[author]Hou Liyang

[content]

Hou Liyang: Alienation and Consequences of Monopoly Behavior in the Perspective of Platform Form Evolution

*Author: Hou Liyang

Professor, Kai Yuan Law School, Shanghai Jiao Tong University

Adjunct Researcher, Institute of Chinese Law and Society, Shanghai Jiao Tong University

Abstract: Treating platforms as a fixed business model makes it difficult to form a joint force in academic research, and also delays the upcoming platform regulation. Platforms have evolved from informationstorage intermediaries in Web 1. 0 to information-exchange intermediaries in Web 2. 0, and to businessaggregation intermediaries in Internet Ecosystem. Such an evolution not only changes the definition of platforms, but also causes expansions and variations of monopoly conduct, i. e. the expansions of relevant markets within Web 2. 0 and Internet Ecosystem, and the variations of the relations between platforms and non-competitors and between platforms and business operators. While the traditional anti-monopoly mechanism still maintains its effectiveness in regulating inter-platform monopoly conduct when integrating an evolutional view on platform models, it reaches its limit in coping with the variations of relations between platforms and business operators, namely the abuse by platforms of their organizing function to strengthen the profit-running function. To bridge the gaps in the anti-monopoly mechanism, it is necessary to adopt categorical regulation based on two platform types, namely to guide and monitor the making of intra-platform rules by big platforms, prohibit super platforms from applying self-preference, and shift the burden of proof to super platforms in cases of predatory acquisitions.

Key Words: Internet Ecosystem, digital platforms, two-sided market, anti-monopoly, regulation

1. Presentation of the problem

Since their creation in the 1990s, digital platforms have rapidly become a pillar industry of the world economy. With its increasing economic share, platform monopoly behavior has begun to trigger focused attention from competition enforcement agencies around the world. Platform antitrust enforcement in the international community began in 2016 with the EU's antitrust investigation into Amazon's implementation of the most-favored-nation clause in the sale of e-books (hereinafter referred to as the “Amazon case”). In China, it was concentrated after the Central Government put forward the requirement of “strengthening antitrust and preventing disorderly expansion of capital” at the end of 2020. Although China's law enforcement agencies have paid relatively short attention to platform antitrust, academic research can be traced back to the case of Tangshan Renren Information Service Co. v. Beijing Baidu Netcom Technology Co. Since then, research can be divided into three phases: the first phase (2009-2016) focuses on the relevant market and dominant position of the platform, with less research on monopolistic behavior and competition damage. Scholars in this phase affirmed the effectiveness of antitrust law in regulating the digital economy, and pointed out the challenges posed by the digital economy to the definition of the relevant market, but did not advocate the direct identification of the dominant position of digital platforms. The second stage (2017-2021) focuses on the study of monopolistic behavior and competitive damage of platforms, and less on the study of relevant markets and dominance. Scholars in this phase seem to completely ignore the research results of the first phase, and directly use the “winner-takes-all” feature of the digital economy to presume the dominant position of the platform. Some scholars have also suggested that the threshold of application of antitrust law needs to be appropriately relaxed to adapt to the changes in the digital economy. The third phase (2022 to present) concludes that some of the problems in the digital economy make antitrust law no longer effective, and that regulation needs to be introduced to make up for the shortcomings of the antitrust mechanism, i.e., the dualistic co-governance of antitrust law and regulation.

The existing literature shows that there is a problem of conflicting views in China's academic research on platform antitrust. This is mainly manifested in the first stage and the second stage of the determination of the dominant position of the platform, which triggers the first question of this paper - whether the platform in different research stages has the same definition. China's academic research on platform monopoly behavior is concentrated in the second stage, which in turn triggers the second question - what kind of transformation of the digital economy is implied in the concentrated research on platform monopoly behavior. Examining the transmutation and alienation of monopoly behavior according to the different forms of platform development leads to the third question to be addressed in this paper - what monopoly behaviors exist in the field of digital economy that are not regulated by the antitrust law, and whether and what kind of regulatory measures are needed to deal with them. These three issues are foundational in platform governance and relate to the future direction of platform governance.

2.the evolution of the platform's form and definitional changes

The starting point of platform antitrust research should be the definition of platform, otherwise it is difficult to conduct academic discussions in the same context. Digital platforms have been in the midst of morphological evolution since their creation in the 1990s. The different forms of platforms imply very different development goals, and the monopoly behaviors triggered by them also have different external manifestations and internal qualities.

2.1Web 1.0 platform as an intermediary for information storage

The emergence of digital economy is closely related to the development of Internet communication technology. The beginning of Internet communication technology is the packet switching technology developed in the late 1960s. Unlike the highly centralized carrier-based system built by the previous circuit switching technology, packet switching breaks information into packets, transmits each packet in a group, and then reassembles the packets once they reach their destination. Thus, this communication technology allows for peer-to-peer transmission between users and represents a decentralized transmission logic. However, the initial purpose of packet-switching technology was to enable the direct transmission of information between users, and did not necessarily require the intervention of a platform.

The commercialization of the digital economy began with Tim Berners-Lee's 1990 invention of the World Wide Web protocol. Commercial operation makes the scale of information transmission suddenly increased, pure point-to-point transmission can not meet the interaction needs of commercialization scenarios, thus, the platform as an information storage intermediary role came into being. This form of digital economy is called Web 1.0 (1990-2000), and Web 1.0 platform is defined by academics as a commercial organization that digitally stores traditional information and transmits or displays it to users in one direction through new Internet communication technologies. The representative platforms are web portals, and China's Sina, Sohu, Netease, etc. also appeared in this period.Web 1.0 platforms are far from the platforms we are currently discussing. The current universal definition of platforms in China is set forth in Article 2 of the Anti-Monopoly Committee of the State Council's Anti-Monopoly Guidelines in the Field of Platform Economy (hereinafter referred to as the “Platform Anti-Monopoly Guidelines”), which states that a platform is “a commercial organizational form that enables interdependent bilateral or multilateral subjects to interact with each other under rules provided by a specific carrier through network information technology, so as to jointly create value”. A platform is “a form of business organization that enables interdependent bilateral or multilateral subjects to interact under the rules provided by a specific carrier through network information technology, so as to jointly create value”. This definition requires that the platform must serve two or more user groups at the same time. Web 1.0 platforms only serve the same type of subjects, i.e., accessors of information. Although early e-commerce platforms also existed during the Web 1.0 period, the services they provided were mainly price lists of goods, which were significantly different from modern e-commerce platforms that facilitate transactions between merchants and users. As a result, Web 1.0 platforms differed little from traditional information intermediaries such as offline libraries and paper media, and only upgraded the traditional mode of storing information to digital storage.

The digital economy in the form of Web 1.0 presents a relatively intense competitive effect, and almost no platforms have gained monopoly status, and there is no monopoly problem in practice.Scholarly research on the digital economy at the end of the 20th century is still mainly limited to its challenges to national security and intellectual property infringement. Scholarship on platform antitrust is very limited, and most of it argues that the digital economy will not change the underlying logic of traditional antitrust mechanisms. For example, Posner points out that although the digital economy (then known as the “new economy”) is different from the traditional economy in many ways, the antitrust laws are still in effect and do not need to be specially adapted. Therefore, the Web 1.0 platform is excluded from the study of this paper.

2.2 Web 2.0 Platform as an Intermediary of Information Interaction

As an information storage intermediary, Web 1.0 platforms can only profit from advertisements, and they have to face competition from other platforms as well as (digitized) traditional media. The lack of innovative profit models caused the bubble of the digital economy to burst centrally around 2000. The mass bankruptcy of platform companies triggered a rethinking of the development model of the digital economy, which led to Web 2.0 (2001-2010). Compared with Web 1.0, Web 2.0 restructured the digital economy in the following three aspects:

First, the Web 2.0 form changed the business model and profit model of the platform. although the Web 2.0 platform used the same communication technology as the Web 1.0 platform and didn't change the digital economy from the technical logic, it increased the options of user participation, which made the platform have great openness. Gradually, users are divided into two groups with different needs, i.e., information providers (merchants) and information acquirers (users.) In addition to providing users with information in one direction, Web 2.0 platforms encourage users to participate in the provision and creation of information, which in turn promotes the creation of user-generated content. The center of gravity has shifted to creating a social network for collaboration and sharing. In this process, the platform gradually evolved into an intermediary for information interaction, promoting the development of the Internet economy together with merchants and users. At this time, platforms of historical significance include various social networking platforms, video and image sharing platforms, and e-commerce platforms. The platforms in the “Platform Antitrust Guidelines” should be Web 2.0 platforms. Accordingly, the source of income of Web 2.0 platforms is based on advertisement fees, with the addition of handling fees for transactions between merchants and users.

