David Evans, Vanessa Yanhua Zhang, Apr 09, 2013
CPI Asia Column edited by Vanessa Yanhua Zhang (Global Economics Group) presents:
The Qihoo v. Tencent Landmark Decision
Evans is Chairman, Global Economics Group and teaches part time at the University of Chicago Law School and the University College London. Zhang is a Principal, Global Economics Group and lectures at Renmin University of China. The authors, together with Howard Chang, also a Principal of Global Economics Group, consulted for Tencent and prepared evidence and analysis on this matter.
On March 28, 2013, in one of the most significant antitrust decisions made by Chinese courts, the Guangdong High People’s Court (Guangdong High Court) dismissed claims on anticompetitive bundling and exclusionary practices brought by Qihoo against Tencent.1 The Court issued a relatively sophisticated and nuanced analysis of market definition and market power that examined Internet-based competition, recognized the importance of multi-sided platforms in this competition, and highlighted the critical role of dynamic competition. It now stands as both a landmark decision in China and an exemplar of serious antitrust analysis of the Internet sector for courts and competition authorities around the world. This article provides a summary of how the Court approached market definition and market power.
China has a large and rapidly growing Internet sector based primarily on domestic Chinese firms. According to China Internet Network Information Center (CNNIC), there were 564 million Internet users in China – 42.1 percent of the population – at the end of December 2012.2 Tencent, which focuses on instant messaging, was the largest firm based on monthly users with 340 million monthly active users in November 2010.3 Qihoo, which focuses on antivirus software, was the fourth-largest with 275 million monthly active users in November 2010.4 Most of the major global Internet players do not have significant shares of traffic in China. Instead, services similar to those offered by international players are offered by domestic companies. Most of these platforms, including Qihoo and Tencent, earn a significant portion of their revenue by selling advertising on the web pages seen by users.
Tencent attracts users by providing a variety of free services including instant messaging (IM), micro-blogging similar to Twitter, online games, online security, social networking, search and e-commerce. QQ, its free instant messaging service, had 399 million monthly active users as of February 2012.5 Qihoo attracts users by providing a variety of free services including online and mobile security such as anti-virus software, a web browser, and a game platform with games developed by third-party game developers. Qihoo’s Safety Guard, which is an Internet security product, had 366 million monthly active users in February 2012.6
Qihoo’s claims relate to a series of events, widely publicized in China, which took place in late 2010. It filed a private antitrust case before the Guangdong High People’s Court on November 15, 2011, accusing Tencent of violating China’s AML on the grounds of tying and exclusionary behavior. Qihoo asserted that the relevant antitrust product market was integrated IM. Integrated IM includes text, audio, and video communication capabilities. Qihoo also claimed that Tencent’s 76.2 percent share of that product market demonstrated that Tencent has the significant market power necessary for a dominant position under China’s AML.
The Guangdong High Court used the evidence before it to conduct a hypothetical monopolist test to evaluate Qihoo’s proposed relevant product market definition. It considered both demand and supply-side substitutability in doing so. Integrated IM services are provided for free. Qihoo’s economic expert argued that it was possible to conduct a qualitative SSNIP test by focusing on the possibility that the hypothetical monopolist could reduce quality or increase the “hidden” price of the user looking at advertisements. The Court rejected this reasoning and focused on evidence related to price: “This Court believes that this case reflects one of the main characteristics of the products and services provided by Internet Service Providers, i.e., almost all of the suppliers set the price of its basic service at zero…” The Court then went on to cite empirical evidence based on consumer surveys that users would switch away if the provider charged for the service. It concluded that while quality and advertising were relevant, “a more important factor to be considered is whether a lot of demand substitution will be generated if a hypothetical monopolist charges the service at a small scale continuously.”7
The Court then concluded based on the evidence on price sensitivity that consumers of integrated-IM would likely switch to free component services if the hypothetical monopolist raised its price modestly from zero. Importantly, from the standpoint of sound antitrust economic analysis, the Court rejected the plea from Qihoo’s economic expert to focus on functional differences between products in favor of quantitative evidence, limited although it was, on the elasticity of demand.
The next market definition issue was whether micro-blogging (Chinese services similar to Twitter) and social networking belonged in the relevant product market. Qihoo said no. To support this conclusion its economic expert calculated the correlation coefficient between the use of social networking and IM software on a weekly and monthly basis over a short period of time. It reported that these correlation coefficients were close to zero and claimed this showed lack of demand substitutability. The Court rejected this conclusion for a number of reasons. Most importantly, from the standpoint of sophistication, the Court criticized Qihoo’s economic expert for not considering dynamic competition: “This Court finds that competition is a dynamic process, and when defining a relevant market in the antitrust lawsuit on the abuse of a dominant market position, we must consider the status quo and future trends of development of relevant industries.” Qihoo’s expert, in particular, had ignored the fact that the use of micro-blogging and social network were rapidly increasing. The Court concluded the micro-blogging and social networking services belonged in the relevant market.
