2.1. Distinction between “repeat players” and “one-shotters”

The theoretical framework of “repeat players” and “one-shotters” was first proposed by Marc Galanter in his 1974 paper. By examining litigation resources, the role of judges, the abilities of lawyers, and various other factors influencing dispute resolution outcomes, Galanter hypothesised that the justice system is more conducive to realising the interests of repeat players. Furthermore, since corporations are more likely to be repeat players and corporate shareholders tend to be the haves, the Great Society program has struggled to effectively implement pro-poor legislation in judicial practice, despite the passage of numerous pro-poor laws (Galanter, Reference Galanter1974, pp. 97–103).

Subsequent research has delved into the specific mechanisms through which repeat players gain litigation advantages, consistently validating these advantages across various contexts (Songer, Sheehan and Haire, Reference Songer, Sheehan and Haire1999, pp. 819–821), including China (Tian, Xu and Wei, Reference Tian, Xu and Wei2021, pp. 172–174). Notably, recent studies have cited this theory to explain contemporary judicial phenomena, such as the high winning rate of tech giants in their local courts (Chen, Reference Chen2023, pp. 129–134). Two of the most influential subsidiary explanatory theories are the party capability theory proposed by Stanton Wheeler et al. and the penetrated court theory proposed by Xin He and Yang Su. Since these two theories will be discussed in detail in Section 5, only their core content will be mentioned here. In essence, the party capability theory posits that the resources wielded by parties are the most significant factor causing outcome differentials between repeat players and one-shotters (Wheeler et al., Reference Wheeler, Cartwright, Kagan and Friedman1987, pp. 439–442). And the penetrated court theory argues that the influence of local governments and behind-the-scenes personal agreements are the primary contributors to these outcome differentials (He and Su, Reference He and Su2013, pp. 138–142).

In sum, most studies on repeat players have implicitly assumed that courts can only play a passive role in the judicial process. Faced with influence from repeat players or local governments, it appears that courts are limited to removing interference and deciding fairly on individual cases. While this is undoubtedly important, the court’s capacity for action extends beyond this. Some studies have noticed the flip side of the long-term relationship between the court and the party. For example, Andrea McAtee et al. found that experienced lawyers do not merely pursue the best interests of their individual clients; they also consider the underlying policy preferences of the judges, thereby expecting judges to give more weight to their arguments in subsequent cases (McAtee and McGuire, Reference McAtee and McGuire2007, p. 262). However, scholarly attention to the leverage enjoyed by the court remains largely limited to observations of individual cases and has yet to be systematically refined and analysed in terms of its exact mechanisms and outcomes.

2.2. Towards a relational approach to analyse judicial activity

The frameworks of relational governance (Williamson, Reference Williamson1981, p. 566) and relational contract (Macneil, Reference Macneil2000, pp. 881–892) were initially and most frequently used to understand how production is organised between firms. When examining transactions and contractual behaviours between two firms, we can approach it from two perspectives. First, we can view each transaction and contract as independent, a perspective supported by classical contract law. Alternatively, we can consider transactions within the context of enduring, adaptable commercial relationships, where opportunistic behaviours are effectively regulated by the dynamics of reputation and social networks. This latter approach is particularly relevant for long-term partnerships. Classical contracts are costly to draft, negotiate, and enforce. More importantly, when firms commit to specialised investments for enduring collaborations—such as acquiring specific equipment or recruiting specialised personnel—they inevitably develop significant dependencies on their partners. Therefore, relying solely on one-off contracts often fails to deter opportunistic behaviour from partners. In other words, relational contracts between two firms with long-term commercial relationships are not primarily designed to create incentives for performance and breach through the prospect of court-imposed monetary damages (Goetz and Scott, Reference Goetz and Scott1981, pp. 1089–1092). Instead, they are intended to confirm and nurture their long-term, mutually beneficial commercial relationship (Bernstein, Reference Bernstein2015, pp. 576–589; Macaulay, Reference Macaulay1963, pp. 62–65).

Some scholars in the field of social policy and administration have integrated relational thinking into the governance of socio-economic affairs by introducing the concept of “interactive governance.” This emerging governance paradigm draws inspiration from long-term cooperative business relationships, suggesting that public authorities and private actors can establish close partnerships and coordinate their behaviours based on mutually beneficial goals (Gu, Reference Gu2019, pp. 78–81; Luo, Zheng and Long, Reference Luo, Zheng and Long2023, pp. 64–66). In legal theory, research on relational governance has predominantly focused on responsive regulatory measures by the executive, such as administrative negotiation (Lu, Reference Lu2019, pp. 145–152), while overlooking the application of relational governance mechanisms by the judiciary.