Secondly, Web 2.0 form has changed the function of the platform, Web 1.0 platform only assume the function of market profitability, and the user is a pure transaction relationship, Web 2.0 platform in addition to this, gradually began to assume the function of market organization. In the traditional offline economy, the market organization function is mainly constructed by the government through public power, but the application of these traditional mechanisms in the information interaction scenarios has greater limitations. For this reason, Web 2.0 platforms need to construct a set of trust and security mechanisms suitable for online transactions, such as online credit evaluation, third-party payment, online dispute resolution and so on. These new functions are presented in the form of rules within the platform, making the platform begin to change “the established state power structure and the relationship between the components of power”.

Third, the Web 2.0 format created the monopoly power of platforms for the first time. The real boom of the digital economy began with Web 2.0, which gave rise to the platforms we now know, such as Baidu, Ali, Tencent, and others.At the beginning of the development of Web 2.0, the competition between various platforms was extremely fierce. For example, Baidu faced strong sniping from Google, Yahoo and other search engine platforms until 2008, and even experienced a decline in market share. As Web 2.0 platforms began to gradually show the characteristics of the “winner takes all”, the trend of market concentration and development is getting stronger. Around 2010, the market position of each head platform was formally established. It was also during this period that platform antitrust cases began to appear in China, with the most representative cases being the “Baidu Case” and Beijing Qihoo Technology Co. v. Tencent Technology (Shenzhen) Ltd (hereinafter referred to as the “3Q Case”). (hereinafter referred to as the “3Q case”). Correspondingly, China's platform antitrust research was formally opened in 2009.

2.3 Internet Ecosystem Platforms as Business Aggregation Intermediaries

Although Web 2.0 has expanded the definition of platforms, platforms in this form are still only the transmission tools of digital information and have not yet been truly integrated with the real economy. Roughly since 2010, the digital economy began to establish the development direction of “Internet+”, i.e. the combination of the Internet economy and the real economy. The combination with the real economy has greatly expanded the business scope of the digital economy, creating a sharing economy that did not appear in the Web 2.0 form, such as online catering and takeout, online car rental, Internet rental, etc., and changing the digital economy once again.

First, the scope of business of platforms has been expanded again; Web 2.0 platforms focus on the operation of a single type of business, and this mode of operation makes the various head platforms inextricably linked to their business. For example, Baidu has even been transformed into a verb in colloquial language, meaning online search. However, under the guidance of “Internet Plus”, platforms have begun to get rid of the development mode of single business operation, and become intermediary platforms for multiple business aggregation. This phenomenon has been named “Internet ecosystem” by academics, that is, “through the construction of a multilateral group cooperation and win-win mechanism to make the platform ecosystem in the core business, driven by the various derivatives of the business modules covered by organic synergy and the formation of the system”. There are two main types of Internet ecosystems: production-based ecosystems and advertising-based ecosystems. The former is built around a device (e.g., Apple) or service (e.g., Taobao), while the latter generates advertising revenue through its main business (e.g., WeChat, Baidu, etc.), which is provided free of charge. While the goals of these two types of ecosystems are the same, there are differences in the way they are monetized. Production-based ecosystems use improving the quality of their core business as their main means of profitability. Advertising-based ecosystems do not derive their main profit from the user side, so they are using increased user stickiness as a way to gain advertising profit. The academic community has not yet a unified name for this, this paper is called the Internet ecosphere form (2010 to the present), that is, through the network information technology, so that a variety of synergistic platform business in the core platform business under the connection of the aggregation, so as to jointly create value of the business organization form.

Secondly, the business model of platforms has been changed again. Web 2.0 platforms operating a single kind of business take the improvement of platform business quality as the main development goal. Internet ecosystem platforms, on the other hand, derive a cross-industry business model and gradually develop into an aggregated platform that unites all kinds of Internet businesses, providing users with a “one-stop” consumption experience and satisfying their multiple needs. Its business expansion is mainly carried out at two levels: the first level is that the platform enters other platforms in addition to its main business. This model can be specifically subdivided into two sub-models, respectively, the platform through their own development to provide other platforms, such as Tencent in WeChat implanted in the circle of friends, public, small programs, third-party payment, etc., and the platform through mergers and acquisitions, shareholding and other ways to absorb other platforms to enter the same ecosystem, such as the Ali Group merged with and acquired the Gaode map, and so on. The second level is that Internet ecosystem platforms utilize the transaction information they have within the platform to select the merchant business areas with the highest profitability to enter the business in person and engage in direct competition with merchants.

2.4 Triple Expansion in the Changing Definition of Platforms

The above three forms are superficially just the upgrading and iteration of the digital economy business model, but at a deeper level they imply a change in the definition of the platform. Specifically, these three forms represent the successive expansion of the platform definition at three levels:

First, the expansion of market players within the platform. web 1.0 platform is mainly an intermediary for information storage, facing users with the same demand for information, and its main function is the one-way display of information, so it involves only two types of market players: the platform and the users. web 2.0 platforms have developed into intermediaries for information interaction, whose purpose is to assist merchants and users in direct or indirect transactions, and to build up a triangular structure between the platform and the users. Web 2.0 platforms have developed into information interaction intermediaries whose purpose is to assist merchants and users in direct or indirect transactions, building a triangular structure between platforms, merchants and users, and upgrading the number of market players within the platform to three or more. The Internet ecosystem platform develops from the core platform business to other platform businesses, forming a structure with the core platform business as the axis and other sub-platform businesses as the spokes, providing merchants and users with multi-business aggregated goods or services. Therefore, in addition to the merchants and users, the market entities within the Internet ecosystem platform have added a brand-new market entity of sub-platforms.

Secondly, the expansion of the legal relationship within the platform, Web 1.0 platform is only a provider of information services, and the contractual relationship between it and its users is the provision of information, Web 2.0 platform in addition to the provision of information services to the merchants and users, but also to ensure that the smooth progress of the transaction between the merchants and users. Therefore, Web 2.0 platform is firstly a contractual relationship of intermediary service between merchants and users. Secondly, due to the platform's new market organization function, a second legal relationship, i.e., market management relationship, has arisen between the platform and the merchants and users. Despite the existence of dual legal relationships, in the Web 2.0 form, the platform creates value with merchants and users, and the three belong to a community of interest. Under the transformation of the Internet ecosystem, the relationship between platforms and merchants has become complicated, and the two are no longer just a cooperative relationship. With the Internet ecosystem platform into the business field of merchants, the development of a third legal relationship between the merchants, that is, the competitive relationship.

Third, the expansion of the platform market scale. the emergence of Web 1.0 form greatly reduces the transaction costs, so that the platform to get beyond the scale of traditional enterprises, that is, the first expansion of the platform scale. However, Web 1.0 platform did not really break through the traditional nature of the enterprise. In contrast, as an intermediary of information interaction, Web 2.0 platform breaks the boundary between the enterprise and the market. Presented in the form of a market, the Web 2.0 platform expanded the scale of the platform for the second time, and initially demonstrated the “winner-takes-all” effect of the digital economy. The third expansion of the scale of the platform in the form of Internet ecosystem makes the platform no longer a decentralized operation of a single kind of business, but an aggregated operation based on the core platform business. Different platform businesses are horizontally connected by the two important assets of attention and data, coupled with the integration with the real economy, the scale of the platform is further expanded. The repeated expansion of platform scale has not only triggered the attention of academia, but also that of the power organizations in various countries. Most countries have begun to change their previous laissez-faire regulatory attitudes toward the digital economy, and are exercising strict control by strengthening antitrust enforcement or enacting new regulatory statutes.

Since the 1990s, digital platforms have evolved from information storage intermediaries in the form of Web 1.0 to information interaction intermediaries in the form of Web 2.0, and then to business aggregation intermediaries in the form of Internet ecosystems today. The morphological evolution of the digital economy has indeed triggered changes in the definition of platforms. In terms of time development, the first stage of China's antitrust research on platforms corresponds to Web 2.0 platforms, and the second stage corresponds to Internet ecosphere platforms. The greater the number of market subjects within the platform, the more complex the functions of the platform, and the larger the scale of the platform, means the greater the type and number of platform monopoly behaviors. Therefore, Internet ecosphere platforms will definitely involve more monopoly behaviors than Web 2.0 platforms. This responds to the phenomenon that the first stage of China's platform antitrust research mainly studies related markets and the second stage mainly studies monopoly behaviors.

