垄断法下的相关数据市场研究

曹阳

Abstract： Data is the profit center， new oil and key driving force of the digital economy. Currently data is almost out of the analysis scope of antitrust reviewing of the data driven economy. Data should be revalued in antitrust analysis. Digital platforms are among the most influential of digital actors in helping to determine the structure of online activity. The disruptive power of data driven platform is revolutionizing business， economics and society. Different from the traditional pipeline business model， the data driven platform market is multi-sided and interdependent. The pursuit of scale means that the platform must make every effort to obtain data resources. Data is the core asset in moderating different sides and the commodity the data driven platform markets. Data's power is manifested in its ability to make profits and enable business model innovation. Power of the data is also reflected in the control of user privacy and manipulation of users' behaviors and thoughts. Data power can bring negative effects on competitions. The damages to competition in relevant data markets by data driven platforms include entry barrier， privacy invasion and detriments to consumer interests. A new thinking is needed to contain the damages of data power. In defining market power of data， the market share of data， the ability to collect， process and commercialize data need to be considered and data market should be treated as a whole in antitrust analysis.

Key words： data driven platform; data power; data market; antitrust behaviors

CLC： D 913                     DC：A                        Article ID：2096-9783（2021）01-0111-16

1 Introduction

Should data be a key factor in antitrust review in the digital economy？ Traditional antitrust review rarely pays attention to competition damages in the data field. Some scholars have conducted research on some major anti-monopoly agencies in Europe and the United States in reviewing data-related antitrust cases， and found that there is almost no authority to analyze market power in the data field[1]. The same is true in China. China's antitrust authority seldom mentions data power in data-related antitrust cases. Recently this trend is changing. On 26 June 2019， the newly established State Administration of Market Regulation of China issued Interim Rules on Prohibition against Abuse of Dominant Market Position （the "DMP Rules"）. Article 11 of the DMP Rules requires the relevant authority to consider the data situation in assessing the dominant position. It is the first time that an official document requires data factors to be considered in the antitrust review process. On 7 February 2019， The German Federal Cartel Office ruled that Facebook abuses its dominant position in the German market for social networks by collecting and aggregating user data. The regulators argue that Facebook uses the data to strengthen its dominant position by increasing its attractiveness to advertisers and impeding competitors which don't have such huge data troves[2].

Data driven platforms （DDPs） have driven up productivity in multiple ways[3]  and play a central role in the digital ecosystem[4]. DDPs can therefore be among the most influential of digital actors in helping to determine the structure of online activity[5]. The disruptive power of the DDPs is revolutionizing business， economics and society. In this article we will first analyze the importance of data in DDPs. We will elaborate on the reason why data becomes the core asset in digital platforms. In part III， we will analyze power the data can bestow upon its owner and how this power shapes the economic pattern of digital economy. In part VI the negative effect of data is analyzed. In this part we will elaborate on how data misuse can terribly hurt the economy and society. In part V， the reason why data misuse is uncontained is analyzed. In this part we will talk about the incapability of current antitrust analysis framework to DDPs and a new paradigm is needed. In this part we propose we need to redefine the relevant market of data related economy. In part VI， we will explore the current data related activities and propose some new ways to analyze market power of data and contain market power of DDPs. Finally， we will use a case to demonstrate how our newly established paradigm works.

2 Data is a Core Asset

In the traditional one-sided market， market operators only need to know their customers' preferences and try to satisfy those kinds of needs. The data is not a key drive for this kind of linear economy and all data collected are served for marketing a specific kind of product or service. But in the DDP economy， flow of data through platform is crucial to maximizing growth and has been recognized for years as a critical driver of economic growth and productivity[6].

The operation of the DDP is conditioned on the so-called same-side and cross-side effects. Platform operators generally try to achieve as many distinct groups of users as possible for each side of the platform to achieve those kinds of network effects. Only scale can produce so-called "multi-sided effects" and "indirect network effects". Therefore， the pursuit of user scale is the basis for the development of all platforms. Only the scale can generate big data that the platform can explore. Only the scale can bring the "multi-sided effect" into reality. The pursuit of scale means that the platform must make every effort to obtain data resources. Big data is the biggest name-changing opportunity for marketing and sales since the Internet went mainstream almost 20 years ago. The data big bang has unleashed torrents of terabytes about everything from customer behaviors to weather patterns to demographic consumer shifts in emerging markets[7]. Having control over， and being able to quickly analyze the big data can provide the platform a key competitive advantage.

The expansionary nature of these platforms means that firms that were operating in completely different areas are now converging together under the pressures of competitively extracting data[8]. Today's DDP， such as social platforms and search platforms， provide free services to certain types of users to collect the data they desire. For example， Google's model is to collect relevant customer's data for advertising purpose. Google can deliver customers more likely to purchase an advertiser's product and， as importantly， help sell those products at the highest price the user may be willing to pay[9]. The core source of value being delivered to advertisers by Google （or any search advertiser） is， by all accounts， its intimate knowledge of its users contained in its vast databases of user personal data[10]. When countries placed restrictions on behavioral data collection， as in parts of Europe， studies found that advertising effectiveness dropped drastically， indicating the critical importance of user data to online advertising[11].

Platforms become dominant not because of what they own but rather because of the value they create by connecting their users[12]. In the DDPs， we can witness several different value units. Some of these value units derive from same-sided interactions， and some from cross-sided interactions. The DDPs is centered on pursuing the growth of these different value units. Different value units in essence are different types of data with various values. Taking social networking platform as an example， the platform providers will integrate and process the data information provided by end users， sell them to the corresponding data sellers. The sellers use these data information for advertisements. In addition， the platform providers also recommend new users to the existing users by exploring the collected data. The scale of the users is thus continuously expanding， thereby generating more data and the platform gaining more advertising revenue. Based on rich data information in the platform， platform operators can easily extend its business to related areas， which may create a relatively independent digital business ecosystem for themselves. Social media platforms promise to connect users person-to-person， entrusted with messages to be delivered to a select audience. But as a part of their service， these platforms not only host that content， they organize it， make it searchable， and in some cases even algorithmically select some subset of it to deliver as front-page offerings， news feeds， subscribed channels， or personalized recommendations. In a way， those choices are the central commodity platforms sell， meant to draw users in and keep them on the platform， in exchange for advertising and personal data[13].

Platforms constantly make moderation decisions， and the very nature of those platforms is itself a kind of moderation[14]. The platform fosters the flow of value by making connections between producers and consumers and data is at the heart of successful matchmaking and distinguishes platforms from other business models. The platform captures rich data about the participants and leverages that data to facilitate connections between producers and consumers[15]. So data becomes the commodity the DDPs marketed in the two-sided markets. Data is the profit center and new oil of the two-sided digital economy[16]. And those data is the core asset and the basis for the operation and the new currency of the DDPs[17].