If we liken the relationship between the court and repeat players to that of two companies engaged in long-term cooperation, we can discern two distinct perspectives to understand this dynamic: one is individualistic, and the other is relational. The individualistic approach mirrors the default mode of judicial proceedings, where each dispute is resolved solely on its own merits without consideration of other cases—a “case-by-case” approach that is well-suited for certain civil matters such as divorce and complex cases.

However, in practice, a significant portion of judicial caseloads consists of interconnected cases. For instance, there may be a series of cases involving the same defendant and identical causes of action. Alternatively, a judge presiding over one case might anticipate issues that could lead to disputes in other cases. In such scenarios, courts may employ flexible measures, such as judicial suggestions, to prompt repeat players to adopt additional compliance measures, thereby pre-empting future disputes or streamlining case resolution. For example, if A sues a platform for unnecessary collection of personal information, and subsequently B and C also file similar suits against the same platform, the court might instruct the platform not only to cease using A’s information but also to stop collecting unnecessary personal information from all users. If the company complies with this directive, all three cases could be resolved collectively, preventing future cases based on the same cause of action. In a word, from a relational perspective, the court, acting as “repeat referees,” can utilise ongoing interactions to encourage “repeat players” to enhance compliance and improve their litigation practices.

3.1. Specialised court

In 2014, three IP courts were launched in Beijing, Shanghai, and Guangzhou. Since then, several specialised courts have been established, including the IP court in Hainan (2020), the financial courts in Shanghai (2018), and Beijing (2021), and the Internet courts in Hangzhou (2017), Beijing (2018), and Guangzhou (2018). While the IP courts and financial courts operate at the appellate level, the three Internet courts function at the basic level. Specifically, Internet courts mainly handle four types of disputes: (1) disputes arising from the signing or performance of online shopping contracts through e-commerce platforms, (2) disputes over network service contracts that are both signed and performed on the Internet, (3) disputes over financial loan contracts or small loan contracts that are both signed and performed on the Internet, and (4) disputes over the ownership of copyrights or neighbouring rights of works published on the Internet for the first time.

Current research highlights three characteristics or functions of specialised courts. First, they increase professionalism for handling complex and novel disputes in specific fields of law. Generally, the criteria for judges appointed to the IP, financial, and Internet courts in China are stricter than those for generalist courts at the same level (Li, Reference Li2022, p. 502). Frequent interaction with specific industries and dealing with particular types of disputes also enhance judges’ expertise. For instance, knowledge gained from resolving one dispute with a company can become valuable experience when handling a similar dispute with another company. Second, the increasing caseloads in these specific fields necessitate more specialised judges. IP, finance, and Internet sectors have undergone significant transformations and are closely tied to the daily lives of consumers and industry operations. Consequently, these fields generate a large number of disputes. Compounding this issue, the legal rules in these areas are often either unclear or entirely absent. Third, various efforts have been implemented to enhance trial efficiency in specialised courts. One prominent measure is that Internet courts conduct cases online, meaning that “the acceptance, service, mediation, evidence exchange, pretrial preparation, court trial, judgment pronouncement, and other litigation steps are generally completed online.”

Based on the research mentioned above, this article seeks to provide a new perspective on the function of specialised courts—the “limited and frequently exclusive jurisdiction in one or more specific fields of the law” (Zimmer, Reference Zimmer2009, p. 46) makes specialised courts an ideal forum for relational governance.

3.2. Judicial suggestions and litigation source governance

In a 1956 reply, the Ministry of Justice stated that “people’s courts may and shall issue suggestions to urge relevant agencies and organizations to make amends when discovering flaws in their work that criminals can exploit. When necessary, [the court] shall suggest that their superiors at higher levels assist in rectifying [the flaw]. If the flaw is generic, [the court] may report to party and government leaders to instruct rectification.” This marked the first official appearance of the concept of judicial suggestions in a Chinese document. Since then, courts have sporadically sent judicial suggestions to government agencies and public institutions, such as banks. About 50 years later, the importance of judicial suggestions was reiterated in two notices issued by the Supreme People’s Court in 2007 and 2012. Consequently, the issuance of judicial suggestions has become much more frequent and regular. According to my fieldwork in Court P, each judge is assigned to draft a judicial suggestion each month. The court selects the best one to be sent in the name of the court.