3. Monopoly Behavior in Web 2.0 Formation

Combining the expansion of the platform definition with the antitrust analysis, two key points for the study of platform monopoly behavior can be found: first, the impact of the change in the platform definition on the definition of the relevant market, and second, the impact of the change in the platform definition on the type of monopoly behavior. The analysis of the latter can be carried out at two specific levels: one is the monopolistic behavior arising from inter-platform (or inter-ecosystem) competition, and the other is the monopolistic behavior arising from intra-platform (or intra-ecosystem) competition.

3.1 Dilemma and Solution of Defining Markets Related to Web 2.0 Platforms

The prevailing theory used by academics to explain the shape of Web 2.0 is the bilateral market. Bilateral markets have three core characteristics: (1) there are two or more different user groups, and these groups must interact through the platform. (2) There is a cross-network effect between these user groups, i.e., an increase in the number of users in one group triggers an increase in the number of users in the other group. (3) The platform imposes an unbalanced pricing structure on these user groups, often in the form of a free-of-charge strategy for a particular user group. From the perspective of Article 2 of the Platform Antimonopoly Guidelines, China's current definition of a platform should be a bilateral market platform based on the Web 2.0 format.

Traditional unilateral market involves only one kind of goods or services, and the competition mainly centers on that kind of goods or services, while the bilateral market in the form of Web 2.0 involves more than two kinds of goods or services, and the competition needs to be centered on a variety of goods or services. Therefore, the scope of competition in a bilateral market is larger than that in a unilateral market, which is the internal reason why the market size of a Web 2.0 platform is larger than that of a Web 1.0 platform. In the first phase of the platform antitrust research, the controversy in China's academic community is mainly about whether the definition of the relevant market in the bilateral market should be based on one of the user groups, or whether it needs to encompass all the user groups. The academic community's dilemma lies in the fact that no matter which one is chosen, there is an enforcement dilemma: on the one hand, if the relevant market is defined based on a certain user group only, it will not be able to reflect the whole picture of the bilateral market. Bilateral marketplaces need to price differently for different user groups, and they often use free or low-pricing strategies to build up the number of such groups in order to attract other paying user groups. If one group of users is considered alone, then a bilateral marketplace's free strategy for users may involve predatory pricing. On the other hand, if all user segments are included in the definition of the relevant market, the relevant market may be too large for a dominant position to be recognized. This is mainly due to the fact that attention has become a “currency” in the digital economy, and if all platforms can be categorized in the same relevant market for the purpose of substitution analysis based on attention alone, then no single platform will have a dominant position.

Judging from recent domestic and international platform antitrust enforcement practices, this theoretical concern has not posed a real challenge to judicial and enforcement agencies. Bilateral markets can be categorized into transactional bilateral markets and non-transactional bilateral markets according to whether there is a direct transactional relationship between merchants and users. Typical examples of the former are e-commerce platforms. The main groups they serve are merchants and users, who must transact through the e-commerce platform. A typical example of the latter is a search engine platform. The groups it serves are advertisers and users, and although the two interact with each other, there is no real transaction activity. In addition, the user groups served by these two bilateral markets have different needs. For e-commerce platforms, the needs of merchants are sales, and the needs of users are purchases. For search engine platforms, the needs of advertisers are to place ads and the needs of users are to obtain information. Given the differences between these two types of bilateral markets, judicial and law enforcement agencies have adopted different approaches to defining the relevant markets.

First, based on the fact that the core of the transaction-based bilateral market is the transaction between merchants and users, the relevant market of this type of bilateral market is defined in practice as the business of the platform that realizes the relevant transaction, i.e., all user groups are included in the definition of the relevant market. For example, in the case of Alibaba Group Holding Co., Ltd.'s abuse of dominant market position (“Ali case”), the State Administration for Market Supervision and Regulation (“AMSR”) defined the relevant market as the “online retailing platform service market”. In the case of Meituan Abuse of Market Dominance (“Meituan Case”), the State Administration for Market Supervision and Administration defined the relevant market as the “online catering and takeaway platform service market”. The theoretical difficulty in defining all user groups as the same relevant market lies in the question of what user groups are to be recognized as a dominant market position. However, this issue does not lead to practical difficulties. The existence of cross-network effects increases the market power of transactional bilateral marketplaces on the merchant side in direct proportion to the market power on the user side. Whether the dominance of such platforms is measured on the merchant side or the user side, the results should be the same. For example, in the Meituan case, the enforcement agency measured Meituan's market share on the merchant side at 70.7% and on the user side at 68.5%, which are nearly equal. Therefore, defining a transactional bilateral market as a unified relevant market becomes a natural choice in practice.

Secondly, there is a lack of direct transactions between merchants and users in non-transactional bilateral markets, and in practice the relevant markets in such bilateral markets are defined based on different user groups, thus including multiple relevant markets. China has not yet had a case of non-transactional bilateral market, but the EU has made several antitrust penalties for Google search engine. In the 2017 “Google comparison shopping case”, given that Google's monopoly behavior acts on search engine services, the relevant market of the case was defined as the search engine services market, which actually defines the relevant market from the user side. In the 2019 Google Ads case, the monopolistic behavior involved in the case was on the Internet advertising side, so the relevant market in the case was defined as the Internet advertising market. Thus, the EU defines the relevant market of non-transactional bilateral markets based on the user groups on which the monopoly acts. In response to the concern that the scope of the relevant market may be expanded indefinitely, the EU argued that the attention generated by the business of different platforms is not substitutable. Therefore, in the Google Comparison Shopping case and the Google Ads case, the EU recognized the attention gathered by search engines as a separate relevant market, which is not substitutable with the attention generated by other platforms. As a result, academic concerns about the definition of relevant market in non-transactional bilateral markets have also been addressed by enforcement practice.

3.2 Monopolistic behavior in inter-platform competition

In the economic model of bilateral markets, platforms face four main types of subjects in their operations: merchants, users, non-competitive platforms, and competitive platforms. The behavior between platforms and merchants and users belongs to the category of intra-platform competition, and the behavior between platforms and competing platforms and non-competitive platforms belongs to the category of inter-platform competition. Among them, Web 2.0 platforms and non-competitive platforms can realize the common enhancement of traffic in the process of interconnection and interoperability, so they belong to the community of interest, and monopoly disputes are less likely to occur.Web 2.0 platforms do not have a trading relationship with competitive platforms, so they can't implement monopoly behaviors on competitive platforms directly, and they can only implement monopoly behaviors indirectly through the merchants and users that they serve.

The theme that cannot be bypassed in the study of platform monopoly behavior is the “winner-takes-all” effect of the digital economy. The digital economy is indeed characterized by a large market size compared to the traditional economy. Although this characteristic has been present since the beginning of the digital economy, firms in the traditional economy are often unable to capture the entire market due to transaction costs. As a result, there has been a distinction between firms and markets in the traditional economy. Because of the inability to capture the entire market, the competition model in the traditional economy is also known as “competition in the market” (competition in the market). In other words, each firm operates with the aim of acquiring a larger share of the market, but is never able to form a monopoly. In a bilateral market, the existence of cross-network effects means that the size of the user base is the key to a platform's competitive advantage. The larger the user scale, the stronger the competitive advantage. As a result, the competition model in the Web 2.0 format has shifted to “competition for the market”. In other words, the platform must capture as much of the market as possible, or even monopolize the entire market, which is the internal cause of the “winner-takes-all” phenomenon.

Although the “winner-takes-all” effect can explain why Web 2.0 platforms have formed a super-large scale, it is difficult to directly prove that digital platforms will definitely form a monopoly. From the point of view of the development of China's digital economy, the number of platforms that can really be “winner-take-all” is very small, and the most important reason for this is the multi-homing effect (multi-homing). The multi-homing effect refers to the fact that users can have multiple applications with similar platforms at the same time, and choose the most suitable platform each time they use it. In contrast, single homing is the tendency for users to use only one platform despite the existence of multiple competing platforms. Platforms with single homing effect will naturally gain monopoly position, but there are very few platforms that naturally have this effect, such as instant messaging platforms. Therefore, in order to gain a monopoly, Web 2.0 platforms need to implement strategic behaviors to break through the limitations of the multi-homing effect, to create heterogeneity of their own platforms, and to artificially create a single-homing effect. To this end, the main monopolistic behaviors that platforms may implement are exclusive dealing and MFN clauses.