3 Data is Power

One author points out that in online markets， the competitive harms that could arise from large firms' access to extensive user data usually exist only in the realm of theory[18]. But the truth is that data is the profit center and new oil of the two-sided digital economy[19]  and changes the rules for markets[20].  Data is connecting point for all sides and also provides value-enhanced resources for all participants. Personal data has become the most prized commodity of the digital age， traded on a vast scale by some of the most powerful companies in Silicon Valley and beyond[21]. With advanced analytics and new data sources， companies in one sector can play a role in the products and services of others， even those far removed from their traditional line of business. This blurs the boundaries between industries and changes competitive dynamics[22]. Soaring flows of data and information now generate more economic value than the global goods trade[23]. As a content and community hub， the DDP will collect data that will give them leverage & power in the marketplace. The platform that controls the data controls the market choices， business opportunities and development models for online business. The data is the fuel that makes businesses smarter and more profitable[24].

Data's power is first manifested in data's ability to make profits. Through data， the DDP can well understand the preferences and conditions of end consumers and send accurate advertisements to targeted customers so as to obtain huge benefits. As a multi-sided market， the DDP has an inherent strong desire to explore its data in the neighboring business areas. Data are kinds of assets that can be explored in non-rival and non-exclusive ways and can be easily used in other areas in a very low marginal cost. Data as a kind of power decide which areas the DDPs can expand to. In other words， the control of the DDP in the data market determines its ability to make profits， expand to adjacent areas， the form and value of its services.

Data also enable business model innovation. A business model， in essence， is a representation of how a business creates and delivers value for a customer while also capturing value for itself， doing so in a repeatable way[25].A business model framework should be based on four interdependent elements： customer value proposition， profit formula， key resources and key processes. Data can contribute in separate ways to a value proposition. Essentially， there are two main directions： data can add value to a key resource， or they can form the key resource itself[26]. With advanced analytics and new data sources， companies in one sector can play a role in the products and services of others， even those far removed from their traditional line of business. This changes competitive dynamics. Companies that transform their business models in parallel with these shifts will find new opportunities for revenue streams， customers， products， and services[27].

Data do not only enable strategies， they become the strategy[28]. Data are valuable if they are used to apply a novel dimension to a prevailing business practice or enable new business types sui generis. The value dimensions can be summarized as product or service （what is offered？）， business processes （how is it offered？） and business model （how is it monetized？）. Business transformation usually entails all dimensions， yet they can also be addressed independently[29]. Through data exploitation， the DDP can easily sell more its current offerings to existing clients; provide new offerings to existing users， or new offerings to new users. The platform operators can also easily provide added functionality to existing services， or combined offerings categories. Of course， the data per se are also kinds of assets that can be commercialized through provision or brokerage. Through data analysis， the DDP can adopt a dynamic price policy to harvest the possible profits.

Power of the data is also reflected in the control of user privacy and manipulation of users' behaviors and thoughts. Through data processing and analysis， the DDP can easily obtain the user's personal privacy information and profile the users. The DDP can easily get information of the user's health status， shopping habits， and financial status through location information and browsing habits. Through data profiling， highly sensitive details can be inferred or predicted from seemingly uninteresting data， leading to detailed and comprehensive profiles that may or may not be accurate or fair. Increasingly， profiles are being used to make or inform consequential decisions， from credit scoring， to hiring， policing and national security[30]. For example， Chinese government is working with Alibaba and Tencent to build a Social Credit System that will rank citizens' and businesses' reputations based on their purchases， movements， and public communications while using that ranking to restrict access to jobs， travel， and credit. By controlling the user's private information， the DDP can commercialize this information to gain a competitive advantage and manipulate users' choices. This is the reason that so much fraudulent data appear on the DDP. And the fraud can bring benefits for the platform[31]. Further， to the extent that such firms compile politically sensitive information about users， and mediate their experience of content， they are also powerful political actors[32].

4 Data Misuse is Harmful

Data enables the digital economy. But abuse of data power is sure to harm the consumer's welfares. Specifically， the market power of DDP in the data market will create barriers to entry， compromise consumer privacy， and harm the competitive order.

4.1 Entry Barriers

The market power in the data market will trigger so-called entry barriers. As we mentioned before， the network effect of the DDP will trigger the user's path dependency and prevent users from moving away from the platform. With the increase in the market share， the barriers to entry will increasingly rise. In addition， there is a so-called first-mover advantage in the DDP. This first-mover advantage is magnified by the existence of network effect， which makes it more difficult for the follow-up platform to enter the relevant competitive field. This is why more and more giant DDP s are emerging today.

Another consideration is so-called data partnership arrangement. As the above-mentioned evidence points out that it seems to be the norm for platform giants to have some data cooperation programs. For example， Facebook weaves its services into other sites and platforms， believing it would stave off obsolescence and insulate itself from competition. Every corporate partner that integrated Facebook data into its online products helped drive the platform's expansion， bringing in new users， spurring them to spend more time on Facebook and driving up advertising revenue. At the same time， Facebook got critical data back from its partners[33]. This kind of tacit coordination creates a kind of entry barrier that benefits only the stakeholders in the digital ecosystem at the cost of potential new entrants. As one U.S. court observed， "Tacit coordination is feared by antitrust policy even more than express collusion， for tacit coordination， even when observed， cannot easily be controlled directly by the antitrust laws. It is a central object of merger policy to obstruct the creation or reinforcement by merger of such oligopolistic market structures in which tacit coordination can occur[34]".  And platforms may use technology and algorithms to support traditional forms of collusion-that is collusion agreed between humans and executed with the assistance of technology[35].

4.2 Damages to Privacy and Consumer Interests

When the DDP gains strong market power in the data market， it will further ignore user privacy and further strengthen its data exploring capabilities. Privacy seems not a much concern for DDP owners. For years， apps and websites have casually harvested personal information for murky ends[36]. Zuckerberg deeply believes that the records of our interests， opinions， desires， and interactions with others should be shared as widely as possible so that companies like Facebook can make our lives better for us-even without our knowledge or permission[37]. Baidu's founder Robin Li even said that Chinese people are willing to exchange convenience for privacy. But this so-called convenience has also caused damages to consumers. Through the control of the data market， the DDP has acquired the ability to send accurate advertisements to consumers. In recent years， the Internet advertising index has been on an upward trend， reaching an all-time high in 2018[38]. Consumers will finally pay the huge investment in the field of digital platform advertising. The consequence is that consumers will pay higher prices for the service， which will undoubtedly harm consumer welfare.

