Web search war: Renren v. Baidu

In 2007, Renren decided to cut back on its investment in Baidu's bidding search. Shortly thereafter, Renren found that its website was no longer displayed as high as it used to be in Baidu's natural search when certain search terms were used, which allegedly had resulted in a significant loss of visitors to Renren's website. Renren claimed that Baidu lowered its website's natural search ranking as a direct response to its decrease in spending on the bidding search and that it had violated Article 17.1(4) of the AML, which prohibits a dominant party from restricting a third party to deal with itself or with selected parties exclusively without valid justification.

The case was filed with the Beijing No.1 Intermediate People's Court. In December 2009, Beijing No.1 Intermediate People's Court ruled in favour of Baidu. Renren appealed the decision to the High Court. In July 2010, the High Court affirmed. In its decision, the High Court addressed three issues common to almost all abuse-of-market-dominance cases: (1) what is the scope of the relevant market for Baidu's services in dispute; (2) whether Baidu has a dominant market position in the relevant market; and (3) whether Baidu abused its market dominance if it is found to have a dominant market position.

Decision of the High Court

Scope of the relevant market

The scope of the relevant market is often crucial to a litigant's case in AML-related litigations. The complainant often favours limiting the relevant market to the accused party's products or services and their close substitutes, and the accused party will argue for a broader and more competitive, relevant market. In this case, Baidu raised an interesting argument that, because it does not charge web users for internet searches, there is no relevant market for Baidu's services. Under the AML, the scope of the relevant market is essentially an economic analysis rooted in the substitutability of goods or services. The price of goods or services should be factored in, but it may be a stretch to argue that there is little or no entry barrier to define the relevant market as long as the relevant goods or services are offered at no charge to consumers. However, the High Court appears to have taken the position that non-profit organisations may operate without a definable relevant market, even though it does rule against Baidu on this issue for the reason that Baidu charges companies for a bidding search, which is part of Baidu's overall services and is inseparable from its natural search system.

The High Court further found that the relevant service market is the search engine market because no meaningful substitute exists for an internet search provided by companies such as Baidu. As to the relevant geographic market, the High Court held that it should be confined to the territory of China. During the litigation, Baidu argued that the relevant geographic market should be global as its services are available to companies and web users around the world. But the High Court found that most of its customers are from China and any company that competes with Baidu most likely will come from China as well.

The High Court's view on the relevant geographic market may have a significant impact to future litigants. In an era of economic globalisation, especially concerning the internet industry, companies accused of monopolisation will argue that any AML allegation should be evaluated against global competition for its products or services. The High Court in this case made clear that if the company's businesses mostly compete with those of other Chinese companies for Chinese customers, the AML analysis should be restricted to the Chinese market. It may be difficult for future litigants to argue for a relevant global geographic market in an AML-related case unless meaningful competition exists abroad between the litigants and foreign companies concerning the goods or services in dispute.

Dominant market position

1.Burden of proof

According to the AML, the accused party's market share in the relevant market is an important factor for evaluating whether it has market dominance. For single company monopoly, Article 19 of the AML provides that if the accused party's market share in the relevant market is more than 50%, it will be presumed to have a dominant market position.

Renren argued during litigation that there should be a presumption that Baidu has market dominance in the China search engine service market because numerous reports have attributed over 60% of the market share to Baidu. For example, a report from China Securities Journal identifies Baidu's market share at 65.8% and an article posted by Baidu on its own website claims that Baidu has over 70% of the market share.

The High Court agreed that the presumption of market dominance may arise if the market share of the company can be accurately determined. However, the High Court does not consider documents submitted by Renren sufficient proof that Baidu's market share does exceed 50% to trigger the presumption. First, some documents were not clear as to how the relevant market was defined or how market share was calculated. The High Court reasoned that market share deduced from a relevant market different from that accepted by the High Court could not form the basis for presumption. Second, for others in which the correct relevant market was applied, the High Court found a lack of sufficient discussion in these documents about the scientific and objective nature of their market share analysis.

Based on the High Court's reasoning, an AML complainant has the burden to prove the accused party's market share through evidence that (1) applies the same relevant market as that adopted by court during litigation, (2) clearly describes how market share is calculated, and (3) uses a scientific and objective method for market share analysis. In practice, litigants may experience significant difficulties satisfying the High Court's standard for proving market dominance. As this case shows, there are plenty of newspaper reports and journal articles placing Baidu's market share at well over 50%, but none submitted by Renren were able to convince the High Court that Baidu has market dominance. In addition, because of the uncertainty surrounding the court's eventual adoption of the scope of the relevant market, a complainant probably has to prepare market share evidence targeting each possible relevant market that may be accepted by court during litigation, which undoubtedly will complicate the complainant' preparation for the case.

2. Party admission rule

With regard to Baidu's own article about its market share being over 70%, Renren argued that this article amounts to party admission and can relieve itself of the burden to prove Baidu's market share. While acknowledging that Baidu indeed boasts its high market share in the China search engine service market, the High Court ruled that such statement does not constitute party admission. According to the High Court, only statements made during litigation can be considered party admission. In addition, because Renren is trying to prove an important element of its claim, it may not use Baidu's statement to satisfy its burden of proof.

The concept of party admission under China evidence rules mainly concerns the claimant's burden of proof. According to Article 75(1) of the Opinions of the Supreme People's Court on Some Issues Concerning the Application of the Civil Procedure Law (最高人民法院关于适用《中华人民共和国民事诉讼法》若干问题的意见), a party to the litigation does not have to prove facts or claims that have been expressly accepted by its opponent. It is generally accepted in judicial practice that the party admission rule is only applicable to statement made in court.