First, the most direct means of restricting the development of a competing platform is to require its own merchants or users not to use the competing platform's services, i.e., exclusive dealing (or “either/or”). As there are multiple user groups in the bilateral market, exclusive dealing can be imposed on users, such as in the “3Q case”, or on merchants, such as in the “Ali case”, the “Meituan case”, and the “Tongfang Knowledge Network (Beijing) Technology Center” case. (hereinafter referred to as the “China Knowledge” case), and the abuse of market dominance by Shanghai Food Pace Commerce and Trade Development Co. The ultimate goal of exclusive dealing by platforms is to prevent competing platforms from achieving a scale advantage and thus being crowded out of the market. Exclusive dealing can be manifested either as a direct restriction, such as forcing or coercing merchants or users not to trade with competitors, or as an indirect restriction, such as attracting merchants or users not to trade with competitors by means of lower fees, loyalty discounts and other benefits. Exclusive dealing has the dual effect of excluding and restricting competition as well as promoting competition, so the general view is that “the principle of per se violation cannot be applied to it, and only the principle of reasonableness can be applied to determine its illegality”. Exclusive dealing exists in both the traditional and digital economies. When hearing cases of exclusive dealing in the traditional economy, the Supreme People's Court pointed out that the determination of the competitive effect of such behavior needs to consider whether it “raises the barriers to entry into the market or increases the costs of competitors, thus creating a market blocking effect”. In cases related to the digital economy, enforcement agencies also focus their analysis on whether exclusive dealing “creates a lock-in effect, reduces its own competitive pressure, and unduly consolidates and strengthens its own market power”. There is no fundamental difference between the two; in other words, the analytical approach of traditional antitrust law is equally applicable to the digital economy.

Second, unlike exclusive dealing, which realizes the single-host effect through the uniqueness of goods or services, the most-favoured-nation clause realizes the single-host effect through the uniqueness of price (the lowest price). A broad MFN clause means that a platform requires a merchant to sell goods or services on its platform at the lowest price on the network, including the merchant's prices on competing platforms as well as on its own website or in offline stores. A narrow MFN clause only means that the platform requires the merchant to set prices on its platform that are no higher than the prices on its own website or in its offline stores. The common academic view is that a broad MFN clause raises the cost of competition for small and medium-sized platforms, especially when the homogeneity of the platform's business is obvious, and creates a strong single-host effect. In contrast, narrow MFN clauses have both pro-competitive and pro-competitive effects. On the one hand, they can prevent free-riding behavior by merchants who promote their goods or services through the platforms but bypass the platforms to conduct transactions, but on the other hand, they also carry the risks of market blocking and price collusion. Therefore, both have to apply the analytical model of the principle of reasonableness. This suggests that the MFN clause implemented by the platform does not exceed the scope of adjustment of the traditional antitrust mechanism. There have been no cases involving platforms with MFN clauses in China, but the behavior has been penalized in the traditional economy. Enforcement agencies have pointed out that MFN clauses are normally not sufficient to have the effect of excluding or restricting competition, and should only be prohibited when they have a market lock-in effect. The EU has recently responded similarly to MFN clauses in the digital economy, both at the antitrust legislative and enforcement levels. At the legislative level, the EU is of the view that MFN clauses in the narrow sense tend to have pro-competitive effects and are therefore subject to wholesale exemptions, but MFN clauses in the broad sense may have market-locking effects and should be subject to a rational principles analysis. At the enforcement level, the European Union found the MFN clause in the Amazon case to be unlawful in the sale of e-books by Amazon, precisely because the conduct in question, combined with Amazon's dominant position, had the effect of creating a market blockade. Therefore, the MFN clause in the digital economy does not show any difference from the traditional economy, and there is no obstacle to the application of the traditional antitrust law.

3.3 Monopolistic behavior in intra-platform competition

Intra-platform competition involves the relationship between platforms and merchants and users. In order to better safeguard the operation of the bilateral market, Web 2.0 platforms will provide better quality services as much as possible, and will not, in principle, have conflicts of interest with merchants and users. However, from the point of view of China's practice, monopolization disputes between platforms and merchants and users may occur in exceptional cases.

One is the refusal to trade. The digital economy has evolved into an attention economy. The importance of attention is such that platforms often do not impose excessive restrictions on users. Therefore, the main subject of conflict with the platform is mainly merchants, mostly involving the platform's blocking of merchants, with typical cases being the “Baidu case” and Xu Shuqing v. Shenzhen Tencent Computer System Co., Ltd. and Others Abuse of Dominant Market Position (hereinafter referred to as the “Xu Shuqing case”). (“Xu Shuqing”). In the Baidu case, the plaintiff was blocked by Baidu for artificially manipulating search results. The court noted that Baidu's blocking was lawful because if Baidu did not control the artificial increase in search rankings, it would lead to disorderly competition in the search business and reduce the quality of users' searches. In the Xu Shuqing case, the plaintiff was blocked by Tencent for publishing WeChat emoticons containing advertising slogans. The court ruled that the plaintiff's behavior violated Tencent's in-platform rules, and therefore “should be resolved under the framework of contract law, rather than directly resorting to antitrust law”. These cases essentially involve the platforms' activities in regulating merchants' misconduct based on their function as market organizations. The existence of common interests makes platforms lack the incentive to implement monopoly behavior on merchants. Therefore, these so-called “monopoly disputes” do not involve real monopoly behavior, but rather conflicts caused by the platform's implementation of market management behavior, which in principle belongs to the category of self-regulation, and should be dealt with by platforms in accordance with the rules of the platform.

Secondly, ultra-high pricing, Web 2.0 platform in theory, there is the possibility of implementing ultra-high pricing. The behavior may occur between the platform and the merchants, may also occur between the platform and the user, depending on the pricing structure of the platform. Either way, however, it can only occur between the platform and the paying transactional subject, and is not beyond the scope of adjustment of traditional antitrust mechanisms. For example, in the China Knowledge case, the enforcement agency relied on the antitrust law to find that China Knowledge's behavior constituted excessive pricing. However, it should be emphasized that Web 2.0 platforms are communities of interest with merchants and users, and the probability of excessive pricing in the digital economy is very low. The report issued by the Organization for Economic Cooperation and Development (OECD) does not even list overpricing as a platform monopoly that national law enforcement agencies must be concerned about. Even the China Knowledge case can hardly be characterized as a pure case of excessive pricing. In that case, the enforcement agency identified two types of monopolistic conduct, exclusive dealing and excessive pricing, but focused its analysis on exclusive dealing, finding that China Knowledge's price exceeded the previous price and the price of its competitors, which constituted excessive pricing. In antitrust law, determining the illegality of excessive pricing requires a measurement of the relationship between price and cost. Any platform always attracts users by losing money at the beginning of its operation, and it is not immediately obvious whether subsequent price increases can cover the sunk costs of the previous period. The enforcement agency's finding that the behavior was unlawful without comparing China Knowledge's price to its operating costs is not entirely consistent with the ultra-high pricing analysis. However, the case also shows that platforms usually do not only implement ultra-high pricing behaviors, but also use other monopolistic behaviors as auxiliary means, at this time, anti-trust enforcement of other monopolistic behaviors can effectively avoid the problem of ultra-high pricing.

Third, big data kills maturity. Big data kill familiarization refers to the platform for different users to personalized pricing behavior. Although personalized pricing can be manifested as setting lower prices for new customers than for familiar customers, the academic community is mainly concerned about the platform's behavior of setting higher prices for users with high consumption capacity. The main reason for the creation of big data killing is that platforms can predict the behavioral patterns of users by analyzing their data. However, so far, big data kills maturity is only a media speculation, and there is not yet any evidence that platforms are generalizing the behavior. Therefore, at least at this stage, it is not yet relevant to explore the antitrust regulation of the behavior. Even if the antitrust law is not applicable, other existing laws in China can effectively address its regulation. For example, Article 14 of the Personal Information Protection Law requires platforms to obtain prior consent from users before processing personal information. If a platform collects user data for the purpose of practicing big data killing, it must inform the user in advance, at which point the user can exercise his or her veto. As another example, Article 18(1) of the E-Commerce Law states, “Where an e-commerce operator provides a consumer with search results for goods or services based on the consumer's interests, hobbies, consumption habits, and other characteristics, the operator shall at the same time provide the consumer with options that do not target his or her personal characteristics.” Big data kills often need to be combined with information searches, so the article similarly gives users veto power. In short, until big data kills maturity is fully proven, the above laws can already better protect the interests of users without having to resort to antitrust laws.

Web 2.0 platforms have created an unprecedented scale of digital platforms by virtue of their function as intermediaries of information interaction, which has changed the competition mode and business model of traditional enterprises. This new business model has indeed brought some difficulties to the enforcement of traditional antitrust laws, but it has not led to the systemic collapse of the latter.The challenge of Web 2.0 platforms to the traditional antitrust enforcement mechanism is mainly manifested in the definition of the relevant market. Once this problem is solved, there is no essential difference between platform enterprises and traditional enterprises in terms of monopoly behavior in inter-platform competition. Compared with traditional enterprises, Web 2.0 platforms are endowed with the function of market organization, which triggers the illusion of monopoly in intra-platform competition, but such disputes belong to the scope of platform self-regulation, and in principle do not require external control by the government. This is also the deep-seated reason why our academia mainly focuses on the relevant market and dominant position and less on the platform monopoly behavior in the first phase of platform antitrust research.