Data power is also manifested in differentiated treatment of consumers through data discrimination. In DDP customer discrimination strategy is increasingly adopted. The essence of the "Killing Cooked" behavior in the DDP is that platforms use the data market power to treat consumers differently. And the customers can hardly notice this kind of discrimination due to surreptitiousness of this strategy.

Data market power also makes predatory pricing possible. Predatory pricing strategy in DDPs includes ultra-low pricing and ultra-high pricing. Ultra-low pricing seems to be good for consumers， but there is no such thing as a free meal. Through so-called ultra-low-cost service the DDP acquires massive data resources， which become the core resources for DDPs to obtain excess profits in the future. The excess profit is ultimately borne by the consumers， and the DDPs become the ultimate beneficiary.

5 Data is Uncontrolled

Typically， market dominance analysis starts with defining the relevant market in which to assess the anticompetitive effects. A relevant product market comprises all those products and/or services which are regarded as interchangeable or substitutable by the consumer by reason of the products' characteristics， their prices and their intended use[40]. Traditionally， determination of the market is based on economic analyses of the flexibility of supply and demand， as well as on market research[41]. The Chicago School approach to antitrust gained mainstream prominence and credibility in the 1970s and 1980s and now becomes the mainstream theory for antitrust policy in various countries including China. The essence of the Chicago School position is that "the proper lens for viewing antitrust problems is price theory".

According to the "price theory" paradigm， the relevant market includes the market involving the product itself and its substitutes. Market definition focuses mainly on demand substitution factors[43]  and most commonly based on the "hypothetical monopoly" test， also known as the SSNIP test[44]. The objective of this exercise is to define the smallest possible markets both in the product and geographic dimension， whereby a hypothetical monopolist could profitably and permanently raise the price of the products by 5 to 10 per cent above the competitive level[45]. SSNIP test is essentially a price test method[46].Defining the relevant market using the above-mentioned "price theory" paradigm will cause serious problems under the data driven economy situation.

Firstly， the price DDPs charged does not reflect the real costs they invested. And this price is not the comparable price that can be used in antitrust view. One author clearly pointed out that personal information collected by a producer but not sold to customers cannot satisfy the hypothetical monopolist test or the Brown Shoe test： there is no sale， no customers， and no product substitution[47]. Price is a poor measure of the value of digital goods and service， which are often paid for by giving access to data[48]. So the standard Lerner equation[49]  doesn't apply in this two-sided DDP market. If the "price theory" paradigm is used to analyze the competitive behavior of these digital giants， there is no doubt that the result will be biased or totally wrong. Even if the DDP already has dominant position， it is impossible to define its market power by price theory because of lacking profits from platform operators. Usually， scale not profits seems a big concern for DDPs. Even if a DDP has a considerable scale and a dominant position in a certain field， it may still have no profit at all. Uber lost $4.5 billion in 2017 and the company's CEO said， "we can turn the knobs to get this business even on a full basis profitable， but you would sacrifice growth and sacrifice innovation."[50]  Meituan， a massive online services platform in China， reveals loss of nearly $8.5 billion in 2018. Amazon， the e-commerce platform giant， has only made a profit in recent years. In fact， some courts have recognized that the SSNIP test has some deficits in defining relevant market under the DDP economy. Beijing High Court held that although the search engine service is a free market， Baidu benefits from that free market， so that market is a relevant market from antitrust perspective.

Secondly， there is a so-called price transfer mechanism in the two-sided market of the DDP and the DDP can subsidize the user on one side through charging the users on the other side. This internal price transfer mechanism makes the price theory based on demand substitution impossible to apply effectively. There are generally free market and paid markets in one single platform market. Defining the relevant market by price changes on either side of the market is inconsistent with the actual market reality. In the above Baidu case， the court held that although the search engine service is a free market， it could also constitute a relevant product market because that market can bring Baidu economic value.

Thirdly， market boundary in the DDP market is not clear. Product cycles are short， borders between "markets" are blurry in the platform market[51]. The DDP market is dynamic， so it is difficult to logically segment it. It is also difficult to find alternative markets for these types of segmented markets. Because of these dynamics， it is perhaps not surprising that competition agencies and other regulators have struggled to define online markets accurately[52]. The service provided by the DDP is a multi-sided service， and its core is to use data resources to integrate platform information to obtain profits. In this sense， the platform should be treated as an information services provider and explorer. Intentional segmentation of the DDP market does not correspond to real business practices. Taking the Baidu search service as an example， the service provided by Baidu appears to be a web search service. In essence， the service provided by Baidu is an information matching service. Baidu collects relevant information from its users and using the information for targeted ads to make profits. It is illogical to separate the matching services provided by Baidu into various sub-services. In Google's case， the FTC suggested that it considered "general purpose search" as the relevant market. This neglected to consider that online search engines are just one way for consumers to get answers and find information[53]. The German Federal Cartel Office appears to be repeating this mistake in investigating Facebook for abuse of dominance in an alleged "market for social networks". Consumers can turn to a wide array of substitute services such as blogs and micro blogs， professional networks， online forums， photo and video sharing services， news aggregators， messaging services， product review sites， social gaming apps， and virtual worlds[54].

Fourthly， it is not easy to find an alternative market in the DDP market. There is always a difference in the content and scope of data collected by DDP. The way each DDP uses data varies a lot. DDPs can generally be divided into data collection platforms， data processing platforms， and data marketing platforms. Different DDPs have big differences in data collecting methods and capabilities. They also have very different ways of exploring the data. So， it is really difficult to compare two different DDPs. Every DDP has its own way of how data is collected， processed and explored based on its own business model. And the relevant laws and regulations also limit how the content and scope of data can be collected and explored. The information it collects is unique and irreplaceable and forms the basis of its own core competitiveness. Thus， in the DDP market， there is no so-called alternative market problem for data.

6 Containing Data Power

6.1 Data Activities

From the industry's point of view， there can never be too much data[55]. DDP not only collects user data information by itself， but also cooperates with third-party data collection companies to collect data on other platforms or through APPs. According to statistics， 88% of the free apps in Google's app store will share relevant data with Google. About 43% of APPs on Facebook will exchange data with Facebook. Facebook can receive highly personal information from certain apps even if the user does not have a Facebook account[56]. Other DDPs such as Twitter， Amazon and Microsoft also share and exchange data with external users of the platform[57]. For years， Facebook gave some of the world's largest technology companies （Microsoft， Amazon， Yahoo） more intrusive access to users' personal data than it has disclosed[58]. Virtually all platform providers track user activity on their sites and collect demographic， behavioral， and other data from users. Data points revealing our habits， social relationships， tastes， thoughts， opinions， energy consumption， heartbeats， even sleep patterns and dreams are correlated ever more ingeniously， extensively， and precisely with still other data points. Then computers sort， analyze， and use it all to refine and target highly personalized ads for us to see online. Data alone cannot guarantee the success of DDPs. The ability to analysis data is also important for DDPs. Big data and big analytics have a mutually reinforcing relationship. Big data would have less value if companies couldn't rapidly analyze the data and act upon it. The algorithms' capacity to learn increases as they process more relevant data. The belief is that simple algorithms with lots of data will eventually outperform sophisticated algorithms with little data. Part of this is due to the opportunity for algorithms to learn through trial and error. Another is seeing correlations from big data sets[59] . Also， algorithms learn through trial and error and finding patterns from a greater volume and variety of data[60].

