Abstract Drawing upon data interviews with Chinese judges who were involved in the decision-making process, we develop two variables for analyzing the influence of social ties, or guanxi, in the judicial setting. The first differentiates the strength of guanxi—whether it is strong or weak. The second distinguishes whether the guanxi source is from a supervisor who has a role in affecting the benefit or the career development of the target judge. We argue that instead of working independently, these two variables interact and often mutually reinforce. With this typology, we contrast the means and outcomes of four types of guanxi. This Article deepens scholarly understanding of the operation of guanxi in Chinese courts. Our framework explains why shady practices that fuel favoritism and undermine the development of legal professionalism are difficult to pin down and stamp out. Everywhere we are embedded in guanxi. Indeed, guanxi has become part of our lives. We make use of guanxi; as a chain of the network, we are also being used by others in guanxi. A Trial Judge from Hinterland China Introduction Guanxi (loosely translated as “social connections” or “ties”) plays a significant role in Chinese society. In his classic, From the Soil, Fei Xiaotong suggests that social ties are crucial to the organization of Chinese society.1 If, as Rousseau suggests, “man is born free and everywhere he is in chains,”2 for the Chinese those “chains” are most manifested in guanxi. Recent scholarship only confirms a prominent role of guanxi in transitional China.3 In cadres’ promotion exercises, it is said that “youth is precious, capability is for reference, guanxi is crucial” (or “nianlin shigebao, nengli zuocankao, guanxi zuizhongyao” in Chinese).4 Junqi Feng suggests that the formal rules work only when being pushed by guanxi.5 In one survey, 52% of judges said that in handling cases they were frequently approached by guanxi, and 61% said that within the scope of discretion, they favored the connected parties.6 Yet, as is often the case with studies of social ties elsewhere,7 researchers of Chinese guanxi tend to rely on data from favor seekers—job seekers, employees, and entrepreneurs—rather than favor givers. Understandably, few favor givers are willing to share their experiences with researchers. This leads to what Doug Guthrie describes as a problem of “information asymmetry”: while favor seekers attempt to use connections to gain an advantage, they do not know the extent to which the connections pulled affect the outcome.8 One consequence of such an approach is that the mechanism of how guanxi works inside the government remains understudied. Most guanxi studies focus on the development of the market and informal economies,9 gift exchanges,10 or job searches,11 where data are more readily available. This asymmetry is particularly acute in the existing studies about the legal system. Chinese judges do not live in a vacuum. In courts, they are judges, but the moment they take their robes off, they are husbands, wives, colleagues, supervisors, subordinates, classmates, and relatives. As illustrated by the epigraph to this Article, judges are ensnared in all sorts of guanxi, which in turn influences their judicial decisions. This coexistence of identities is particularly consequential in China because of the nascent status of the legal professional: the professional identity of a judge does not take precedence over other social identities. Few studies address the role of guanxi in the Chinese legal system. They predominantly rely on reports from disputants who alleged that their adversaries had insider ties, but the disputants were not privy to the decision-making process. For example, Michelson and Read’s survey reported that Chinese litigants often lament the edge of their adversaries with insider guanxi in exploiting procedural loopholes and unduly influencing outcomes.12 Ling Li, in her otherwise provocative argument, relies on secondhand data, such as media reports and a political novel, to demonstrate how lawyers and auctioneers have tried to build up guanxi with judges.13 Clearly, the problem of information asymmetry has put a veil over the impact of guanxi on the operation of the Chinese legal system, in which its institutional contexts are different from the market.14 As a result, while scholars generally see guanxi as an obstacle to the country’s efforts toward the rule of law,15 and as a facilitator to corruption,16 little is known about the mechanisms behind how it operates. One exception is Xiaoqiang Wang.17 A judge himself, Wang divides guanxi into three categories: emotional, mixed, and instrumental, with emotional guanxi being the closest between the guanxi source and the guanxi target.18 He argues that the three types of guanxi have varying impacts on judicial decisions. He also points out that guanxi will influence judges through their supervisors and colleagues inside and outside the courts.19 While Wang’s ideas are fascinating, he does not provide empirical evidence to illustrate the processes that judicial decisions are affected by. In this Article, we draw upon data to which existing scholars have hitherto had limited access—interviews with judges who were involved in the