Recently, judicial suggestions have become a key mechanism within the Litigation Source Governance program in Chinese courts. To alleviate the heavy caseloads faced by these courts, Litigation Source Governance (Suyuan Zhili) encompasses a comprehensive set of measures designed to address disputes without relying solely on case-by-case adjudication. According to several official documents issued by the Supreme People’s Court since 2021, the Litigation Source Governance program in China includes pre-emptive resolution of disputes, Alternative Dispute Resolution (ADR), judicial suggestions, and other methods.

Most existing studies have concluded that the success of source-of-complaint governance relies primarily on the active cooperation of local governments and the high-level promotion by party committees. Judicial governance, based on the relationship between the court and repeat players, can to some extent be regarded as another effective mechanism for realising Litigation Source Governance.

4.1. Judicial suggestions for tech giants

Judicial suggestions in China are recommendations made by courts to address issues identified during legal proceedings. These suggestions are not legally binding; their implementation depends on the willingness of the recipient. As a key mechanism for Internet courts to engage in the comprehensive governance of platforms, these judicial suggestions encompass topics such as information privacy, user agreements, protection of minors, and e-commerce transactions, primarily targeting platform enterprises. According to publicly available data, the response rate to judicial suggestions from the three Internet courts exceeded 70%, with one court achieving a 95% response rate.

During field interviews, judges also indicated that most platforms were willing to adjust their business practices in line with judicial suggestions. While a response letter may be superficial, genuine implementation entails substantial compliance costs for the company. An example of the effectiveness of judicial suggestions is a judicial suggestion from Court P to Platform J. In this case, the plaintiff sued Platform J in Court P, requesting the Court to declare a boilerplate clause in the user agreement invalid. Court P supported the plaintiff in the judgment, and recommended Platform J to revise its user agreement in the judicial suggestion sent after the judgment. Specifically, the Court recommended changing a clause that stated, “as long as you purchase from this platform, you are deemed to consent to the clauses below.” In response, Platform J promptly amended the user agreement so that the clauses would only come into effect after users ticked a consent box.

The primary issue facing most judicial suggestions is the low rate of response and implementation. Common solutions to this problem include high-level promotion by Political and Legal Affairs Commission, strengthening inter-departmental cooperation, and public monitoring. Some articles also propose legislation to allow courts to enforce judicial suggestions in the same manner as they enforce judgments. Several of these measures are already in practice. For example, the People’s Court Newspaper has published a selection of judicial recommendations, and Shanghai and Jiangsu have established databases for judicial recommendations.

However, judicial suggestions sent by the Internet court to tech giants are generally not made public. Even when they are, they are often anonymised or published alongside the platform’s response letter, minimising public pressure on the platform to comply. Additionally, Internet court typically sends judicial advice directly to the company and rarely seeks assistance from Political and Legal Affairs Commissions or relevant administrative organs.

Therefore, the high compliance rate of judicial suggestions from Internet courts to tech giants cannot be explained by public opinion pressure or by leveraging power from Political and Legal Affairs Commissions or the government. So, why are tech giants willing to comply with judicial suggestions from Internet courts? Exploring their long-term relationship may provide the answer.

4.2. Relation between Internet courts and tech giants

Tech giants typically act as “repeat players” in terms of litigation frequency, winning rates, and litigation resources. Several studies have examined the advantages that platform companies may enjoy in litigation, approaching these advantages from various angles. First, most apparently, platform enterprises often boast robust legal teams and engage experienced external counsel. Then, government may influence the judiciary to meet local economic development objectives, inadvertently granting advantages to tech giants, which are major taxpayers. Finally, their investments in cyberlaw research also provide them with substantial discourse resources in litigation. Some studies have highlighted a series of concepts that either have been or continue to be favourable to reducing company liability and safeguarding their interests. Examples include “technology neutrality” and “internet decentralization” (Dai, Reference Dai2023a, pp. 87–91). These concepts can influence court decisions directly or indirectly. And such influences may further solidify and perpetuate through judicial rulings.

However, from another perspective, frequent interactions with the court also bind tech giants more tightly and provide them with greater incentive to comply with court suggestions. On one hand, platforms aim to avoid repeated losses in similar litigation. Cases involving platforms typically share three distinct characteristics: first, they often revolve around common issues such as the validity of user agreement terms and the legality of handling personal information; second, they are numerous in quantity; and third, they tend to be concentrated in specific jurisdictions. Under provisions of the Civil Procedure Law and other regulations, a significant number of platform-related lawsuits are consolidated in a few courts. This concentration means that each court, and even each judge, faces a substantial volume of similar cases involving specific platforms. When cases are dispersed among different courts and relevant legal rules are unclear, platforms may perceive less risk in failing to adjust their behaviour in line with judicial suggestions from a particular court, as decisions in other courts could differ. However, when cases are consolidated in a few courts, and if the company fails to respond or implement guidelines proposed within former judicial suggestions, subsequent cases of the same nature may lead this court or judge to decide against the company with a high likelihood.