As DDP collects more data on its users， and as its algorithms have more opportunities to experiment. The ability to monetize data effectively — and not simply hoard it — can be a source of competitive advantage in the digital economy. By processing all available information and thus monitoring and analyzing or anticipating their competitors' responses to current and future prices， competitors may easier be able to find a sustainable supra-competitive price equilibrium which they can agree on[61]. Theoretically， companies can pursue more than one approach to data monetization at the same time[62].

6.2 The Data Market is Integrated

For the same search service， the Chinese court defined the relevant market of Baidu service as a "search service market" （"Baidu Case"）[63]. The FTC thinks "general purpose search" is the relevant market for Google service （"Google Case"）[64]. For the social networks services， the Chinese court defined the relevant market as "Digital platform online promotion service market" （"WeChat Case"）[65]，  but the German Federal Cartel Office defined the relevant market as a market for social networks （"Facebook Case"）[66].  However， in the Baidu and Google search platforms， there are at least the following related markets： information search market， the internet advertising market （bidding ranking business market）， etc. On the WeChat and Facebook platform， there are at least two related markets： the instant messaging market， social software and service markets， etc. There is no doubt that the various distinct markets of the DDPs are not isolated. Is it reasonable to divide the DDP market into some sub-markets？ In the above-mentioned Baidu case， there are at least the following related markets： Internet information search market， Internet advertising market （keyword auctions market）， and so on. The court of this case finally defined the relevant market as the "search engine service market" on the grounds that Baidu provided Internet information search services to ordinary network users.

Baidu search platform is a typical two-sided market. In this market Baidu undoubtedly occupies a dominant position in information search market. However， whether Baidu occupies the dominant position in the Internet advertising market or keyword auctions market is a question worthy of discussion. If Baidu's keyword auctions market is identified as a branch of the traditional advertising market， Baidu clearly has no dominant position in this advertising market. In any case， Baidu is not likely to have a dominant position in the advertising market， even if the advertising market is further subdivided into the Internet advertising market， online Internet advertising market， and so on[67]. In this case if the relevant market is defined as an information search market， it is unfavorable to Baidu. While the relevant market is defined as an advertising market， it is unfavorable to other stakeholders. In the Tencent case， the court defined the value-added segmentation service market as the relevant market. It is certainly irrational to define a two-sided market （search platform market） as a one-sided market （information search market， online ads market， etc.） regardless of interconnection and inter-effect of different sides of markets. In fact， the two-sided market is an inseparable market. Without internet search services， there is no keyword auctions market; without keyword auctions market， there is no online search service. Intentionally dividing an essentially interdependent and indivisible market is not in conformity with business reality. The DDP market as a multi-sided market is an integrated market in which markets on all sides are interdependent and inter-linked. Therefore， the definition of the DDP's relevant market must treat the DDP market as a whole[68].

6.3 Defining Data Power

6.3.1 Data Power is Not About Price Change

The market power assessment of the data is different from the linear markets. Traditionally market power refers to the ability of a firm （or group of firms） to raise and maintain price above the level that would prevail under competition. Price theory focuses entirely on price and excludes non-price dimensions of competition. Firms may adjust quality and other attributes to compete instead of price and engage in other non-price strategies not considered. The exercise of market power leads to reduced output and loss of economic welfare[69]. We have clearly pointed out that the pricing model of the DDP differs substantially from the traditional linear economy， and the price structure has a so-called asymmetry characteristics. Cross-subsidy is a common phenomenon in the DDP economy. And the business structure of the DDP which is focused on data flow and harvesting rather than profiting made the price factor almost meaningless in assessing the market power of the DDP. So a traditional price change model cannot be used for market power analysis of the DDP. The conclusion based on the price theory is sure to be inconsistent with the actual monopoly status of the DDP.

6.3.2 Data Power is About Market Share

In the traditional economy， market share is an important indicator for market power analysis. In Europe， the Commission's view is that the higher the market share， and the longer the period of time over which it is held， the more likely it is to be a preliminary indication of dominance[70]. If a company has a market share of less than 40%， it is unlikely to be dominant[71]. Some scholars believe that under the digital environments， there may be no necessary connection between high market share and market power. They hold that it is a normal state that market share of the DDP is highly concentrated in a small number of firms. It is difficult to judge whether a firm has a dominant market position based on a high market share. Predicting market dominance by market share may not be in conformity with the current status of the digital industry， and may even hurt the normal development of the digital industry[72]. The Supreme Court in China also held that market share is only a rough or even misleading indicator for judging dominant position in digital environment[73]. However， in the DDP economy， high market share is undoubtedly the key indicator in analysis of market power. High market share means more users and more interactions in the platform， and the network effect in the DDP will attract more users and more interactions. Due to the interdependence of users from different sides of the platform and users' path dependence， when more and more users engage in the platform， the less and less users will depart from the service the platform provided. That means the platform will become a magnet for users when it has a high market share， and the barriers to entry will increasingly rise. When a fair number of users have strong attachment with the platform， the platform will get strong market power in manipulating the data and users to harvest monopoly profits and exclude others to enter the competitive service.

6.3.3 Data Power is About Data Exploring Ability

The collection and control of massive amounts of personal data is an important source of market power for core players in the global marketplace[74]. In determining market power of the DDP market， the controlling and monetizing ability of data resources should be primary factors to be considered.  The quality and quantity of data a DDP boasts is a prerequisite for platforms to stimulate users' engagement. The market power in data collecting can be proved by the quality and quantity of data harvested by the DDP. Non-rivalrous and non-exclusive character of data is beneficial for a platform to explore the data in other areas， but it also limits the market power of the DDP in some degree. Non-rivalrousness means that one party's use of data does not prevent another party from collecting and using that same data， even from the same source. Non-exclusiveness means that a firm cannot exclude others from collecting that same data. As a result， no single firm controls all， most， or even a significant amount of the total universe of user data[75]. The ability of limiting the effects of non-rivalriousness and non-exclusiveness of data and enclosing the data determine the power market of the DDP.