decision-making process. Echoing the findings that strong ties are crucial in illegal or socially disapproved transactions,20 we argue that strong guanxi alter case outcomes in China. We further argue that due to the hierarchical nature of Chinese courts, supervisory guanxi, a variable overlooked by Yanjie Bian, is no less important. If trust usually comes from strong guanxi characterized by intimacy and obligation, then pressure often comes from supervisory guanxi characterized by the hierarchical patron–client relationship. We develop two axles analyzing the influence of guanxi in the judicial setting. The first differentiates the strength of guanxi—whether it is strong or weak. The second distinguishes whether the guanxi source is from a supervisor who has a role in affecting the benefit or the career development of the target judge. We argue that instead of working independently, these two axles often interact. With this typology, we contrast the means and outcomes of four types of guanxi. Our goal is to identify and differentiate the effects of the four types of guanxi on judicial outcome. As will be shown, this framework allows us to address the intricate, and at times counterintuitive, relationship between guanxi and bribery. This Article deepens scholarly understanding of the operation of guanxi in Chinese courts. In so doing, we view the courts as essentially no different from other Chinese bureaucracies. After all, the hierarchical nature of the courts, which shapes the supervisory dimension of guanxi, is shared by other types of Chinese bureaucracies. Our synthetic framework explains why shady practices that fuel favoritism and undermine the development of legal professionalism are difficult to pin down and stamp out. I. Two Variables of Guanxi As Thomas Gold and colleagues characterize, there are two prevailing views on guanxi in the literature.21 The first, the institutionalist perspective, sees guanxi as the structural consequence of an authoritarian state filled with powerful officials in a society with weak internal control. It treats guanxi as a strategic resource and a form of soft power in the institutional context. Specifically, it argues that guanxi is instrumental for navigating institutionally uncertain environments in the reform economy.22 Exemplifying this institutionalist thesis, Andrew Walder describes the relationship between superiors and subordinates in work units as a form of patron–client relationship.23 In a centralized command economy, guanxi became important when powerful officials were in full control of scarce necessities such as housing, job openings, and other nonwage benefits. The second view, commonly known as the cultural thesis, sees guanxi as a broad, diffused, and deep-rooted phenomenon of Chinese society. According to Fei, Confucian ethics are a form of tiered social ethics: familial relationships, long-term friends, classmates, schoolmates, in-laws, colleagues, business partners, and acquaintances.24 Society is composed not of discrete individuals and organizations but of overlapping networks of people linked together through differentially categorized social relationships.25 Each individual is at the center of an egocentric network with no explicit boundaries, always involving guanxi of varying magnitudes. Applying the cultural thesis to his analysis of judicial guanxi, Wang suggests that judges are surrounded by “the emotional, mixed, and instrumental.”26 “The emotional” is closest to the center, satisfying emotional needs, such as love, care, security, and belonging. “The instrumental” is farthest from the center, following the rule of business transactions, while “the mixed” lies in between.27 Furthering the cultural thesis, both Bian and Yang study how individuals cultivate and utilize their guanxi to advance personal interests.28 In return, those benefiting from guanxi are obligated to assist those connected.29 For example, in assessing the relative efficacy of network ties in job searches, Bian argues the need for “bringing strong ties back in.”30 In debating the nature of guanxi, the two theses are often presented as oppositional—guanxi is either institutional or cultural.31 This precludes the theoretical strategy of treating the two as analytically distinct dimensions that differentiate types of guanxi. In this Article, we treat the two distinct dimensions as superimposed on each other. We develop two analytical axles. First, the nature of guanxi is in part determined by the institutional positions occupied by favor seeker and favor giver. This first axle refers to the degree of direct control or supervision a favor seeker has over the favor giver. We ask whether a connection comes from the supervisors of the judge who decides, and whether ties can be classified as supervisory and nonsupervisory. The Chinese court is a hierarchical institution. A judge’s income and promotions, two key components in Chinese judges’ utility functions,32 are to a large extent determined by her supervisors. A judge is under strong pressure to deliver if she is approached by someone who is in a supervisory position; in other words, that person has direct control over her in the bureaucracy.33 Such pressure is weakened if the favor seeker does