Having decided that the party admission rule does not apply to the Baidu article, the High Court went one step further, noting that the burden to prove an important element of a litigant's claim may not be relieved by the party admission rule. It is not clear why the High Court believes it is necessary to limit the scope of party admission to exclude important elements of a litigant's claim, and such holding apparently finds no support in statutory languages or SPC judicial interpretations. However, it does reveal the High Court's reluctance to decide important issues in dispute through party admissions. From the Court's perspective, a litigant has to prove its case through objective evidence, especially with regard to key elements of its claims, and cannot meet its burden simply by pointing to a careless statement made by its opponent out of court.

Abuse of market dominance

Since the High Court did not find Baidu to have market dominance, it is not obligated to delve into the issue of whether Baidu's alleged lowering of Renren website's ranking violates AML requirements if Baidu were dominant in the relevant market. However, probably under the belief that this is an opportunity to lay out its views on some important AML issues, the High Court addressed this issue as well. According to Baidu, the drop of Renren website's ranking is unrelated to the change of Renren's payment, and instead was caused by the proliferation of spam links on the Renren website in violation of Baidu's “spam link policy”.

Renren did not deny the existence of spam links, but argued that the spam links can be easily added to its website by any web operator, including Baidu, and Renren should not be penalised for these. Furthermore, Renren noted that the spam links had been showing on its website for several months before Baidu lowered its website's ranking, which took place coincidently only after Renren cut back on its investment in Baidu's bidding search.

In its decision, the High Court held that, because Baidu had presented a justifiable cause, it would not have violated the AML provisions by lowering Renren's ranking in the results from a natural search if Baidu had the dominant market position. The High Court noted the proliferation of spam links had a negative impact on the quality of the search results and companies such as Baidu were entitled to take certain steps to maintain the quality of their products or services.

The AML certainly recognises a company's right to exercise its discretion with regard to the parties with whom and the conditions under which it wants to deal. However, the company's right to refuse to deal with or set forth dealing conditions is not absolute, especially if such business decisions may result in the monopolizing of the relevant market. In this case, the High Court took Baidu's proffered business justifications at their face value and noticeably did not address the issue of whether they were simply pretextual to Baidu's attempt at monopolisation. Without further discussion on the impact to market competition and other options available to Baidu to deter spam links, the High Court may have missed an opportunity to provide clear guidance on the relationship between a monopolist's duty to maintain market competition and its right to exercise independent business decisions.

Draft SPC AML Interpretation and its potential impact on future cases

In response to the evidentiary standard set forth by the High Court, Renren lamented that future complainants will face an insurmountable task to prevail in abuse-of-market-dominance cases. The High Court also acknowledged the difficulty for complainants to meet their burden of proof, but it claimed that the standard laid out in its decision is mandated by the statutory requirement that civil litigants must prove each element of their claims. The SPC apparently has taken notice of this issue and several provisions of the latest Draft SPC AML Interpretation have attempted to make it easier for the complainants to meet their burden of proof.

In one of its provisions, the Draft SPC AML Interpretation provides that if the complainant has submitted preliminary evidence to prove the accused party has dominant market position, the court shall rule in favour of the complainant on this issue if the accused party is unable to rebut with sufficient evidence. This provision essentially allocates the burden to prove market dominance to both parties. The complainant has the initial burden to show that the accused party is likely to have market dominance and, if the complainant meets such burden, the burden is then switched to the accused party to show that it does not have market dominance. According to the Draft SPC AML Interpretation, information disclosed by listed companies, party admission, market research, economic analyses, special studies and statistical findings provided independently by qualified third parties may all be used by the complainant as preliminary evidence to prove market dominance. Thus, if this case were decided under the Draft SPC AML Interpretation, Renren probably would have satisfied its initial burden through evidence about Baidu's market share and the burden would then have been on Baidu to prove that it does not have market dominance.

The “party admission” which may be used as preliminary evidence under the Draft SPC AML Interpretation to prove market dominance does not appear to be limited to an in-court statement. On the contrary, if read in context, this could include all statements made publicly by the party accused of an AML violation. It is unclear whether this provision, if issued officially verbatim, should be read in conjunction with other evidence rules so that the scope of party admission includes out-of-court statements in all circumstances, or whether it is limited to situations where the admission may only prove market dominance. In any event, if the Draft SPC AML Interpretation was to govern in this case, it is likely that the article posted by Baidu on its website could have been used by Renren to prove that Baidu has market dominance.

Further, the Draft SPC AML Interpretation identifies market research, economic analyses, special studies and statistical findings made independently by qualified third parties as potential evidence to support allegations that the accused party has market dominance. The Draft SPC AML Interpretation does not go into detail as to how much information has to be provided in these documents. But since the accused party carries the burden of rebuttal, it is reasonable to conclude that the SPC does not intend to require a detailed description of an analytic method in these documents that the market share calculation is scientific and objective as the High Court declared in this case.

Final thoughts

A number of cases have been filed with Chinese courts since the AML came into effect in August 2008, most of which claim abuse of market dominance. Nonetheless, the Baidu case can be distinguished from previous litigations because a provincial high court, for the first time, conducted a substantive analysis of several important AML issues: the definition of the relevant market, the burden to prove dominant market position, and the finding of an abuse of market dominance. In this case, the High Court sets forth a high evidentiary standard for the complainant to prevail in AML cases.

Even though court decisions in China are not precedential, due to the High Court's position and influence, the Baidu case may have a profound impact on how other courts adjudicate AML cases in the future. However, some of the High Court's holdings in this case may not survive if the Draft SPC AML Interpretation is passed without major modification. The Draft SPC AML Interpretation makes a clear attempt to lower the burden of AML complainants to prove their case and large companies such as Baidu will surely face more AML scrutiny from courts in the future.

Fang Qi and Tingting Liao, Fangda Partners, Beijing