The DDP cannot prevent other parties from collecting and using same data， but it can collect and use unique data that is specifically tailored for its own use. The ability of harvesting unique data for platforms determines the extent and scope of market power of DDPs. Some platforms have powerful ability to use big data and algorithms to acquire desired data， while some platforms have poor ability and have to purchase data from third parties. The big DDPs such as Google， Facebook have very strong power in data harvesting because those kinds of platforms have some type of tacit collusion to control the data collecting market to exclude the competition in that field.

Access to data alone is not normally at the root of an online platform's success. For that， a DDP must provide real value that motivates user engagement[76]. Data processing and monetizing abilities are other factors that need to be considered when examining the market power of the DDP. The ability of data processing， aggregating and monetizing determines the type and development prospects of the business model of the DDP. Data analytics， the capacity to extract actionable information from data， is becoming an important source of competitive advantage[77]. There are big differences in the data processing and monetizing capabilities of the DDP. To assess the market power of data processing and monetizing， the relevant authority should pay attention to the ability of data processing， data matching， data mining， data analysis， data aggregation and data exploring ability. The stronger data processing and commercialization capabilities of a market power， the more powerful it ultimately is.

Data extension is another factor that needs to be considered when valuing the market power of the DDP. Data's non-rivalrous and non-exclusive character means that same kind of data can be used across different areas spontaneously with a very low marginal cost. Technically， there are no obstacles to the cross-application of data and information except some legal restrictions. Even these legal restrictions can be easily circumvented in various ways. As a result， almost all DDPs have the power and incentives to apply data resources to other areas. Then the analysis of market power of DDP should consider the possibility and scope of data extension. The more possible the data can be cross-applied， the more powerful of the platform will be.

It is worth noting that the various data markets for the DDP sometimes cannot be separated in a clear-cut way. The data exploring process is mostly integrated. Data harvesting is not alone process， which is usually accompanied by data mining， processing activities. The same is true with data processing， which may also be accompanied by data collecting activities. Therefore， data collection and processing activities may be found in data commercializing activities. When evaluating data power of the DDP， the relevant authority should not disregard this reality.

7 Case Study： WeChat Case

In the case of WeChat[79]，  Shenzhen Intermediate Court defines the relevant market of WeChat service as a "Digital platform online promotion service market"， and excluded the WeChat's social service market and data market into analysis of the antitrust behaviors of WeChat. The court held that in the online promotion market WeChat has no dominant position and that refusing third parties' use of its platform is justified based on WeChat's own rules. We will probably see a different result if we use the above-mentioned new paradigm.

7.1 Data Power of WeChat

Tencent's 2019 Wechat Data Report shows that WeChat monthly active accounts reached 1.2 billion， an increase of 6% comparing with 2018. That means that Tencent has controlled more than 1 billion users' data information. The data includes at least the personal identity information provided by users and other personal information Tencent harvested. Through WeChat value-added services， Tencent can easily command users' financial， interests， and personal orientation information.At the same time Tencent can also discover some unique data information through its powerful data mining capabilities. There is no doubt that WeChat has become the most comprehensive digital platform company that boasts the largest number of users' data in the world. In short， Tencent not only holds the information directly submitted by users， but also has the details of our daily life through data analysis. In a sense， in the face of Tencent， users are nothing but useful data information.

The core value of WeChat lies in the data. Through this data information， WeChat can not only achieve huge advertising revenue， but also has the potential to provide seamless value-added services through its one-stop platform[80]. Without strong user data support， WeChat is unlikely to receive more than 68.4 billion RMB in online advertising revenue in 2019. In addition， it is worth noting that the value-added service provided by WeChat is a one-stop service. This one-stop service provides users and partners with a whole solution to online service discovery， delivery and settlement. The precondition for all these value-added services is WeChat's huge data. JD， the giant E-commerce platform in China， has to seek co-operation with WeChat due to WeChat's huge data flow. Undoubtedly， WeChat， which has nearly 1 billion user personal data， has absolute market dominance in the personal data market. The data power of WeChat is as follows：

The first is to obtain ultra-high profits through data. Through the analysis of personal data， Tencent can conduct accurate advertising， sell data to third parties to obtain advertising revenue.

Second， Tencent can easily expand into related fields using the data. WeChat is increasingly expanding its value-added services through its strong data market power， which are the most important profit areas of Tencent today. Now it has become a giant entity of instant messaging， social， e-commerce， gaming， payment and travel services， etc. The basis for the development of these extended services is the huge data information owned by WeChat. The relationship between the data information and Tencent's extended service is a typical skin-to-hair relationship. The essence of WeChat's market power in these value-added services is the extension and expansion of market power in user data.

WeChat's market power is undoubtedly derived from its power in the data market. With the help of data WeChat can be easily successful in the instant messaging market and the social service market. And it can also quickly gain market power in the relevant value-added services market. When analyzing the market power of WeChat， if we completely ignore the data market's power and only pay attention to the segmented value-added service market， which is inconsistent with the WeChat business model， this artificially narrows the scope of market power analysis. The analysis of WeChat's market power focusing on so-called segmented value-added market completely ignores the interdependency and relevance between WeChat's related markets. The so-called value-added service provided by WeChat is actually just one part of WeChat's one-stop seamless service. The one-stop seamless service， not the segmented service， is the business model that WeChat claims. The service provided by WeChat is an integrated social data collection， integration and utilization service. Therefore， the relevant market of WeChat should be the market for the collection and utilization of social data.

7.2 Re-evaluation of Tencent's Refusal to Trade

Article 17 （3） of the Anti-Monopoly Law in China expressly prohibits firms with market dominance from refusing to trade with third parties without justifiable reasons. The refusal to trade under the current monopoly law in China is based on traditional economics. Under the traditional economic model， although the refusal of trading behavior is widely investigated and prosecuted as an act of abuse of market dominance， the final proportion of illegality is very low[81]. For the digital platform economy， the digital platform is a software-based internet service. Based on the openness， scalability， two-sided effect， network effect and locking effect of platform services， digital platforms are generally willing to open their platforms to third parties to obtain more data resources. Usually it is reasonable for a digital platform to reject the third party's request to use its platform because the digital platform has invested a lot in developing the platform. The third party's desire to join in the platform is actually to take advantage of the data resources in the platform， which are very critical to the platform's development. The platform certainly has the right to decide who can use its precious data resources. Therefore， generally it is not an antitrust activity as even the digital platform rejects the third party's attendance in its platform. Under what circumstances does the digital platform provider's refusal to trade constitute a monopolistic act？ We think that when the DDP with market dominance needs to open the service to third parties in accordance with relevant industry practice and refuses to open， its behavior constitutes antitrust behavior. Some platform's business models are inherently open， and they allow third parties to freely use the platform data to gain benefits. Such a platform's refusal to trade will be an antitrust activity if it refused to open its service to a third party. Some super DDPs such as Amazon and Alibaba can be defined as key infrastructures in the platform economy. Those kinds of platforms are open platforms and all qualified third parties can join the platforms. If these platforms do not justify its refusal to trade， then the selective transactions constitute monopolistic behavior. In addition， the DDP prohibiting others from obtaining user data published on their platforms may also constitute a misuse of the data market power.