not occupy such a position. The term “supervisor” goes beyond the immediate boss of the judge; it includes any officials within the judicial system who are in a position to assess the performance of the judge and thus affect her welfare and career prospects. Supervisors thus include division directors or presidents of the court. Supervisors can also include senior judges from the corresponding appellate court. These senior judges determine if a lower-court judge’s decision is reversed or remanded, either of which is a strike against her performance record. Supervisors can also be local political leaders who influence appointments to the court on which the judge works. Local administrative leaders and Party officials are responsible for the appointment and removal of court officials. Local officials in charge of powerful bureaus in the economic sector and the People’s Congresses also play “supervisory” roles as they have de facto control over the court budget and staffing decisions. In Wang’s words, they indirectly share in the judicial decision-making power.34 Favor seekers are nonsupervisory when they do not have a direct say in the judge’s performance and career development. They include court colleagues of the same, or lower, rank, as well as judges from neighboring courts who are not in a supervisory position in the court. Nonsupervisory guanxi may also involve a judge’s classmates or schoolmates, or her former colleagues. They may also include her relatives or friends. As we will show, whether a relationship is supervisory makes an important difference in influencing the decision-making process. The second axle gauges the strength of guanxi: whether a social tie is strong or weak.35 A tie between two individuals can be classified as strong or weak, depending on the time spent in interaction, emotional intensity, intimacy, and reciprocal services.36 Strong guanxi is also multiplex. Multiplex ties exist in multiple contexts,37 and are more likely to extend to a new setting, playing the role of a “bridge.”38 Analytically, multiplexity can be operationalized as the degree of overlap between roles, exchanges, or affiliations in a social relationship.39Guanxi that appears in many settings is harder to resist, and is likely to extend its effects to the new setting. For instance, judges have many colleagues but only with those in their inner circles will they spend time together over the weekend or share family vacations. Only with those who are in her personal circle will a judge share her experiences and opinions on difficult cases. This overlaps with what Wang called “the emotional.”40 The same applies to the relations between a judge and her friends. Some friends are close and some are casual. The relations between a judge and his family members are usually strong because familial ties straddle different settings. This is the innermost circle of Fei’s original thesis of differential modes of association. The relation of a judge with his former classmate may also be strong if they see each other frequently and assist each other. On the other hand, uniplex guanxi is weak and often superficial: the ties do not extend beyond that setting. Of course, the line differentiating what is close and what is superficial is fluid. Our approach nonetheless provides a framework to examine the impact of guanxi (Table 1). Table 1. A Typology of Guanxi and Its Influence on Judicial Decision Making. Strong tie Weak tie Supervisory Enormous impact, a judge’s help may go beyond the scope of discretion. Neither money nor gift is offered. Significant to moderate impact, judges help by exercising discretionary power. Nonmonetary gift or no gift is offered. Nonsupervisory Significant impact, judge’s help usually within the scope of discretion. Nonmonetary gift offered. Uncertain impact, judge’s help, if any, is within the scope of discretion. Money and gift offered. Strong tie Weak tie Supervisory Enormous impact, a judge’s help may go beyond the scope of discretion. Neither money nor gift is offered. Significant to moderate impact, judges help by exercising discretionary power. Nonmonetary gift or no gift is offered. Nonsupervisory Significant impact, judge’s help usually within the scope of discretion. Nonmonetary gift offered. Uncertain impact, judge’s help, if any, is within the scope of discretion. Money and gift offered. View Large Influence is most intense when guanxi comes from a relation that is both supervisory and strong. The relationship between the favor seeker (higher) and the favor giver (lower) is both formally hierarchical and multiplex. Walder has attributed this to a form of patron–client relation that mingles personal loyalties, institutional assessment on performance, and material interest. The patron offers not only advantages of official bonuses, raises, and promotions to their clients, but also informal “perks,” such as better housing allocations and scarce commodities unavailable on the market. Under this circumstance, the target judge has to deliver for her supervisor to show that she is both capable and loyal. She regards the matter raised by her supervisor as her own business.41 The influence subsides when the connection is supervisory but weak. This kind of guanxi can be found