Since the 1990s, the society has been overtaken by a new era of technological development driven by information technology revolution and globalization. It has changed or is changing the way of creating and distributing wealth, the operation model of economy, and the modes of production and lifestyle of people. The information industry highlighted by ICT (Information and Communications Technology) is changing and advancing rapidly. The network externalities of the information industry are also remarkable. As network technology develops, some startup companies have become less dependent on the traditional Marketing Theory of 4Ps (Product, Price, Promotion, and Place) while expanding their influence by increasing the number of users with the aid of network. Platform enterprises, such as Alibaba, Baidu, Tencent, Microsoft, and VISA, not only have different types of customer groups, but also use a pricing discrimination strategy targeting different customer groups (search engine users and advertisers). In addition to the information industry, an increasing number of industries in the modern economy are organized based on platforms, such as hypermarkets, the traditional media industry, and intermediary agents. Consumers now can have access to a variety of products thanks to the benefits of externalities in two-sided or multi-sided markets. The network effect of winner-take-all accelerates the growth of platform enterprises in two-sided markets. Also, the strategies of cross subsidization and freemium adopted by these platform enterprises have drawn attention from the public, scholars and the government. VISA and other payment card companies were slapped with antitrust lawsuits; Microsoft’s practice of bundling was accused as illicit competition. The recent Didi-Uber merger also aroused a lot of discussion – would it lead to monopoly? Who is right and who is wrong? With the unique characteristics of two-sided markets taken into consideration, which strategy should be chosen in the management of monopolistic two-sided markets, laissez faire or government intervention? Opinions vary.

David S. Evans and other experts suggest that regulations over monopolistic two-sided markets might lead to lower policy effectiveness (Evans, 2003). On the one hand, in monopolistic two-sided markets, the price difference between monopoly price and socially optimal price is smaller than that between two prices in a traditional market (Evans, Trirole, 2003), so the effectiveness of regulations is limited. On the other hand, with government intervention implemented in monopolistic two-sided markets, the consequences of the policy might be more complicated than those in a standard market and prone to more potential risks (Wright, 2004). According to Wang Hongtao and Lu Weigang (2012), the current regulations were made in the case of one-sided markets with no consideration of the characteristics of the telecommunications network industry as a two-sided platform, resulting in the regulation failure. Generally speaking, mergers between competitors enhance their market power, raise prices and hurt customers. However, such a theory does not necessarily apply to two-sided markets. Przemyslaw Jeziorski (2014) employs data from the 1996-2006 merger wave in the U.S. radio industry and examines mergers in two-sided markets. He identifies the incentives for merged firms to exercise market power on both listener and advertiser sides of the market and disaggregates the effects of mergers into changes in product variety and advertising quantity. His research finds 0.2% listener welfare increase and 21% advertiser welfare decrease. Chinese scholars Li Xinyi and Wang Haohan (2010) carry out an empirical study of the pricing decision-making and welfare of horizontal mergers in two-sided markets based on the empirical data from network media industry from 2005 to 2007. They find out that horizontal mergers between platform enterprises have no significant influence on the pricing of both sides of the market. Yet Weyl and others hold different views. According to Weyl, government regulations over monopolistic two-sided markets play an important role in that they can improve social welfare to a great extent. In monopolistic two-sided markets, socially optimal price is usually lower than market price, so the external market power, together with the twist effect of the internal monopoly, has impact on the optimal pricing policy of platform enterprises. Compared to traditional monopolistic markets, monopolistic two-sided markets have an invisible hand that is detrimental to the improvement of efficiency and welfare. In addition, the difference between personal and social benefits twists the pricing of the platform, as was mainly manifested in two aspects – the total platform price and the balance of prices on both sides. For this reason, in monopolistic two-sided markets, the price intervention of the government aiming to improve social welfare plays an important role. In reality, opposing voices of sellers as well as buyers against the monopoly of two-sided markets are also heard. Examples are found in the issues of slotting allowances and fees of hypermarkets and too many advertising slots in TV programs.