On the other hand, the active implementation of judicial suggestions by platform companies helps to maintain a favourable litigation image and social reputation, which is particularly important for large platforms under high public scrutiny. Implementing judicial recommendations serves as an effective signal to both the courts and society that the tech giants are actively complying with the law. Through judges’ social networks, reputation of the company is disseminated within the courtroom and even throughout the judicial system. This dissemination can occur in both formal settings, such as during judges’ meetings, and informal settings, such as during gossip over tea. Additionally, courts may sometimes post the judicial advice along with a note on the company’s implementation in their public accounts. This practice helps establish a socially responsible image for the tech giants among the public.

Furthermore, from a company’s perspective, judicial suggestions serve both as a deterrent and a commitment. Although drafted by one or two judges, these suggestions are uniformly delivered in the name of the court and can potentially influence the behaviour of other judges. Consequently, compliance efforts by platform companies in response to judicial suggestions often yield tangible outcomes. Essentially, repeat players stand a better chance of winning subsequent cases if they adapt according to judicial suggestion. For instance, in a recent case, a user contested a boilerplate clause in a platform’s user agreement, alleging it to be unfair. Following an adverse ruling against the defendant, the judge issued specific guidance on how to amend the clause. If the defendant modifies the clause accordingly, the company is unlikely to lose subsequent cases involving the same issue. In sectors characterised by regulatory uncertainties, such as emerging technologies, judicial suggestions often act as de facto safe harbour rules for major tech firms (Dai, Reference Dai2023b, pp. 33–37).

5.1. Preconditions for relational adjudication

Most parties seek smooth dispute resolution to minimise disruption to their daily lives, business activities, and administrative work, aligning with courts’ objectives to settle cases and reduce caseloads. However, professional litigation groups such as patent trolls or professional counterfeiters represent a notable exception. These groups profit from generating disputes and lack the reciprocal basis necessary for a constructive relationship with the courts. In contrast, their persistent engagement in litigation can hinder their profitability over time. For instance, in the case of professional counterfeiters, courts may approach liability of the defendant cautiously if they suspect the plaintiff intentionally purchases large quantities of dubious-quality goods solely to profit from legal damages. Consequently, these parties often conceal their litigation history and, when recognised by the court, may even deny their past actions to avoid scrutiny.

Additionally, intentions to cooperate must be substantiated with effective signals exchanged between parties. Based on field works in court and interviews with corporate lawyers, both the court and the party typically convey goodwill through “bonus moves” beyond their mandatory obligations, signalling their readiness to collaborate. Specifically, from the plaintiff’s or defendant’s perspective, these “bonus moves” include: timely and comprehensive submission of evidence to streamline judicial review processes; systematic adjustments to business practices based on court judgments and recommendations, particularly in cases with high homogeneity, to support the court’s objective of systemic resolution (“handle one case, govern one area”); and proactive engagement in mediation and open communication with adversaries in sensitive cases prone to future conflicts or public scrutiny.

From the court’s perspective, engaging in “bonus moves” primarily involves: gaining a thorough understanding of the industry’s overall landscape and specific business processes to propose more practical behavioural norms or recommendations; actively assisting parties in overcoming evidentiary obstacles by issuing subpoena or extending evidence submission deadlines within legal limits; respecting parties’ desires to maintain their reputations by refraining from publicising specific case outcomes; and dedicating additional time and effort to facilitate mediation when a party seeks to settle due to lower win probabilities or enforcement challenges, aiming to achieve mutually satisfactory mediation outcomes with reduced communication costs.

5.2. Conditions for maintaining relational adjudication

The cooperative dynamic between the court and repeat players also hinges on the frequent and long-term exchange of information. First, it is imperative for the court to clearly communicate its compliance expectations to the parties. Beyond relevant rulings and judicial suggestions, seasoned parties also glean tacit insights into the court’s compliance standards by monitoring its latest verdicts or engaging with the judge’s interactive discourses. Notably, judges and heads of corporate legal affairs often attend the same conferences. Furthermore, experienced attorneys frequently act as information brokers, ensuring that their clients swiftly and accurately grasp the court’s intentions. They also facilitate the dissemination of relevant information on a wider scale by publishing legal research reports on their official accounts.