In China's first vertical search case[82]， the court prohibited third parties from collecting user data information published on the platform; but in Hiq Labs v. LinkedInd[83]， the US court held that the plaintiff could not prevent the defendant from accessing the public available data in the platform. From the perspective of the data market， the platform monopolizing the open data resources and prohibiting others to obtain those kinds of open data is harmful to the normal exploitation of data， which is the core resource for the development of the digital economy. Refusing a third party's participant in the platform and prohibition of using of freely available open data can only strengthen the market power of the platform to some extent and undermines the platform economy competition.

As far as Tencent's case is concerned， the court held that the relevant product market in this case should be an online promotion service market of internet platform. The plaintiff claimed that the relevant product market in this case should be defined as instant messaging， social software and service market. The court held that the plaintiff's definition failed to clarify the independent relationship between value-added services and basic services provided by WeChat， and deviates from the promotion nature of the plaintiff's demand for the WeChat public account. However， according to the single market theory of the DDP， Tencent's basic service market and value-added service market are not independent but interdependent markets. It is impossible to have a so-called value-added services market without the basic service market of WeChat. In fact， What Tencent provides is an integrated internet information service based on data information. Therefore， the relevant market for this case should be the market for collecting and exploring social networking user's data information. In this data market， Tencent undoubtedly has a high market share and has a dominant market position. Based on this precondition， does Tencent have any justification for the ban on the plaintiff's WeChat public account？ In this case， the court held that the plaintiff's operation of the WeChat public account was not only in violation of the "Service Agreement" and "Operational Regulations" of the WeChat's public platform， but also seriously undermined the WeChat user's communication environment and the normal use of WeChat software. However， can the court rule out the application of antitrust rules based on Tencent's own rules？ Undoubtedly， Tencent can't exempt itself from an anti-monopoly review on the grounds that the users violate relevant operating rules. Otherwise， Tencent itself becomes the law enforcer of its monopolistic behavior. In this case， the court should not condition the legality of Tencent's refusal to trade on Tencent's own rules. The correct analytical logic of the court in this case should be whether Tencent's refusal to trade in the data market is detrimental to competition. From the existing facts in this case， at least Tencent's selective trading behavior can be easily found. Such selective trading behavior is undoubtedly not justified when Tencent has a dominant position in the relevant data market. The misjudgment in the Wechat case established a terrible precedent for the platform economy. These days， we can see that Wechat is becoming more aggressive and more and more likely to refuse others.

8 Conclusion

The DDP market has some unique characteristics that are completely different from traditional brick-and-mortar market. The most notable feature is that the market power of the platform data market has a self-reinforcing function， which does not have a self-correcting mechanism. Based on the network effect and the fundamental value of data in the platform， the DDP will continue to strengthen data aggregation capabilities and data control， which is determined by the most essential business model of the DDP. Therefore， as far as the data market is concerned， the DDP will not weaken but will only strengthen the control of data. The lack of such a self-correction mechanism will inevitably lead to an intensification of the concentration of the DDP， and in a sense it will be easier for the platform to develop into a network service infrastructure. This will ultimately undermine consumer choice， hinder the entry of new competitors， and thereby undermine the competitive order. The nature of data makes the antitrust remedies of the past less useful. Breaking up a firm like Google into five Googlets would not stop network effects from reasserting themselves： in time， one of them would become dominant again[84]. With that in mind， careful intervention may be necessary to remedy market failure and promote customer welfare[85]. A radical rethink is required. A new "integrated data market" theory not only conforms to the commercial reality of the DDP market， but also effectively solves the problem that giant platforms fail to receive antitrust reviews. The new theory held that defining the relevant market of DDP data shall be a primary factor concerned. In assessing the market power of DDP， the relevant authority should consider the platform's data collecting， processing and exploring ability. The damages arising from platforms' abuse of power in data market are omnipresent. It is the right time to contain these digital empires and digital authoritarianism from abusing our data.

References：

[1] Reuben Binns， Elettra Bietti. Dissolving Privacy， One Merger at a Time： Competition， Data and Third Party Tracking[J]. Computer Law & Security Review， 2020， April （36）： 1-39.

[2] [66] Bundeskartellamt Prohibits Facebook from Combining User Data from Different Sources. [DB/OL]. [2019-2-12]. https：//www.bundeskartellamt.de/SharedDocs/Publikation/EN/Pressemitteilungen/2019/07_02_2019_Facebook_FAQs.pdf; jsessionid=E9D7480D3905A89B569E24B223CD8B76.1_cid387？ __blob=publicationFile&v=4.

[3] The Center for Global Enterprise. The Rise of the Platform Enterprise： A Global Survey [DB/OL]. [2020-10-30]. https：//www.thecge.net/app/uploads/2016/01/PDF-WEB-Platform-Survey_01_12.pdf.

[4][5] European Commission. A Digital Single Market Strategy for Europe-Analysis and Evidence [DB/OL]. [2020-9-13]. .

[6] Robert Pepper， John Garrity， Connie Lasalle. Cross-Border Data Flows， Digital Innovation， and Economic Growth[DB/OL]. [2020-1-08]. http：//reports. weforum. org/global-information-technology-report-2016/1-2-cross-border-data-flows-digital-innovation-and-economic-growth/#view/fn-3.

[7] Mckinsey. Big Data， Analytics， and the Future of Marketing & Sales [DB/OL]. [2020-12-08]. https：//www.mckinsey.com/～/media/McKinsey/Business% 20Functions/Marketing% 20and% 20Sales/Our% 20Insights/EBook% 20Big%20data% 20analytics% 20and% 20the% 20future% 20of% 20marketing% 20sales/Big-Data-eBook. ashx.

[8] Nick Srnicek. The Challenges of Platform Capitalism： Understanding the Logic of a New Business Model[DB/OL]. [2020-12-08]. https：//www. ippr. org/juncture-item/the-challenges-of-platform-capitalism.

[9] [10] Nathan Newman. Search， Antitrust， and the Economics of the Control of User Data [J]. Yale Journal on Regulation， 2014， 2（31）： 401-454.

[11] Avi Goldfarb， Catherine Tucker. Privacy Regulation and Online Advertising [J]. Management Science， 2011， 1（57）： 57-71.

[12] Alex Moazed， Nicholas Johnson. Modern Monopolies： What It Takes to Dominate the 21st Century Economy[M]. NEW YORK： St. Martin's Press， 2016： 101.