4. the monopoly behavior in the form of Internet ecosphere

The emergence of the Internet ecosphere, on the one hand, is related to the new user dividend period after the decline of the platform revenue generation pressure surge, on the other hand, is related to the development of common cost convergence between different platform business. The Internet ecosphere form not only further expands the scope of the digital economy, but also deepens the interaction between platforms and other market players, thus triggering more new types of monopolistic behaviors.

4.1 Expansion of Relevant Markets and Their Responses in the Internet Ecosphere

In the form of Internet ecosphere, the competition between platforms is no longer the competition of single kind of business, but the competition of aggregated business based on the core platform business. The real business purpose of the ecosphere platform is to continuously test the development boundaries of the two important assets, attention and data, in the digital economy around the core platform business, so as to maximize the operation of both. This new business model expands the scope of the ecosphere platform's operation while also expanding the scope of the relevant market.

For one thing, the center of gravity of competition in the Internet ecosphere form has begun to converge on the core platform business. When the platform develops to the form of Internet ecosphere, the role of the core platform business begins to highlight, and the non-core platform business which is difficult to form the aggregation effect in principle does not need to be the object of antitrust law. Article 2 of the EU Digital Market Law only lists online intermediary services, online search engines, online social networks, video sharing, instant messaging, operating systems, web browsers, virtual assistants, cloud storage, and online advertisements as core platform businesses, and does not regulate other platform businesses. China's law enforcement agencies have yet to pay attention to the significance of core platform businesses in the shape of the Internet ecosystem. For example, at the end of 2021, the State Administration for Market Supervision and Administration (“SAMSA”) issued the “Guidelines for the Classification and Grading of Internet Platforms (Exposure Draft)” (“Platform Classification and Grading Guidelines Opinion Draft”) and the “Guidelines for the Implementation of the Main Responsibility of Internet Platforms (Exposure Draft)” ("Platform Responsibility Guidelines Opinion

Among them, the “Opinion Draft on Guidelines for Classification and Rating of Platforms” provides ideas for defining the relevant market, but it defines a total of 31 types of platform business, bringing almost all types of platform business into the scope of regulation. The logical assumption behind this should be that each type of platform business has the potential to become the core of the ecosystem. However, from an objective point of view, it is difficult for most of these businesses to become core platform businesses, such as vertical commodity trading, super group purchasing, home services, financial information, and so on. This categorization of platforms means that China's law enforcement agencies are still stuck in the Web 2.0 form of platform regulation, and have not yet noticed the major competitive transformation brought about by the Internet ecosystem form.

Second, the Internet ecosystem is accelerating the integration of traditional platforms. China has yet to see a case of monopolization by an ecosystem platform, but the recent case of Tencent's banning of Jitterbug implies a scenario of competition between ecosystems. The case involves Tencent's banning of ShakeMe video links in WeChat. According to bilateral market theory, WeChat is in the instant messaging market and ShakeMe is in the short video market, and they are in different related markets. However, it is strange that among the many ecosystems, the Tencent ecosystem is mainly targeting the ShakeEn ecosystem to implement the banning behavior, and other ecosystems are relatively in a state of peaceful development. This means that the emergence of the Internet ecosphere form has led to the beginning of an overlap in the scope of business between Tencent and ShakeEn. This shift has not yet been brought to the attention of our law enforcement and judicial bodies.On September 6, 2023 the European Commission designated Tiktok as a platform gatekeeper pursuant to Article 3 of the Digital Market Law and characterized it as a social networking platform, in the same relevant market as Facebook, Instagram and LinkedIn. WeChat's circle of friends is also a social network. According to this definition, Jieyin and WeChat are in the same relevant market after the expansion of the boundaries of social networks. This also explains why Tencent is specifically targeting ShakeMe and not other ecosystems. It seems that the Internet ecosystem format has indeed triggered the convergence of traditional platforms, thus creating a wider range of related markets.

Unlike Web 2.0 platforms, which mainly expand the scale of the digital economy from the perspective of economies of scale and seek to “compete for the market” by reducing transaction costs through Internet technology, the emergence of Internet ecosystem platforms has once again strengthened the new competitive model of the digital economy from the perspective of economies of scale. From the perspective of the platform, the Internet ecosystem form can break through the boundaries of attention and data, enabling the platform to reduce the operating costs between different businesses. From the user's perspective, the “one-stop” consumption ecosystem allows users to log in to multiple Internet platforms at the same time through a single account, which improves the experience of using the services of different platforms. In addition, the simultaneous provision of multiple services can also have a monopoly effect by limiting competition from platforms operating a single service. Confined to the decline of the dividend period of new users, platforms enter the Internet ecosystem form to reduce operating costs on the one hand, and to grab users with competitors on the other. As a result, competition in the Internet ecosystem form is more intense and more likely to raise concerns about abuse of dominance by platforms.

For business aggregation in the Internet ecosystem, there is a cluster market in the antitrust law, which defines multiple businesses that can generate strong economies of scope as one relevant market. The best example of a cluster market is a supermarket. Although each product in a supermarket should be defined as a separate relevant market, national judicial and enforcement agencies have always defined it as a package of supermarket-type goods. The Internet ecosystem is structured precisely as a grouped product market, not as multiple markets for a single platform business. Judging from the current market development, each ecosystem is still mainly built around the core platform business. Therefore, the relevant market definition in the Internet ecosystem should also be centered on the core platform business, and other non-core platform business should be included in the group product market composed of the core platform business. However, one of the difficulties at this point is that the type of core platform business is under continuous development and evolution, and it is not yet known what the final result will be. Another difficulty is that business aggregation is still changing the boundaries between core platform businesses, and recognizing Internet ecosystem platforms as Web 2.0 platforms will inevitably lead to false-positive errors. In this regard, judicial and law enforcement agencies need to constantly observe the trend of platform business convergence in the creation of the Internet ecosystem, so as to determine what core platform business should be based on to define the relevant market in China.

4.2 Monopolistic behavior in inter-ecosphere competition

After entering the form of Internet ecosphere, the relationship between platforms and competing platforms has not changed much, and it is still monopolistic behavior triggered to create a single-host effect. However, the relationship between platforms and non-competitive platforms starts to alienate, especially when entering into the business of non-competitive platforms, Internet ecosphere platforms may weaken or even stop the cooperative relationship with non-competitive platforms, thus generating new types of monopolistic behaviors, mainly self-preferential treatment and platform blocking.

Self-preferential treatment and platform blocking have different behavioral appearances, but they are unified in terms of monopoly purpose. Self-preferential treatment means that in order to promote the development of “self-owned/affiliated sub-platforms”, “core platform 1” grants them more favorable treatment than “competing sub-platforms” that are subordinate to “core platform 2”. Competitive sub-platforms” in order to promote the development of ‘self-owned/affiliated sub-platforms’ over ‘core platform 2’. This preferential behavior develops to the extreme of platform blocking, i.e., in order to give preferential treatment to the “self-managed/affiliated sub-platforms”, the “core platform 1” no longer conducts any transactions with the “competing sub-platforms”. This is the aforementioned “Tencent's blocking of Jitterbug” case. Most Chinese academics believe that self-preferential treatment belongs to the use of leverage to crowd out competitors, and that these behaviors can be subdivided into refusal to deal, tying, and differential treatment according to the manner of implementation. Most of the existing studies believe that the competitive harm of self-preferential treatment is the leverage effect of the behavior, i.e., the core platform, through the implementation of self-preferential treatment, transmits its dominant position to the relevant market where the sub-platforms are located.The analysis emphasizes that “self-owned/affiliated sub-platforms” have gained a perverse competitive advantage with the help of “core platforms1”.

Figure 1 Inter-ecosystem Competitive Transformation

The above analysis has fragmented the role of “Core Platform 2” in the inter-ecosystem competition. With the horizontal integration of two important assets, attention and data, “Core Platform 2” creates an aggregated business platform, i.e., through business integration, it promotes the development of the business of “competing sub-platforms” on the one hand, and obtains the “competing sub-platforms'” business advantages on the other. That is, through business integration, on the one hand, it promotes the development of the business of “competing sub-platforms” and on the other hand, it obtains additional attention and data brought by “competing sub-platforms” to strengthen the development of the core platform business. Therefore, when platform competition is upgraded to Internet ecosystem competition, the enforcement agency should include “core platform 2” in the competition analysis, focusing on whether it can respond to various monopolistic behaviors of “core platform 1”. Failure to consider the competitive role of Core Platform 2 may lead to false-positive errors in the competition analysis.