Second, courts overseeing multiple cases involving the same party can gain precise insights into whether the company alters its behaviour in response to judicial suggestions. Consequently, to avoid losses in future disputes with similar causes of action and to maintain a positive reputation with judges, repeat players like tech giants often choose to adhere to judicial suggestions. Furthermore, intra-agency information exchange mechanisms, such as routine meetings or informal discussions over tea, play a critical role in mitigating the impact of judicial rotations or changes in legal representation. These mechanisms help maintain robust interactions between the court and repeat players.

Third, communications off the record help to keep the interaction between the court and the parties flexible. If the relationship between the court and the parties is analogous to that of two enterprises in a long-term partnership, judicial suggestions are the equivalent of the quality and manufacturing manuals issued by the buyer to the supplier. In business-to-business cooperation, a formal contract clarifies the rights and obligations of both parties, whereas non-mandatory materials, such as production standards manuals, provide additional information to facilitate smooth cooperation (Bernstein, Reference Bernstein2015, p. 573). Similarly, unlike judgment documents, judicial suggestion is not merely a unilateral order from the court to the parties but also an important channel for communication and consensus-building. To reduce the risk of misunderstanding or miscommunication, judicial suggestions and their replies (to which the public usually has no access) allow the parties to gain a deeper understanding of the specific compliance measures required by the court, while the court gains a clearer understanding of the business considerations and difficulties faced by the parties. This information, essential to maintaining a healthy and cooperative relationship, is often unstructured and difficult to convey through formal, public processes such as the filing of evidence and the issuance of rulings.

5.3. Limits and perils

As mentioned above, relational governance can have important positive effects, such as relieving caseloads in the courts, promoting stable market expectations, and incentivising compliant behaviours from repeat players. However, the potential downsides of relational governance should not be overlooked.

First, the court’s relational governance over the parties carries de facto behavioural incentives but lacks de jure coercive authority. Even when the court’s decisions are rational and its recommendations concrete, the feasibility of program implementation must be weighed. If the upfront cost of compliance outweighs the potential losses from an adverse ruling in future litigation, rational parties may opt against compliance. This partly elucidates why judicial suggestions from Internet courts consistently emphasise areas such as privacy, child safeguarding, identity verification, and boilerplates, while rarely addressing issues that touch the roots of the platform economy, such as platform monopoly or market concentration.

Second, another factor to consider is that while the information garnered through case-by-case adjudication may adequately address conflicts of interest between parties, it may not be comprehensive enough to inform decisions on broader dimensions such as market dynamics, social equity, and the like.

Third, relational governance may undermine judicial integrity. Some communications between the court and repeat players, such as tech giants, occur off the record, potentially concealing inappropriate transactions. Additionally, since documents like judicial suggestions lack coercive power, they are not scrutinised as strictly as formal judgments. The procedure for issuing these documents is also less formal. However, as mentioned earlier, these documents have the de facto power to influence both the behaviour of the parties involved and subsequent court rulings. This imbalance of significant influence and minimal oversight raises serious concerns.

Fourth, even in the absence of corruption, the very existence of relational governance may undermine judicial credibility. Studies on perceived justice highlight that judges’ conduct during legal proceedings significantly influences parties’ perceptions of trial fairness, often irrespective of the final decision outcome (Tyler, Reference Tyler1992, pp. 439–442). For example, if a judge inadvertently reveals familiarity with a party’s representative during a court session, the opposing party may perceive it as evidence of improper collusion, even if the interaction is entirely innocent and unrelated to any irregularity or malpractice.

6.1. Party capability theory

The party capability theory posits that, rather than a normative bias within the law or prejudiced attitudes among judges, the resources wielded by parties—such as significant disparities in access to legal counsel between the haves and have-nots—exert a more substantial influence on the outcome differentials between repeat players and one-shotters in the United States court system (Wheeler et al., Reference Wheeler, Cartwright, Kagan and Friedman1987, pp. 439–442). This theory underscores the idea that “repeat players,” who have the advantage of frequent interactions with the legal system, are often better equipped to navigate the complexities of legal processes. They have the financial means to hire experienced attorneys, invest in comprehensive legal strategies, and endure lengthy litigation processes, which gives them a significant edge over “one-shotters,” or those with limited encounters with the legal system.