[13] Tarleton Gillespie. Platforms Are Not Intermediaries[J]. Georgetown Law Technology Review， 2018， 2（198）： 198-216.

[14] James Grimmelmann. The Platform Is the Message[J]. Georgetown Law Technology Review， 2018， 2（217）：217-223.

[15] Mark Bonchek， Sangeet Choudary. Three Elements of a Successful Platform Strategy[DB/OL]. [2020-8-05]. https：//hbr.org/2013/01/three-elements-of-a-successful-platform.

[16] [19] Joris Toonders. Data Is the New Oil of the Digital Economy[DB/OL]. [2020-5-09]. https：//www.wired.com/insights/2014/07/data-new-oil-digital-economy/.

[17] James Kanter. Antitrust Nominee in Europe Promises Scrutiny of Big Tech Companies[N]. New York Times， 2014， 3 October.

[18] [53] [54] [55] [56] [70] [81] Daniel Connor. Understanding Online Platform Competition： Common Misun Derstandings[DB/OL]. [2020-1-01]. https：//ssrn.com/abstract=2760061.

[20] Economist. Data Is Giving Rise to a New Economy[DB/OL]. [2020-12-09]. https：//www.economist.com/briefing/2017/05/06/data-is-giving-rise-to-a-new-economy.

[21] [33] [58] Gabriel Dance， Michael Laforgia， Nicholas Confessor. As Facebook Raised a Privacy Wall， It Carved an Opening for Tech Giants[DB/OL]. [2020-1-08]. https：//www.nytimes.com/2018/12/18/technology/facebook-privacy. html？ emc=edit_na_20181218&nl=breaking-news&nlid=69176904ing-news&ref=headline.

[22] [27] Elias Baltassis. Transforming Business Models with Big Data[DB/OL]. [2020-12-04]. https：//www.bcg.com/capabilities/big-data-advanced-analytics/transforming-business-models.aspx.

[23] James Manyika， Jacques Bughin， Jonathan Woetzl. Digital Globalization： The New Era of Global Flows[DB/OL]. [2020-8-30]. https：//www. mckinsey. com/business-functions/digital-mckinsey/our-insights/digital-globalization-the-new-era-of-global-flows.

[24] Kate O'Neill. Facebook's' 10 Year Challenge' Is Just a Harmless Meme—Right？ [DB/OL]. [2020-1-16].  https：//www.wired.com/story/facebook-10-year-meme-challenge/.

[25] Mark Johnson， Clayton Christensen， Henning Kagermann. Reinventing Your Business Model[J]. Harvard Business Review， 2008， 12： 57-68.

[26] [29] Gabriel Seiberth， Wolfgang Grundinger， Data-driven Business Models in Connected Cars， Mobility Services and Beyond[DB/OL]. [2020-12-09]. https：//www.bvdw.org/fileadmin/user_upload/20180509_ bvdw_accenture_studie_datadrivenbusinessmodels.pdf.

[28] Michael Schrage. How the Big Data Explosion Has Changed Decision Making[J]. Harvard Business Review，2016， 8： 18-35.

[30] Data Is Power： Profiling and Automated Decision-Making in GDPR[DB/OL]. [2020-12-09]. https：//privacyinternational. org/sites/default/files/2018-04/Data% 20Is% 20Power-Profiling% 20and% 20Automated% 20Decision-Making%20in%20GDPR.pdf.

[31] Behind Fake Data by Mafengwo Travelling Platform[DB/OL]. [2020-12-05]. http：//finance. sina. com. cn/stock/usstock/c/2018-10-23/doc-ifxeuwws7414888.shtml. In Wei zexi case， the false information in Baidu leads to death of a college student.

[32] Ariel Ezrachi， Maurice Stucke. Virtual Competition： The Promise and Perils of the Algorithm-Driven Economy[M]. Cambridge， Harvard University Press， 2016： 31.

[34] Ariel Ezrachi， Maurice Stucke. Algorithmic Collusion： Problems and Counter-Measures[DB/OL]. [2020-12-02]. https：//www.oecd.org/officialdocuments/publicdisplaydocumentpdf/？cote=DAF/COMP/WD%282017%2925&docLanguage=En.

[35] Price-fixing： Guidance for Online Sellers[DB/OL]. [2020-9-08]. https：//www.gov.uk/government/uploads/system/uploads/attachment_ data/file/565424/60ss- price- fixing- guidance- for- online- sellers. pdf2020; Jonathan STEMPEL. U.S. Announces First Antitrust E-commerce Prosecution[DB/OL]. [2020-12-04]. http：//www.reuters.com/article/us-usa-antitrust- ecommerce-plea-idUSKBN0MX1GZ20150406.

[36] Facebook to Pay Record $5bn to Settle Privacy Concerns[DB/OL]. [2020-12-29]. https：//www. bbc. com/news/business-49099364.

[37] Siva Vaidhyanathan. Violating Our Privacy Is in Facebook's DNA[DB/OL]. [2020-12-12]. https：//www.theguardian. com/commentisfree/2018/dec/20/facebook-violating-privacy-mark-zuckerberg.

[38] The Online Ad Revenue Index[DB/OL]. [2019-1-12]. https：//adrevenueindex.ezoic.com.

[40] Definition of Relevant Market [DB/OL]. [2020-10-04]. https：//eur-lex.europa.eu/legal-content/EN/TXT/？uri=LEGISSUM%3Al26073.

[41] Daria Kostecka-Jurczyk. Determination of the Relevant Market as a Criterion of Assessment of Concentration Effects in the Practice of Antitrust Authorities[J]. Wroclaw Review of Law， Administration and Economics， 2012， 2（2）： 129-138.

[43] Mark Jamison. Defining Relevant Markets in Evolving Industries[DB/OL]. [2020-10-04]. https：//bear.warrington.ufl. edu/centers/purc/docs//papers/1317_Jamison_Defining% 20Relevant% 20Markets% 20in% 20Evolving%20Industries%20Final.pdf.

[44] [45] Defining the Relevant Market in Telecommunications[DB/OL]. [2020-11-12]. https：//www.oecd.org/daf/competition/Defining_Relevant_Market_in_Telecommunications_web.pdf.

[46] Zhong Chun. Defining Relevant Market for Antitrust Enforcement in Internet Industry[J]. Legal Science， 2012， 4.

[47] Darren Tucker， Hill Wellford. Big Mistakes Regarding Big Data[DB/OL]. [2020-12-09].  https：//www.americanbar. org/content/dam/aba/publishing/antitrust_source/dec14_tucker_12_16f.authcheckdam.pdf .

[48] Economists Focus too Little on what People Really Care about[DB/OL]. [2020-4-02]. https：//www.economist.com/finance-and-economics/2018/05/03/economists-focus-too-little-on-what-people-really-care-about.