The Court then examined an argument presented by Tencent that the relevant product market for evaluating Qihoo’s claims consisted of “Internet application platforms.” The Court concluded that the dominant form of competition in the Chinese Internet industry involved platforms competing for the attention of users and then selling that attention to advertisers:8
The Court finds that, firstly, Internet application platform as a form of business model is becoming more and more common. Consequently, users, traffic, and usage time become the main focus of competition on the Internet…. Obviously, [a number of Chinese Internet companies] provide free services to attract a large number of users, and then take advantage of the huge user resources in the operation of value-added services and advertising to make profit. In turn, they use the profit generated from value-added services and advertising to support the survival and development of their free services. This has become the typical business model in the Internet industry. In this business model, the real competition among service providers is about the number of users, page views and effective usage time. The reason is that more users generate greater traffic and more effective usage time, which lead to higher profits from advertising and value-added services. Vice-versa, those companies can survive and grow their business only by providing an integrated platform to attract more users and increase their effective usage time.
While the Guangdong High Court did not conclude that application platforms was the relevant product market, it relied on the competitive constraints coming from attention rivalry, among the other factors discussed above, to reject the integrated IM product market put forward by Qihoo and its economic expert.
Having rejected Qihoo’s claim that the relevant product market consisted of integrated IM in China, the Guangdong High Court also found that the fact that Tencent had a 76.2 percent share of the overall instant messaging segment could not prove that it had monopoly power. The Court went on, however, to consider whether Tencent could have significant market power even under the assumption that the relevant product market consisted of integrated IM services in China. Importantly, it rejected the notion that market share data was sufficient to establish market power in this instance. “[D]ue to special market conditions of the Internet industry, market share, in particular cannot be deemed as a decisive factor in the determination of a dominant market position.”
The Guangdong Court dismissed Qihoo’s bundling and exclusionary abuse claims since Qihoo had not succeeded in identifying a relevant antitrust market in which Tencent held a dominant position. However, the Court’s analysis of the tying claim is worth noting. It found that combining the installation of IM and security software helps users by increasing the performance of their computers and, therefore, was an economically rational decision for Tencent to make. It also found that Qihoo had not shown any evidence that Tencent’s alleged tying has excluded competition from the market.
The Guangdong High Court’s decision in Qihoo v. Tencent is a landmark in several respects. It provides the most detailed antitrust analysis thus far by a Chinese court concerning the proper approach to market definition generally, and in abuse of dominance cases in particular. It demonstrates that the Chinese courts, barely five years after the AML went into effect, are already adept at rendering sophisticated antitrust opinions and applying modern economic concepts. It suggests that economic analysis, and economists, will play a significant role in private antitrust litigation in China as they have in US private antitrust litigation and in cases brought by the European Commission. Qihoo and Tencent both retained economists and introduced economic evidence before the Court. Finally, it is one of the world’s leading and most sophisticated antitrust decisions concerning competition in Internet-based industries. Although some commentators will certainly disagree with specifics of the analysis, the Court provided a relatively nuanced analysis of multi-sided platform competition, rivalry in attention markets, the application of the SSNIP test to products that are offered free of charge, and the role of dynamic competition.
1 See the Guangdong Province High People’s Court’s Civil Judgment No. Yuegaofaminsanchuzi 2/2011. The Chinese version is available at http://www.gdcourts.gov.cn/gdcourt/front/front!content.action?lmdm=LM43&gjid=20130328040159946185.
2 See the 31st Report of Internet Development in China, CNNIC, January 2013, Chinese version available at http://www.cnnic.cn/hlwfzyj/hlwxzbg/hlwtjbg/201301/P020130122600399530412.pdf
3 iResearch data on number of users per month (November 2010).
5 iResearch data on number of users per month (February 2012).
7 The correct SSNIP test should consider the impact of price changes on both sides of the platform—users and advertisers—but the Court focused on the single-sided SSNIP analysis presented by Qihoo.
8 The Court’s analysis is consistent with the analytical framework described in this Evans, David S., Attention Rivalry among Online Platforms and Its Implications for Antitrust Analysis, forthcoming Journal of Competition Law and Economics. Available at SSRN: http://ssrn.com/abstract=2195340.