To illustrate this point, Wheeler et al. have examined the success rates of larger businesses compared to smaller businesses, as well as large government organisations versus small government organisations. Studies have shown that larger businesses and government entities tend to win more often in court. For example, businesses with substantial legal and financial resources are better positioned to pursue litigation aggressively and can afford to hire specialised legal counsel. Similarly, large government organisations benefit from more extensive legal departments and greater familiarity with the legal system, which enhances their ability to achieve favourable outcomes in court.

Moreover, the disparities in legal outcomes are not merely a matter of case volume but also of strategic advantages. Repeat players, due to their continuous engagement with the legal system, develop expertise in managing legal risks and shaping favourable precedents. They are adept at selecting cases that are more likely to succeed and settling those that might set adverse legal precedents. This strategic case management further amplifies their advantages over one-shotters, who may lack the resources and experience to engage in such sophisticated legal manoeuvring.

6.2. Penetrated court theory

Xin He and Yang Su’s study suggests that the decisions rendered by Chinese courts are not impervious to external influences (He and Su, Reference He and Su2013, pp. 138–142). One prominent example of this influence is the judges’ reliance on local governments for appointments, promotions, and, until recently, funding and material security. This dependence creates a scenario where judges are more accountable to bureaucratic and political hierarchies than to the law and judicial ethics. This phenomenon is not unique to China; however, the degree to which local governments can influence judicial decisions through control over judges’ career paths highlights a significant divergence from systems where judicial independence is more rigorously protected.

Additionally, the study highlights the role of personal agreements and behind-the-scenes influences in shaping court decisions. These informal arrangements often involve government officials, businesses, and wealthy individuals, leveraging their connections and resources to sway judicial outcomes. For instance, government-related parties and firms consistently exhibit a remarkably high win rate in court cases, underscoring the permeable nature of the Chinese judicial system to external influences. This pattern suggests that the courts are not merely favouring the government due to resource disparities but are structurally aligned to serve governmental interests.

However, the study also notes that recent judicial reforms have aimed to mitigate the extent of governmental control and reduce judicial corruption. These reforms include efforts to improve the professional qualifications of judges and to restructure court financing to decrease reliance on local governments. While these changes have made some progress in curbing direct interference in court proceedings, the deeply entrenched institutional and social channels of influence remain a challenge. The reforms are a step towards greater judicial independence, but their effectiveness in fundamentally altering the power dynamics within the Chinese legal system is still evolving.

In cases involving big tech, the impact of party capacity and government penetration cannot be overlooked. However, their explanatory power is most evident in cases concerning unfair competition, platform monopolies, or market concentration. And as outlined previously, these issues relate to macroeconomic developments, where the court often lacks the requisite expertise or regulatory authority. Therefore, the court tends to accord full respect to the overall strategic directions made by the government concerning national and local economic and social development.

6.3. Relational governance theory

Given the high compliance rates and insights from the theory of relational governance, in most cases involving big tech (typically as defendants) and ordinary individuals (typically as plaintiffs)—such as those pertaining to privacy, child protection, identity verification, and standard contract terms—the victories of big tech largely stem from their incorporation of legal risk prevention mechanisms into their daily self-management practices. Consequently, most potential violations are pre-emptively addressed before reaching formal hearings. Contrary to traditional perceptions, within a relational governance framework, the court’s intervention in behaviour extends beyond ex post, case-by-case adjudicative decisions to encompass ex ante, systematic compliance advice. This shift reflects the adoption of “litigation source governance” (Suyuanzhili) within the Chinese court system and has proven more efficient in many cases (Chen, 2024, p. 49).

Based on the illustration above, we can view the efforts by repeat players to improve their litigation capacity from another perspective—they objectively reduce the cost of information exchange between the court and the parties, thus providing a solid foundation for relational adjudication. In other words, a strong litigation capability is accompanied by an equally strong compliance capability (Edelman and Suchman, Reference Edelman and Suchman1999, p. 943), and the litigation advantage of repeat players can become an important asset in the court’s relational governance.

For example, tech giants often hire former judges to serve as law clerks and frequently invite the courts to visit their companies. While this helps platform companies build their litigation advantage, it also fosters a valuable exchange of information and understanding between courts and companies. This circulation and exchange of personnel create opportunities for the court and the company to comprehend each other’s situations and needs better. A similar phenomenon can be observed among long-term business partners, with some original equipment manufacturers (OEMs) sending experts to their suppliers and increasing the exchange of personnel through regular gatherings (Bernstein, Reference Bernstein2015, p. 574).