For one thing, when encountering the self-preferential treatment of Core Platform 1, the “competing sub-platforms” are able to take countermeasures with the help of Core Platform 2. The most famous platform blocking dispute in the history of China's platform development was Tencent's blocking of Alipay red packets in WeChat on the occasion of the Spring Festival in 2015. This behavior can be interpreted as Tencent's ecosystem's self-preferential treatment in support of WeChat's red envelopes. However, based on its own economic and technological advantages, the Ali ecosystem launched the “Alipay password” overnight, and succeeded in breaking through in a short period of time. At this time, if the law enforcement agencies in accordance with the antitrust law to punish Tencent's behavior, on the one hand, in the innovation-intensive digital economy, there is the problem of lagging effect of law enforcement, on the other hand, can not stimulate the market players to find their own solutions to the motivation. In addition, the fact that Ali's ecosystem was able to break through the Tencent ecosystem's ban overnight through the “Alipay password” not only proves that the competition between the ecosystems is still in a benign state, but also means that the intervention of the antitrust law is not necessary.

Secondly, the role of “core platform 2” may limit the development of small ecosystems. In 2017, shunfeng and caijiao closed each other's data interface, for example, although shunfeng does not have a dominant position in the express delivery industry, but is China's largest logistics platform. Although the rookie is a small logistics platform, it is subordinate to the Ali ecosystem, which is much larger than that of SF. If the competition law enforcement agency looks at SF's blocking behavior in a fragmented way and determines that it constitutes a refusal to deal under Article 22 of the Anti-Monopoly Law, SF must open up its data to Cainiao. And as a small platform in the express delivery industry, Cainiao is not required to open its data to SF. In this way, the Ali ecosystem can obtain all of Shunfeng's data via Raijiao, and may be able to utilize the advantages of inter-industry operation to exclude or limit Shunfeng's competition in the ensuing competition. In fact, the rookie dares to challenge the scale is much larger than its own SF, objectively also shows that its competition in the full use of the Ali ecosystem market position. Therefore, the platform blocking behavior between Shunfeng and Caijiao is commercially reasonable, and the rigid use of antitrust law to deal with it is not conducive to competition in the relevant market, and will not really solve the problem. Subsequent facts proved that, although SF and Caijiao continued to cooperate on the surface under the coordination of the State Post Bureau afterwards, they did not realize full interconnection in practice.

4.3 Monopolistic behavior in the competition within the ecosphere

After entering the form of Internet ecosystem, platforms began to operate their own business and compete with merchants, and the relationship between platforms and merchants began to be alienated. At this time, the main concern is that the platform may use the market organization function to strengthen its market profit function by giving preferential treatment to self-operated business. Thus, unlike inter-ecosystem competition, which focuses on the competitive role and profitability function of the core platform, intra-ecosystem competition involves how the market organization function of the core platform is perceived.

Figure 2 Competitive transformation within the ecosystem

Self-preferential treatment in intra-ecosystem competition is superficially similar to self-preferential treatment in inter-ecosystem competition: competing sub-platforms and merchants are on an equal footing and are both classified as intra-platform operators, but there are fundamental differences between the two: “Competitive sub-platforms are in the same Internet ecosystem as Core Platform 2 and are protected by the latter, whereas “merchants” are independent market players that lack the protection of the ecosystem. The author believes that antitrust can be used as a means of protection. In the author's view, the concept of control in antitrust law can be used to distinguish between “competing sub-platforms” and “merchants”. Operators on platforms controlled by other Internet ecosystems belong to “competitive sub-platforms”, while operators on platforms not controlled by other Internet ecosystems belong to “merchants”. The main purpose of distinguishing between the two is that “competing sub-platforms” can engage in antagonistic behaviour under the protection of the Internet ecosystem, whereas independent “merchants” are subject to the market organization function of the platform, making it difficult for them to engage in direct antagonism with the platform.

In order to promote the development of self-owned businesses, platforms may give preferential treatment to self-owned businesses in two ways: one is to increase the exposure of self-owned businesses on the homepage, to improve the search rankings of self-owned businesses, and to lower the search rankings of merchants' businesses, etc. These behaviors are explicitly self-preferential. These behaviors are explicit self-preferential behaviors. The second is to collect and analyze merchant data to improve the competitive advantage of self-owned business. This behavior is similar to self-preferential treatment, but there are differences. Self-preferential treatment means that the platform provides more favorable services to its own business, while the platform may not directly involve the provision of favorable services to its own business when it uses merchant data to compete with it. Of course, even if they do not offer preferential services to their own businesses, platforms can still analyze merchant data to gain an advantage over their competition. These behaviors are implicitly self-preferential. They are explicitly regulated in the EU by Article 5 of the Digital Marketplace Act. However, implicit self-preferential treatment has not yet been included in the scope of China's antitrust law.

Even for explicit self-preferential treatment, whether the antitrust law can regulate it is full of controversy. The differential treatment rule in the antitrust law only prohibits the monopolist's discriminatory behavior among third-party trading entities, while self-preferential treatment involves the behavior of treating oneself better and “treating” third-party trading entities worse. In principle, market mechanisms do not prohibit operators from giving themselves trading conditions that are superior to those of others. Therefore, the rule of differential treatment is hardly a legal basis for regulating such self-preferential behavior. Another possible basis for regulation in antitrust law is the essential facilities doctrine, which requires a monopolist to deal with a third party when an asset constitutes an essential facility. However, the conditions for determining essential facilities are extremely stringent, requiring that the platform in question must lack both actual and potential “substitutability”. To date, there has been no case in China involving a platform constituting an essential facility. In the first international self-preferential treatment case, the Google Comparison Shopping case, the EU did not recognize Google's search engine as an essential facility. EU academics have strongly criticized this approach of determining the illegality of self-preferential treatment in the absence of a finding of essential facilities, arguing that it significantly exceeds the scope of the traditional antitrust laws. Self-preferential treatment involves the platform's use of market organization functions to strengthen the profit-making function of the market, which is essentially a problem arising from the unequal trading status between the platform and the merchant. However, the goal of antitrust enforcement is to promote economic efficiency, not to protect vulnerable groups. Therefore, the issue of self-preferential treatment in intra-ecosystem competition cannot be effectively addressed in antitrust law.

Predatory acquisition is a special kind of monopolistic behavior that exists in the competition within the ecosystem, which refers to the process of platforms building the Internet ecosystem, “acquiring startups with competitive potential in the market and preventing them from being acquired by other large platforms or growing into competitors that are not controlled by them “. These strangled startups lack the protection of the Internet ecosystem, so such mergers and acquisitions do not fall into the category of inter-ecosystem competition, which this paper categorizes as monopolistic behavior in intra-ecosystem competition. The creation of the ecosphere can not only produce the attack effect of entering the business of other platforms, but also achieve the defensive effect of protecting the main business. Stifling mergers and acquisitions are the embodiment of this defensive effect. Competition in the digital economy focuses on “killer apps”, or breakthrough innovations. Breakthrough innovations are costly and risky. For the top platforms, one of the effective ways to achieve a defensive effect is to nip in the bud breakthrough innovations that could impact their core business. Stifling mergers and acquisitions not only limit competition from other platforms, but also have a negative impact on startup innovation. While the competitive harm of stifling M&As is not difficult to understand, there are implementation-level difficulties. The difficulty lies in the fact that stifling mergers do not have an obvious exclusionary or restrictive effect on competition at the time of their occurrence, but tend to raise concerns only after a certain period of time, especially when startups achieve commercial success after being merged. By that time, however, it is too late to prohibit them under the antitrust laws. As a result, the problem of stifling mergers and acquisitions is also difficult to address effectively in the antitrust laws.

Although the Internet ecosystem form reduces the operating cost of the platform, its business aggregation model triggers fierce competition among heterogeneous business platforms. This mode of competition is not like the “competition for the market” in the Web 2.0 form, but is closer to the “competition within the market” in the traditional economy. In this competition mode, the Internet ecosystem platform began to “fight” with other platforms with vested interests, triggering more inter-ecosystem monopoly behavior. At this time, the antitrust law, which aims at regulating competition, is not invalid, but only needs to be adjusted based on the new situation. However, the monopoly behavior within the Internet ecosystem involves both competition and transaction relations, and the handling of antitrust law has begun to highlight the inadequacy.

