may be an appropriate response.
As the most important factor of production in the “new” or “digital” economy, data flows are necessary to achieve efficiencies and to encourage innovation. As an intangible asset, data can flow across many domains. Due to the different state of development of various data-related industries, the United States, the European Union, and China have shown different approaches towards the digital economy. However, there is no doubt that there will continue to be a trend towards international cooperation in antitrust enforcement in the digital economy.
1 Chenying Zhang, Associate Professor, Tsinghua University. This article is a result of research under the 2018 National Social Science Fund’s major project “Research on the Legal Protection of the Internet Economy” (project number: 18ZDA149).
2 According to Article 18 of the AML and Articles 6 to 10 of these Provisions, factors such as the competition characteristics of relevant industries, their operating mode, the number of users, network effects, lock-in effects, technical characteristics, market innovation, the ability to grasp and process relevant data, and the market power of operators in related markets may be considered in determining the market dominance of online operators. See http://www.gov.cn/gongbao/content/2019/content_5430507.htm.
3 Paragraph 2 of Article 21 of the AML (Revised Edition) sets out factors such as network effects, economies of scale, lock-in effects, and the ability to grasp and process relevant data that should also be considered in determining the market dominance of online operators.
6 See CHENG Xiao: On Personal Data Rights in the Big Data Era, Social Sciences in China, Issue 9, 2016.
7 The European Commission released a research report entitled Challenges to Competition Policy in the Digital Economy in July 2015; it launched a survey on the e-commerce industry in May 2015, covering online sales of daily necessities and online distribution of digital content. The final report of the survey was released in May 2017.
8 The United States also expressed its concern about potential algorithmic cartel behavior in its comments on the OECD report on algorithms and collusion in May 2017.
9 Article 9 of the Anti-Unfair Competition Law
10 Article 2 of the Anti-Unfair Competition Law.
11 (2016) J. 73 M.Z. No. 588 Second Instance Civil Judgment.
12 (2019) J. 0116 M.C. No. 2091 Civil Judgment.
13 (2015) P.M.S. (Z.) C.Z. No. 528.
14 Article 12 of the Anti-Unfair Competition Law: Operators shall not hinder or disrupt the normal operation of network products or services legally provided by other operators by influencing users’ choices with technical means.
15 Paragraph 3 of Article 5 of the Interim Provisions on Prohibiting Monopoly Agreements.
16 Administrative Penalty Decisions of National Development and Reform Commission,  No. 1 and 2.
17 Administrative Penalty Decisions of State Administration for Market Regulation, G.S.J.C.  No. 21 and 22.
19 See supra note 2.