Moreover, the inclusion of former judges can help tech companies with self-censorship and pre-compliance, which can also reduce the courts’ workload. If tech enterprises effectively conduct advanced compliance reviews of user agreements, improve identity authentication procedures, and establish robust internal dispute resolution mechanisms, the probability of lawsuits occurring will be significantly reduced. Additionally, tech giants’ active involvement in legal research and dissemination of the results can be seen as a way for tech companies to express their willingness to comply with the law to public authorities, including the courts, despite potentially biasing legal norms in their favour.

7.1. Emerging technology and institutional changes to the legal system

Using Internet courts in Beijing, Hangzhou, and Guangzhou as examples, these courts have demonstrated their systemic influence on platform business behaviour and the digital industry as a whole. Their influence extends beyond case-by-case adjudication, taking on roles similar to administrative regulation. For instance, in many information privacy disputes, courts have ruled that Platform A’s collection of a plaintiff’s information constituted a legal infringement and ordered the deletion of the relevant information. In theory, the direct impact of such rulings should be limited to Platform A deleting the specific plaintiff’s information. However, in practice, Platform A often chooses to stop collecting and deletes similar information from all users. Additionally, platforms B, C, and D, upon learning of the case’s outcome, are likely to adjust their information-handling practices accordingly. Consequently, a court’s judgment in one case can influence information processing norms across the entire industry.

How does this happen? We can explain it from two perspectives: why the courts undertake extra work such as issuing judicial suggestions, and how they manage to do so without de jure coercion. From an incentive perspective, Internet courts are willing to undertake “bonus moves” primarily because they face an enormous number of disputes. With the proliferation of new technologies and the commercial or organisational innovations they bring, many old norms have become inapplicable. This has led to unclear normative expectations and, consequently, a large number of disputes. For courts handling these cases, the imbalance between the high volume of cases and the limited number of judges is increasingly evident. For instance, in Court P, it is not unusual for a judge to handle five to eight cases a day, yet cases still overrun from time to time. Therefore, the need for courts to manage litigations systematically has become particularly urgent.

Despite incentive, the Internet court is also well-equipped to handle the task. First, Courts specialising in specific dispute types possess a nuanced understanding of industry dynamics, making their regulation both actionable and beneficial. Furthermore, unlike the legislature and the executive, the judiciary directly confronts various types of disputes arising in social life. This enables the judiciary to pinpoint critical normative gaps based on the volume of relevant cases, allowing for targeted intellectual investment and the rapid creation of corresponding behavioural norms. This approach enables the judiciary to promptly respond to the normative needs of the relevant parties.

Second, unlike formal norms, especially laws, which emphasise completeness and coherence, the judiciary’s case-by-case decisions and the behavioural norms they establish can focus on specific points where disputes are likely to arise or have already arisen. This ensures that these norms are urgently utilised and that priority is given to reconciling major contradictions. Moreover, formal norms, particularly legal systems, often take a long time to draft, revise, adopt, and implement. In contrast, the judiciary can respond much more quickly, making it better suited to adapt to the rapid iteration of new technologies and the issues they generate.

Third, flexibility stands out as another functional advantage of relational adjudication compared to other forms of governance. Emerging technologies and the organisational innovations they spawn are often fraught with uncertainty during their inception and early development stages. Effectively regulating such entities requires balancing the need to foster innovation while addressing potential negative impacts (Wu, Reference Wu2011, pp. 1848–1851). This necessitates governance measures that are not only binding in behaviour but also adaptable at the margins. In contrast, overly rigid regulatory approaches at the outset of intervention can stifle the growth of new industries due to incomplete information, potentially leading to costly adjustments later on. Relational adjudication generally allows more room for negotiation between parties and the court. While long-term relationships may exert a certain pressure on parties to conform, this pressure differs from normative coercion. For instance, in cases involving judicial suggestions, parties facing high implementation costs may opt to fulfil only the obligations mandated by court rulings while explaining their constraints in the response letter to the judicial suggestions.

Fourth, the leverage and influence of Internet courts primarily stem from their centralised jurisdiction over cases involving tech giants. Currently, China’s three Internet courts are located in Beijing, Hangzhou, and Guangzhou, all of which are home to major technology companies and their branches. This “platform-based jurisdiction” model means that each court, and even each trial team, handles numerous similar cases involving specific platform companies. This institutional and long-term interaction enables Internet courts to obtain information about the past compliance behaviour and litigation performance of tech giants at low costs, thus establishing a solid foundation for relational adjudication. Additionally, different tech companies often share similar business models, allowing courts to transfer the information and knowledge gained from one enterprise to better understand the operations of others.