With all these disputes and complaints put aside, one of the challenges for the government is how to decide whether monopolistic platform enterprises exist in two-sided markets. In one-sided markets, two principal factors of market structure – market concentration ratio and market entry barriers – can help to see whether monopoly exists. Market concentration ratio is the most common economic indicator in industrial economics. It is a function of the number of firms and their respective shares of the total production in a market. The top ones are leading enterprises or oligopolistic enterprises within the industry. The higher the market concentration ratio is, the smaller number of enterprises within the industry is. Entry barriers can be used to discuss the dynamics of market relations from the perspective of potential new entrants into a market. It can contribute to the investigation of the competition between incumbent enterprises and potential new entrants. It reflects the intensity of potential competition in a market. In an industry with high market concentration ratio, if the entry is easy for potential new entrants, then the competition within the industry will not be reduced. According to Michael Porter’s Diamond Model, potential new entrants may enter the market at all times and yet there is still pressure of competition. However, these two factors may not be applicable in the discussion of Internet plus platform enterprises in two-sided markets. As for traditional intermediary platform enterprises, such as real estate agencies, matchmaking agencies, and household staffing agencies, the start-up costs are relatively low and they take up a small portion within the industry. As for platform enterprises whose business is based on the Internet, such as payment card systems, e-commerce payment systems, car rental booking systems and matchmaking agencies, the construction costs of platforms are huge, including costs on infrastructure, maintenance, two-sided or multi-sided client access, two-sided client matching and coordination, platform operation and the formation of information ecological chain. In addition, the construction involves the Internet, so there exist technological barriers. A TCP/IP-based network includes at least the following layers: the basic physical layer accessed by such networks as fixed network and mobile network, logical link control layer, application layer and session layer. In such a layer system, once the dominant operator takes control of the basic physical layer and adopts the vertical integration strategy or partial vertical integration strategy, it gains power in the marketplace within such a network. Then it can use exclusive strategies, such as bundle application and market foreclosure. In the end, the geographic market in the Internet industry is composed of many IP protocols. With IP addresses, relevant links can be extended to users by those IP addresses. In one-sided markets, some strategies are prohibited by antitrust laws, such as bundling, price discrimination and exclusive dealing, which turn out to be common yet useful strategies used by platform enterprises in two-sided markets. Such a logic can be found in the following cases – the Google-DoubleClick merger, the planned merger between two leading Dutch flower auctions Bloemenveiling Aalsmeer and FloraHolland, the Tangshan Renren’s lawsuit against Baidu, the Qihoo 360’s lawsuit against Tencent, the China Netcom V.S. China Telecom broadband network issue and etc.

The research of two-sided markets has been on the rise around the world since the conference “The Economics of Two-sided Markets” co-organized by IDEI (Institut D’Economie Industrielle) and CEPR (Center for Economic and Policy Research) in France in 2004. Rochet Tirole defines a two-sided market from the perspective of price structure as one in which the volume of transactions between end-users depends on the structure and not only on the overall level of the fees charged by the platform. Armstrong introduces the concept of cross-group network externalities to the discussion of two-sided markets – positive network externalities of one group arises with the increase of the number of agents from the other group. A brief list of such two-sided markets includes hypermarkets, intermediary agencies, and such companies as Taobao, Amazon, JD and LeTV with the popularization of Internet plus. Rochet Tirole is the first one to define and explain the concept of two-sided markets. He points out that in a market, the price charged by the platform to the buyer (PB) and the price charged by the platform to the seller (PS) adds up to the total price charged by the platform to the two sides (P=PB+PS). In the context of a buyer-seller interaction mediated by a platform, if the volume of transactions depends on the price level of the platform (P), but not on its allocation between the two end-users, the market is one-sided. If the total price (P) remains unchanged, any change in the price structure (price allocation) has impact on the demands and degree of participation of end-users on both sides, and then affects the volume of transaction in a platform. Armstrong has a further explanation of network externalities. The network externalities of two-sided markets not only depend on the number of agents from one group in a platform, but also on the number of agents from the other group who join the same platform. To take the credit card market as an example, the demand for credit cards not only depends on the cost on purchasing a credit card and the number of consumers who use a credit card, but also on the number of retailers who accept a credit card. In the same way, the demand of retailers for credit cards not only depend on the cost on accepting a credit card and the number of retailers who accept a credit card, but also on the number of consumers who carry a credit card.