[49] The Lerner Index， Formalized in 1934 by Abba Lerner， Describes a Firm's Market Power. For the Details of This Equation， See A. P. LERNER. The Concept of Monopoly and the Measurement of Monopoly Power[J]. The Review of Economic Studies， 1934， 1 （3）： 157–175.

[50] Katie Roof. Uber Could Be Profitable if It Wants to， Says CEO[DB/OL]. [2020-12-18]. https：//techcrunch.com/2018/02/14/uber-could-be-profitable-if-it-wants-to-says-ceo/？guccounter=1.

[51] [52] [53] [54] [64] Daniel Connor. Understanding Online Platform Competition： Common Misunderstandings [DB/OL]. [2020-1-01]. https：//ssrn.com/abstract=2760061.

[55] Ronald Deibert. Three Painful Truths About Social Media[J]. Journal of Democracy， 2019， 1（30）： 25-39.

[56] Sam Schechner， Mark Secada. You Give Apps Sensitive Personal Information. Then They Tell Facebook [DB/OL]. [2020-2-23]. https：//www. wsj. com/articles/you-give-apps-sensitive-personal-information-then-they-tell-facebook-11550851636 .

[57] Reuben Binns， Ulrik Lyngs， Max Van Kleek， Jun Zhao， Timothy Libert， Nigel Shadbolt. Third Party Tracking in the Mobile Ecosystem[DB/OL]. [2020-1-02]. https：//arxiv.org/pdf/1804.03603.pdf.

[59] Ariel Ezrachi， Maurice Stucke （n 32） 24-25.

[60] Ariel Ezrachi， Maurice Stucke （n 32） 22.

[61] Competition Law and Data[DB/OL]. [2020-9-09] . http：//www.autoritedelaconcurrence.fr/doc/reportcompetition lawanddatafinal.pdf. Ariel EZRACHI， Maurice STUCKE. Artificial intelligence and collusion： when computers inhibit competition[J]. Illinois Law Review， 2017， 5：1776.

[62] Barbara Wixom， Jeanne Ross. How to Monetize Your Data[DB/OL]. [2020-9-05]. https：//sloanreview.mit.edu/article/how-to-monetize-your-data/.

[63] Beijing Number 1 Intermediate Court： Number 845 Minchuzi （2009）; Beijing High Court： Number 489 Gaominzhongzi （2010）.

[64] Daniel Connor. Understanding Online Platform Competition： Common Misunderstandings[DB/OL]. [2020-1-01]. https：//ssrn.com/abstract=2760061.

[65] Shenzhen Intermediate Court： Number 250 Guangdong 03 Minchu （2017）.

[67] [72] Jiao Haitao. The Constraint of Antitrust Enforcement in Internet Industry[J]. Shanghai Jiaotong University Law Review， 2013， 2.

[69] R. S. Khemani， D. M. Shapiro. Glossary of Industrial Organisation Economics and Competition Law[DB/OL]. [2020-6-10].  http：//www.oecd.org/regreform/sectors/2376087.pdf.

[70] Dominance and Market Power are Linked Concepts. To Be Regarded as Dominant in an Economic Sense， a Firm，or Group of Firms， Must Have Sufficient Market Power to Enable It to Raise Price or Act in some other Way Independently of its Rivals. Economists Focus on the Issue of Market Power rather than Dominance.'Market Power and Dominance（2011） [2020-9-07]. http：//www.compecon.ie/attachments/File/Dominance（1）.pdf.

[71] Antitrust Procedures in Abuse of Dominance[DB/OL]. [2020-10-03]. http：//ec. europa. eu/competition/antitrust/procedures_102_en.html.

[73] Xu Shuqing V. Tencent， China Supreme Court： No 5955 Minshen （2017）.

[74] Privacy and Competitiveness in the Age of Big Data[DB/OL]. [2020-8-05]. http：//europa.eu/rapid/press-release_EDPS-14-6_en.htm？locale=en.

[75] Daniel Connor. Understanding Online Platform Competition： Common Misunderstandings[DB/OL]. [2020-1-01]. https：//ssrn.com/abstract=2760061.

[76] Eliana Garces. Data Collection In Online Platform Businesses： A Perspective For Antitrust Assessment[DB/OL]. [2020-12-03]. https：//www. competitionpolicyinternational. com/wp-content/uploads/2018/05/CPI-Garces. pdf.

[77] David Kiron， Rebecca Shockley， Nina Kruschwitz， Glenn Finch， Michael Haydock. Analytics： The Widening Divide[J]. MIT Sloan Management Review， 2012， 53： 1.

[79] Shenzhen Micro SourceCode Development （MSCD） v. Tencent by Shenzhen Intermediate Court （（2017） Yue 03 Min Chu No. 250 （Shenzhen Intermediate Court））.

[80] WeChat—From Messaging App to Profitable Ecosystem[DB/OL]. [2020-10-20]. https：//www.fungglobalretailtech.com/research/deep-dive-wechat-messaging-app-profitable-ecosystem/.

[81] Zhang Zhiwei. Antitrust Regulation on Refusal to Trade of Internet Firms in China[J]. Jiangxi Journal of Jiangxi University of Finance and Economics， 2015， 3.

[82] Dazhong Dianping v. Aibang， Beijing Number 1 Intermediate Court， No 7512 Minzhongzi （2011）.

[83] hiQ Labs， Inc. v. LinkedIn Corporation， No. 3：17-cv-03301 （N.D. Cal. 2017）.

[84] The World's Most Valuable Resource Is No Longer Oil， But Data[DB/OL]. [2020-9-08]. https：//www.economist.com/leaders/2017/05/06/the-worlds-most-valuable-resource-is-no-longer-oil-but-data.

[85] Ariel Ezrachi， Maurice Stucke （n 32） 40.

壟断法下的相关数据市场研究

曹  阳

（上海政法学院，上海201701）

摘    要：数据是互联网平台经济的利润中心与关键驱动力。在对平台经济的反垄断审查中，相关机构很少将数据要素纳入审查分析范围。平台经济的反垄断审查分析中需重新审视数据要素的价值。互联网平台是在线经济结构的最有影响力的参与者。与传统的管道业务模型不同，平台市场是多方且相互依存的市场。追求规模化意味着平台须尽一切努力获取数据资源。数据不但有利于改善平台的获利能力，还有利于促进平台业务模型创新。数据市场垄断可能引发进入障碍、隐私侵害和消费者利益损害等。遏制数据市场力对市场竞争的损害需将数据要素纳入反垄断审查范围。在反垄断分析中应将数据市场视为整体。在定义数据市场力时，应考虑数据的市场份额以及收集与处理数据的能力。

关键词：数据驱动平台;数据力;数据市场;垄断行为