between a supervisor and a subordinate in more impersonal types of bureaucracies. This guanxi is more formal and narrower. It does not span a broad range of settings. The target judge still works to show loyalty to the supervisor, but she does not go out of her way to get the job done. She may not want to risk her career, and is hesitant to go against the law in order to satisfy her supervisor. The third type of guanxi is strong but nonsupervisory. Since it is nonsupervisory, the target judge does not worry about how the favor seeker can impact her career. But the guanxi is strong, thus it calls upon the judge’s obligations in other social capacities, as son or daughter, sibling, relative, or friend. Compared with the first type of guanxi, this type of nonsupervisory but strong tie also has significant influence on the favor giver. However, the extent of the influence seldom goes beyond the scope of her discretionary power. As the guanxi is not supervisory, her supervisors are not “in” on the exercise. To protect herself, the judge who offers favor much prefers to work within the confines of the law. Usually no immediate favor is needed. The influence is weakest when guanxi is nonsupervisory and weak. This is however the most common form of guanxi. This type of guanxi lies on the periphery of a judge’s egocentric network; ties of this type are so numerous that they are easily replaceable. This type of guanxi is also uniplex. It does not straddle across settings. The potential favor giver is reluctant to bring weak guanxi to her job setting. Furthermore, since this type of guanxi is nonsupervisory, the target judge is not worried about the feelings of the favor seeker. The target judge has little incentive to make efforts, not to mention breaking the law. She may only reply to questions that the other party could also get the answers to from official websites or published documents, or offer a “personal take” on what the party could do, without promising anything. In short, she only provides “information” without exerting “influence.”42 II. Data and Methods In this Part, we examine the operation of guanxi from the understudied perspective of the favor giver—the judges. In fact, our interest in studying guanxi cases developed gradually as we worked on related topics on Chinese courts including domestic violence, courtroom discourse, mediation, and criminal reconciliation in the past decade. While we were researching these topics, judges sometimes brought to our attention cases in which their decisions were clearly influenced by guanxi. We found these cases fascinating since they offered a rare glimpse into the workings of guanxi from the perspective of the favor giver. Our compilation of these cases steadily grew, to the point where we thought we could study the topic more systematically. To achieve this goal, we reached out to these judges again to further ask them about the strength and nature of the guanxi that were at play in the cases that they had previously mentioned to us. Some of them also mentioned new cases to help our understanding of guanxi. We further generated a broader set of cases by snowball sampling, that is, by asking judges to refer us to their colleagues who have encountered similar cases. Because of our years of fieldwork, we have developed a high level of trust with the judges who shared with us the cases discussed below. The judges were aware that our purpose is limited to academic research, and they were willing to share their stories with us. Specifically, we asked these judges how they deal with renqing (favor), guanxi, and money (jinqian) cases—the so-called three types of cases (san’an) in the lingo of the Chinese judiciary. We focused on cases in which the judges were centrally involved, usually as the responsible judge (chengbanfaguan), or cases in which they participated in making the decision, since in those cases they were in the best position to understand the impact of guanxi. We paid special attention to the character of the guanxi involved, the nature of the requests from their guanxi sources, and the means by which requests were conveyed. We also paid attention to the relationship between the guanxi source and the target judges, the responses of the target judges, the outcomes of the efforts, and the means of compensation (gifts, banquets, money, or other nonmonetary compensation). We asked the judges to evaluate each type of guanxi in their career development and the risk entailed. We also asked what protective measures they had taken in fulfilling such requests, and the distinction, if any, made between guanxi and corruption. We asked how the judges viewed their supervisors and how they characterized their relationships with them. Additionally, we asked them to engage in several counterfactual thinking exercises; for example, we asked how they would handle certain cases if no guanxi were involved. Questions we asked judges include: “What was your relationship with the person who asked you for a favor?”; “Did you feel pressure?”; “Did guanxi affect the outcome of the ruling?”; “Why did you turn down the request (in cases where guanxi had little influence)?”; “Was there any favor given in return?”; and “Was money involved?” In the actual