5. The institutional framework of antitrust law and regulatory dichotomy

From 2009 to now, China's academic research on platform antitrust has been more than ten years, and basically formed a general theory of dualistic governance of antitrust law and regulation to adjust the monopolistic behavior of platforms. At present, the main controversy is how to construct the institutional framework of the dualistic division of antitrust law and regulation. The previous study on the correlation between the evolution of platform forms and the types of monopolistic behaviors can provide a preliminary answer.

5.1 The Establishment of Antitrust Law and Regulatory Dichotomy

The digital economy started as a storage intermediary platform in the form of Web 1.0, evolved into an information interaction platform in the form of Web 2.0, and then evolved into a business aggregation platform in the form of the current Internet ecosystem. From the perspective of evolutionary theory, these three stages are not the convergent evolution of mutual substitution, but the compound evolution of forward development. In this process, the digital economy of the later generation always develops new business models on top of the previous generation. Although China's academic research has paid attention to the monopoly problem in the digital economy, the lack of evolutionary perspective makes the research results of different phases fractured and has not yet formed a synergy.

The type of platform monopoly behavior is closely related to the evolution of the shape of the digital economy. The aforementioned studies show that the morphological evolution of the digital economy has brought about two expansions of platform monopoly behavior as well as two-fold alienation. The two expansions involve the scope of platform-related markets: the first one is the Web 2.0 form which expands the unilateral market in the traditional economy into a bilateral market, and the second one is the Internet ecosystem form which expands the single-type platform business operation in the Web 2.0 form into an aggregated platform business operation. However, these two expansions have not exceeded the scope of adjustment of the antitrust law, and have only brought about challenges within the system. The two-fold alienation occurs mainly in the form of Internet ecosystem and is closely related to the expansion of platform functions. The first alienation occurs between platforms and non-competitive platforms, but the antitrust law can still effectively deal with it, only that it needs to incorporate the new features of the Internet ecosystem pattern into the enforcement mechanism. The second alienation occurs between platforms and merchants and cannot be effectively addressed by antitrust laws.

The expansion and alienation of platform monopoly behavior not only affects the competition pattern in the digital economy, but also is continuously affecting the research, enforcement and legislation of platform antitrust in China and even around the world. As far as research is concerned, China's academia has recognized that “strengthening antitrust” in the digital economy is not only about strengthening the enforcement of antitrust law, but also about the need for a dualistic division of antitrust law and regulation. The previous analysis shows that the traditional antitrust mechanism is indeed powerless in the face of heterogeneous monopoly behavior in intra-ecosystem competition, and there is an urgent need to introduce a regulatory mechanism for joint law enforcement. Before 2020, China's law enforcement agencies will hold an “inclusive and prudent” regulatory attitude towards the digital economy, and the governance of platform monopoly behavior will be led by the judiciary at this time. Since Web 2.0 platforms only brought about the expansion of the scale of the digital economy, and had not yet triggered the alienation of monopoly behavior, there were no judicial cases in which platforms were found to be in violation of the antitrust law. 2020 end of the central government put forward the requirement of “strengthening antitrust and preventing the uncontrolled expansion of capital,” and since then, it has entered a new stage of platform governance dominated by the law enforcement agencies. Since then, it has entered a new phase of platform governance led by law enforcement agencies. Among these cases, the platform monopoly behaviors that were penalized were mainly exclusive deals. However, exclusive trading is only a monopolistic behavior in Web 2.0. For the monopoly disputes in the Internet ecosystem that have emerged in practice, China's law enforcement and judicial agencies have not yet launched substantive actions. Perhaps also realizing that there are platform monopoly behaviors that cannot be effectively adjusted by the antitrust law, the State Administration for Market Supervision and Administration (SAMSA) released the “Opinion Draft of Guidelines for Classification and Rating of Platforms” and the “Opinion Draft of Guidelines for Responsibility of Platforms” at the end of 2021. Although these two guidelines have not yet been promulgated, they formalize China's legislative and law enforcement concept of implementing the dualistic division of antitrust law and regulation in the digital economy.

Around the world, the regulation of the digital economy has also experienced a change in attitude from “inclusive and prudent” to “strengthened antitrust” to the dichotomy of antitrust law and regulation. Until about 2016, there were few cases of platform monopolization worldwide. Since 2016, the EU has been on a platform antitrust kick, and has since used antitrust law to impose large penalties on platform companies on a number of occasions. However, as it entered 2020, the EU began to realize the shortcomings of its antitrust regime and introduced draft legislation for a Digital Marketplace Act, which was enacted in 2022 and has become representative legislation for platform regulation today. The U.S., another important antitrust jurisdiction, is also actively promoting a dualistic approach to antitrust law and regulation. For example, the appointment of Lena Khan, a leading figure of the New Brandeis School (the school of positive regulation), as the Chair of the U.S. Federal Trade Commission, and the introduction of five pieces of draft legislation on platform regulation by the U.S. House of Representatives in 2021, modeled on the EU's Digital Marketplace Act. Even the Organization for Economic Cooperation and Development (OECD) has continued to call for countries to back up their antitrust laws with regulation. This shows that a consensus has been formed on the implementation of the binary division of antitrust law and regulation in the digital economy.

5.2 Effectiveness of antitrust law

Although both antitrust law and regulation can realize the adjustment of monopoly behavior, the antitrust mechanism is based on the law of market economy, and the intervention in economic activities implies the self-restraint and modest tolerance of the power institutions. Especially in the digital economy, which contains great uncertainties and unknown risks, antitrust law can best reflect the basic requirements of the principles of the rule of law and science. In contrast, regulation is more destructive to the market mechanism. Once there is a regulatory failure, it will cause damage to the relevant economic field for a longer period of time. Therefore, the general view is that antitrust law should be preferred when adjusting monopoly power, and regulation should be resorted to only when there is structural failure. In determining the scope of their respective adjustments, it should first be clarified what the antitrust regime can and cannot do, and then the complementary adjusting function of regulation can be explored in areas where the antitrust regime cannot.

In the process of dominating platform antitrust, China's law enforcement agencies have deviated from the traditional antitrust law analysis method, and even showed a tendency to go beyond the antitrust law to conduct antitrust enforcement. For example, in the Ali case, the law enforcement agencies pointed out that the key to analyzing competition damage was to examine whether the conduct in question “creates a lock-in effect, reduces its own competitive pressure, and improperly consolidates and strengthens its own market power”. Unfortunately, however, the enforcement agency did not conduct a quantitative analysis of the market-closing effect of the conduct in question.The administrative penalty decision in the case showed that Taobao's market share declined by a total of 15% instead of increasing during the five years in which the exclusive dealing was implemented. This means that the market share of Taobao's competitors increased by 15% during the period in which the monopolistic behavior was implemented, and competition in the relevant market was not blocked. Although the competitor's market share might have been higher in the absence of the monopolistic conduct at issue, it is difficult to conclude that the relevant market might have been blocked from competition in the medium to long term without a counterfactual analysis. As noted earlier, China's practical and theoretical communities have recognized that exclusive dealing must be enforced in accordance with the principle of reasonableness, but the analysis in this case borders on treating it as a per se violation of the law. “The China Knowledge case also lacked sufficient reasoning in its analysis of ultra-high pricing. Strengthening antitrust enforcement is not to go beyond antitrust enforcement; if it were, it would lose the basic qualities of the antitrust regime, which are modest and inclusive, and would instead be suspected of being a form of regulation in the name of antitrust law.

Market development has shown that this kind of law enforcement does not play a role in ensuring the healthy and sustainable development of the digital economy. From the perspective of law enforcement effects, China's recent platform antitrust cases have led to three aberrant law enforcement phenomena: first, antitrust penalties have not weakened the monopoly position of the head platform. To this day, all the penalized platforms, such as Ali, Meituan, China Knowledge, and Food Pace, continue to maintain their monopoly position in the relevant markets. Second, the deterrent effect of antitrust penalties is insufficient. China's history has repeatedly seen the phenomenon of share prices rising rather than falling after platforms have been penalized. For example, although the “Ali case” set a record of 18.228 billion yuan in administrative fines, Ali's stock price rose about 6% on the day the penalty decision was issued, with profits far exceeding the amount of the fine. It seems that the antitrust penalty is not a negative evaluation, but a positive endorsement of the platform's investment prospects. Third, there is a problem of overly strong deterrent effect of antitrust investigations. Immediately after China's platform enterprises were initiated with antitrust investigations, their stock prices began to fall back continuously; in comparison, the impact of antitrust investigations in Europe and the United States on the stock prices of the relevant platform enterprises was minimal. The reason behind this may be that this enforcement model is more likely to result in unlawful penalties that are difficult to overturn, while antitrust investigations in Europe and the United States are more uncertain. This enforcement effect seriously deviates from the original purpose of platform antitrust enforcement.