Evans’ division of three types of platform enterprises – market makers, audience markers, and demand coordinators, is based on the service types provided by two-sided markets. At the pioneering stage of entrepreneurship, platform enterprises have to induce customers by charge-free strategy or even subsidies. Cailaud Julien points out that at the initial stage of the establishment of two-sided markets, platform enterprises usually use preferential policies or even subsidies to attract a large number of customers on one side, which in turn boosts development on the other side. Such promotion strategies raise controversy on dumping. When a platform enterprise grows to a certain phase, its operation is at a certain scale. To sum up, platform enterprises at two-sided markets have their own unique competitive strategies, such as pricing structure – the price charged by the platform to the users on one side is lower than the marginal cost, or even it is a negative price, while the price charged to the other side is higher than the marginal cost. In the context of traditional one-sided markets, such strategies are considered by antitrust law enforcement agencies as predatory pricing and monopoly. Other strategies, such as cross subsidization and bundling, are also controversial. The disputes over the Windows operating system and the VISA/MASTER payment cards interchange fee litigation are the examples. Similar examples can also be found in the recent development of Didi and Uber.

Past cases, recent examples and researches by various scholars all suggest that new criteria, which are different from those applicable to traditional one-sided markets, should be adopted in the context of two-sided markets. Based on a review of various research results, suggestions are given as follows:

1. According to the theory of industrial organization, there exists a positive correlation between equilibrium price and market share, but such a correlation is uncertain on one side in two-sided markets where market share has limited influence on the identification of market dominance and the correlation between market concentration and competition is uncertain. Therefore, it is suggested that the SSNIP (Small but Significant and Non-transitory Increase in Price) test used by the EU, namely the hypothetical monopoly test, is gradually applied. The SSNIP test seeks to identify the smallest relevant market within which a hypothetical monopolist or cartel could impose a constant and profitable significant increase in price. This is the mainstream method used in many countries to define a relevant market, although it is accused of the inaccurate estimate of quality and hidden prices.

2. A competition simulation test is suggested. We should refer to the environmental impact assessment in the “Not-In-My-Back-Yard” effect and have a critical loss analysis simulation of the stakeholders and end-users involved during the merger of competition. Before approving the merger application of two leading Dutch flower auctions, the Netherlands Competition Authority (Nederlandse Mededingingsautoriteit, NMa) did a critical loss analysis and also tried to inquire the intention of transfer of the planters and consumers to determine the actual loss caused by the reduction of competition. The Netherlands Competition Authority also took into consideration the impact of cross-group network externalities on the critical loss analysis. Such a method is recommended to be localized.

3. The core problem of the antitrust analysis based on direct evidence lies in the identification of market dominance and the assessment of the competitive effects on enterprises. The definition of a relevant market and market share is only one way to identify market dominance. Any direct evidence that helps to identify the market dominance of an enterprise can be used. As pointed out in the case of the Federal Trade Commission v.s. Indiana Federation of Dentists and the Horizontal Merger Guidelines jointly released by the Federal Trade Commission and the Department of Justice of the United States, the market delineation is not a necessity.

4. The focus of attention in market regulation should be on the price charged to end-users and the quality of service. To take Didi as an example, what matters is the price and quality of the ride-hailing service. On July 28, 2016, seven departments, including Ministry of Transport, Ministry of Industry and Information Technology and Ministry of Commerce, issued the Interim Measures for the Administration of Online Taxi Booking Business Operations and Services. The Measures not only have set the threshold for market entry into the tailored taxi service and online taxi booking service market from the perspectives of both software and hardware, but also oriented the management mode from the entrance management (e.g. inspections) to results-based management (e.g. data tracking, trace identification) with the help of platform data.

Two-sided markets constitute a significant part in the modern economic system, which relies more and more on the operation of such complicated markets. Researches of two-sided markets are still preliminary. The government should take the monopolistic characteristic of platforms into consideration when making policies of price regulation over market entities in monopolistic two-sided markets. The complexity of the pricing strategy in two-sides markets, the internal price transmission mechanism on both sides in particular, results from the profit target of platform enterprises, the interdependence of both sides in the market and network externalities. Regulations over two-sided markets should be different from those over one-sided markets. Only price regulations that accord with the pricing strategy in two-sided markets can help to reduce distortion and raise the effectiveness of antitrust regulations. Compared to horizontal integration, the vertical integration of platform enterprises in two-sided markets deserves more attention from the government when making regulation policies. This is also a key issue for further research.

Note: The author is a professor of the School of Economics and Management, Tongji University and the director of the MPA Center of the School of Economics and Management. This research was supported by Project NSFC71473177, the National Social Science Foundation (Project No. 13&ZD041), and Tongji University Think Tank for Sustainable Development (Project No. 1200219317). The author is responsible for all consequences of this article. This article first appeared at Prosecutorial View (November 2016).