Finally, the scarcity of Internet courts indirectly enhances their prominence in information dissemination networks. In the absence of a unified database of decisions or judicial suggestions, the influence of individual cases spreads through small-scale information dissemination. Three Internet courts undoubtedly occupy key positions within the information network composed of relevant legal counsels, various courts with jurisdiction over Internet-related cases, and other public authorities. The information they release gains significant attention and can be widely processed within the network. In my interviews, I found that some legal heads of tech giants have the habit of including the latest Internet court judgments in their dynamic weekly reports. Based on these judgments, they prompt the company to implement compliance measures or prepare for potential legal risks in advance. Additionally, the Internet court’s efforts in publicity and the model effect of tech giants have facilitated the diffusion of the Internet court’s relational governance of specific companies into a network-based governance of the digital industry as a whole.

7.2. Distinct roles of Chinese and US courts in social governance

The distinct roles assumed by Chinese and American courts in social governance help explain why the dynamics between courts and repeat players differ between the two countries. Generally speaking, courts in both China and the US serve a function of social governance. However, the difference lies in the explicit duty of Chinese courts to actively participate in social governance, whereas American courts engage in social governance more implicitly.

According to Mirjan Damaška, conflict-solving justice and policy-implementing justice represent two distinct approaches within legal adjudication. Conflict-solving justice focuses on resolving disputes between parties, with the state playing a reactive role as an arbitrator to ensure fairness and adherence to rules in a contested process. This model emphasises the neutral position of the judge and is prevalent in systems with traditionally limited state intervention. In contrast, policy-implementing justice is oriented towards enforcing state policies, reflecting a proactive state role where the legal process serves to implement governmental objectives. Judges in this model actively oversee adherence to state directives, often initiating investigations themselves. This approach is common in legal systems where the state takes a more managerial and interventionist role, aiming to shape societal behaviour according to broader governmental goals (Damaška, Reference Damaška1986, pp. 97–180).

As a country with the tradition of continental law, China’s judicial system has long been considered to adhere to policy-implementing justice. This conception is supported by former research on repeat players. For example, Xin He and Yang Su have found that where legal norms are unclear, the trade-offs between different policy objectives in the courts can affect the success rates of repeat players. Generally speaking, if the goal of economic development takes precedence, companies as repeat players are more likely to win. Conversely, if maintaining social stability is a more pressing concern, the outcome of litigation may favour one-shotters (He and Su, Reference He and Su2013, p. 140).

However, as noted earlier, the current role played by courts like Internet courts in China not only exceeds the scope of conflict-solving justice but also differs significantly from policy-implementing justice. Internet courts are actively shaping policy rather than merely implementing policies developed by the government or party committees. One frequently cited slogan of Internet courts is “establish rules through adjudication, promote governance through rules, and foster development through governance.” By handling numerous pioneering cases involving the application of emerging technologies, Internet courts are shaping game rules in the field of emerging technology. Through relational governance, these courts ensure that these rules not only apply to individual cases but also encourage tech companies to systematically adjust their business operations.

The institutional changes described above are not the result of deliberate design or rooted in legal tradition but rather stem from practical necessity. The rapid iteration of new technologies and business models continually generates numerous legal issues within normative gaps. Given this backdrop, the traditional slower process of receiving information, processing it, establishing rules, and applying them fails to meet the urgent demand for a large volume of rules in a short timeframe. Moreover, amidst disruptive innovation, legal issues not only emerge rapidly but also dissipate quickly. Consequently, it often happens that before formal regulations can be formulated or promulgated, the subjects they aim to regulate are already inundated by market dynamics. For instance, the introduction of Article 12 of the “Anti-Unfair Competition Law” in 2018 aimed to classify online unfair competition to reduce dependence on general provisions. However, due to the proliferation of non-standard forms of competition, the provision was swiftly rendered ineffective. Therefore, rules and policies must be produced and disseminated in a high-frequency interactive manner between regulators and the regulated. This underscores the crucial role of relational adjudication in governing emerging technologies and their applications.

As early as his 1974 paper, Galanter suggested that a more activist role for courts could potentially mitigate the influence of resource disparities on litigation outcomes. However, he promptly identified significant barriers to implementing such a solution within the American legal framework, including resistance from vested interests in the traditional litigation system and ideological critiques from jurists. Considering these challenges, Galanter placed his hopes on organisational reforms and a cultural shift within the legal profession. While the notion of a more activist judiciary may encounter obstacles in the US, it finds support in China, where a proactive judiciary is both a reality and a mandate. This distinctive legal culture and institutional framework provide insights into why the relationship between courts and repeat players differs from that in the US.