Other countries or regions that have initiated antitrust enforcement against platforms lack international head platforms. The EU is the jurisdiction with the most resolute platform antitrust enforcement, but the EU's largest platform, Spotify, is not even among the top 20 global platform companies. Currently, all EU platform antitrust targets are U.S. platforms, and no local EU platforms have been included in the scope of enforcement. And as the most developed digital economy in the United States, there has not even been a single final monopoly case in which a platform was found to have violated the law. China must remain vigilant against this kind of law enforcement posture that deviates from or even exceeds the antitrust law. In China, the digital economy not only bears the responsibility of deepening innovation and development, but also shoulders the important task of enhancing foreign trade in the context of the trade friction between China and the United States. Continuing enforcement beyond the antitrust law is not the optimal choice to promote the development of China's digital economy. Although the evolution of the shape of the digital economy has caused many challenges to the implementation of the antitrust law, it is only a problem to be dealt with within the antitrust law system, and does not require a breakthrough in the analytical paradigm of the traditional antitrust law.

5.3 Complementary Function of Regulation

The complementary function of regulation to the antitrust law is mainly manifested in that the scope of regulation should be limited to the behaviors that cannot be dealt with by the antitrust law, i.e., all kinds of monopoly disputes presented by the competition within the ecosystem. The “Draft Guidelines on Classification and Rating of Platforms” and the “Draft Guidelines on Platform Liability” provide a basis for the introduction of regulation, but there are problems that do not fully take into account the effectiveness of the antitrust law and the characteristics of the development of the Internet ecosystem, and need to be adjusted.

5.3.1 Clarify the purpose of platform classification and grading

The purpose of platform classification is to delineate the platform business to be regulated. The competition model in the form of Internet ecosphere indicates that the object of platform governance should be located in the core platform business rather than the non-core platform business which is difficult to form an aggregation effect. From the perspective of the development of China's digital economy, core platform businesses may include comprehensive e-commerce, online search engines, social networks, video sharing, instant messaging, operating systems, application stores, third-party payments, cloud computing, and so on. In view of the fact that the Internet ecosystem is still in the process of development and evolution, the list can be kept open and added or removed as the core platform business develops and integrates.

The purpose of the platform hierarchy is to determine which platform enterprises need to be regulated, which must be set in accordance with the inadequacy of the antitrust regime. The previous analysis shows that monopoly disputes that go beyond the traditional antitrust mechanism mainly occur in the competition between platforms and merchants. According to the different forms of platforms, this can be divided into two specific situations: (1) disputes between platforms and merchants in the form of Web 2.0. In principle, disputes between Web 2.0 platforms and merchants belong to the scope of platform self-regulation and generally do not need to be regulated. However, with the reshaping of the social structure of the digital economy, platforms (including Internet ecosystem platforms) have gained the power traditionally controlled by the government through digital technology, such as the rules of the platforms have begun to replace the government in enterprise qualification and responsibility supervision. Therefore, it is necessary to regulate the self-regulation of platforms in a certain form to prevent the uncontrolled expansion of capital. According to the new regulatory theory, in addition to government regulation and self-regulation, there exists a third state of regulation, namely meta-regulation. Meta-regulation is to realize the regulation of self-regulation by the external guidance of the government. Scholars generally agree that meta-regulation is especially needed when “the problem to be solved is too complex, or the industry to be regulated is very special and in dynamic evolution”. The self-regulation of platforms belongs to this situation and can be regulated by meta-regulation. (2) Disputes between platforms and merchants in the form of Internet ecosystem. The platform's self-regulation in the form of Internet ecosphere triggers alienated monopoly behavior, and the behavior cannot be adjusted by the antitrust law. However, the need to introduce regulatory legislation has not yet been agreed upon by the academic community in China. Some scholars suggest that the theory of publicity can be invoked for regulation, but others point out that the theory has serious logical loopholes. Despite the theoretical controversy, failure to regulate platform businesses with structural market failures will leave these areas in a state of monopoly for a long time.

Based on the two different regulatory scenarios mentioned above, the author suggests setting two levels of large platforms and super platforms for categorized regulation. The purpose of setting up large platforms is to cope with disputes between platforms and merchants in the Web 2.0 form, and to guarantee that the regulator can supervise the rule-making process within the platform. The classification criteria for large platforms can be referred to Article 3.2 of the current “Platform Classification Guidelines”, i.e., having a large user scale (no less than 50 million active users in China in the previous year) and a high economic volume (market capitalization or valuation of no less than RMB 100 billion at the end of the previous year). The purpose of setting up a super-platform is to deal with disputes between platforms and merchants in the form of Internet ecosystems. Superplatforms should be set up with great care, so that even if there is alienated platform monopoly, merchants can effectively protect their interests by switching to a competing platform when there is sufficient competition. Only when there is a serious lack of competition, the alienated platform monopoly behavior can not really be adjusted by the market mechanism. Therefore, in addition to a few areas such as search engines where competition is seriously lacking, there is no need to designate a super-platform in most areas where there are two or more competing platforms. In addition, in view of the fact that there are only a small number of areas where competition is seriously lacking, there may not be quantitative criteria for super-platforms, but only the above limitations.

5.3.2 Regulatory measures for large platforms

The regulation of large platforms is mainly limited to self-regulation, which is based on the formulation of rules within the platform. This can be achieved in three ways: first, by requiring large platforms to establish an on-platform rule-making mechanism. While regulators do not actively control the content of these rules, large platforms could be required to establish an on-platform rule-making process first. The rule-making process can be modeled on the national legislative process, follow the principle of openness and transparency, and involve a wide range of stakeholders such as merchants, users, industry associations, experts and scholars, and then be formed after sufficient deliberation. Second, establish an external supervision mechanism for the rules within the platform. In order to prevent large platforms from neglecting to formulate rules on specific issues, regulators can require them to report their planning for the formulation of intra-platform rules on a regular basis, or they can be required to formulate specific rules at any time. Third, establish an external review mechanism for intra-platform rules. Disputes between platforms and merchants should be resolved by the large platforms themselves, and the regulator should be positioned as an external reviewer of the in-platform rules, responsible for dealing with disputes arising from the reasonableness of the in-platform rules, and empowered to review whether the in-platform rules that are in dispute are fair and reasonable, and whether they will promote the development of the digital economy.

There is also a need to impose disclosure obligations on large platforms. Currently all the behavior of platforms is implicitly manipulated by algorithms, which is difficult to detect by other market players. The disclosure obligation is the mildest regulatory measure, with a low level of intervention in the platform's operations. However, the disclosure obligation allows merchants to understand the operation of the platform and make rational choices based on it. Given that directly requiring large platforms to disclose all the contents of their algorithms may involve infringement of the platform's trade secrets, they can be required to disclose the main principles of the relevant algorithms and provide risk warnings to users.

5.3.3 Regulatory measures for mega-platforms

The lack of an effective competition mechanism means that the market is unable to play a decisive role in resource allocation, and therefore special obligations should be imposed on superplatforms that go beyond those of other platforms. The market organization function of platforms makes them go beyond the scope of traditional enterprises and become a third sector that undertakes part of the government's functions. Although traditional theory requires that the third sector must assume non-profit obligations when performing social adjustment functions, platforms operating their own businesses will not only enhance total social welfare, but also consumer welfare. Therefore, super-platforms should not be prohibited from engaging in self-operated business altogether, but they can be required to treat other merchants equally and not to practice self-preferential treatment when operating self-operated business. Given the secretive nature of the implementation of self-preferential treatment, super-platforms should also be subject to the same algorithmic disclosure obligations as large platforms.

To address the issue of stifling mergers and acquisitions, some scholars have suggested adjusting the reporting threshold for operator concentration to include it in the scope of antitrust scrutiny. The difficulty with stifling mergers and acquisitions is that, on the one hand, we cannot completely prohibit platforms from merging with startup platforms; on the other hand, the competitive harm of the behavior is difficult to detect at the time of the merger or acquisition. Therefore, adjusting the notification threshold only expands the scope of scrutiny by the enforcement agency, but still fails to address the issue of entity characterization of stifling mergers and acquisitions. In this regard, Nobel Prize-winning economist Tyrrell has proposed a relative compromise by reversing the burden of proof on the effect of exclusion or restriction of competition to the merging parties. However, this is too high a burden of proof for the merging firms, and the harm of stifling mergers will only be more pronounced in the absence of a serious lack of competition, so it is sufficient to impose relevant requirements on superplatforms only.