interviews, we did not follow a rigid format to go through all questions one by one. This would be too interrogative, given the sensitive nature of the topic. We tried to keep the interview as open-ended as possible. We also tried, as much as possible, to let the judges speak. The judges generally touched on many of the abovementioned questions in their accounts. In cases where some of the important details were left unmentioned, we asked follow-up questions to make sure their accounts were comprehensive and sufficiently detailed for our purpose. With a few exceptions, judges we spoke with came to know the cases from discussions in division meetings or in private conversations, or from both. The majority of the judges that we interviewed were frontline judges. A few of them were deputy division heads and division heads. In total, we talked to about twenty judges in depth on the subject. We also had shorter discussions with another twenty judges on matters related to guanxi. A few of them in the latter group were reluctant to recount specific cases. But all provided important background information. In view of the sensitivity of the topic, we did not resort to elaborate sampling techniques. We generated our sample of cases by means of referral sampling. We also found that judges were more amenable to discussing guanxi cases in casual settings. Our interviews took place in restaurants, canteens, the judges’ own offices, and occasionally even cars we rode in with them during the course of our fieldwork. In short, our reports were from insiders with whom we developed a relationship of trust. We did not, and could not, observe directly the dealings that influenced the outcomes of the cases we reported below. Since few judges would make up these stories and their accounts were more consistent than not, we believe our data presents a fairly accurate picture of how guanxi works inside the courts. We did verify reliability by asking the same question (or the same question in a slightly rephrased form) at different times during an interview to make sense of the judges’ accounts. Most cases used in this Article come from two basic level courts, one located in the rural region of an inland province, where the GDP per capita is only one-third of the developed coastal areas; the other located in one of the most affluent areas of the country. The locations of the courts allow us to identify the effects of the market economy on the judicial process. Towards the end of the Article, we seek to compare the main cases with interviews of judges from courts in other parts of the country. III. The Impact of Guanxi A. Supervisory and Strong Few judges are immune to the influence of guanxi when it is supervisory and salient. In many rural courts, the court leaders who occupy supervisory positions retain broad and substantial power. They vet decisions proposed by junior judges. More important, in one way or another, they determine the benefits, positions, and career development of a judge. A veteran judge who worked in the civil division of a grassroots court in an inland province recalled the following case. A woman was injured at work at a big brick manufacturer. Workers there had been used to heating up drinking water by putting a kettle on the top of the kiln. One day, when the woman tried to climb up the kiln to heat up another pot of water, the top of the kiln collapsed and she fell inside. She was pulled out of the kiln immediately, but a large area of her body was burned. The accident left her in a critical condition. After a long series of treatments and multiple skin grafts, she survived. Her medical expenses amounted to 190,000 yuan, an enormous sum by the local living standard. She filed a tort suit against the brick manufacturer for compensation (Case of the Collapsed Kiln). For five years, the court did not render a judgment. The brick manufacturer was one of the biggest enterprises in the region. The owner was a delegate in the local People’s Congress. His relationship with local officials was well known: when the local government had claimed the manufacturer’s land for redevelopment, it was rumored that the manufacturer had been overcompensated by millions of yuan. The owner had previously had a close relationship with the vice president of the trial court as well as with the senior officials in the region’s intermediate court. The ties were so strong that the responsible judge was asked to directly report the case to the vice president. Counsel for the manufacturer requested to include more defendants. They argued that the manufacturer had contracted the construction of the kiln to a work team, and thus should not be held liable for the injury. In informal meetings with the judges on the case, the vice president would not approve of any rulings that held the manufacturer liable for damages. On the eve of the 2013 Spring Festival, the responsible judge and her division director proposed that the manufacturer prepay 10,000 yuan to the victim as a temporary relief, a small fraction of the medical expenses incurred. The vice president was angry. In the presence of the manufacturer’s owner, he scolded the responsible judge and the division director. The vice president did hint during the meeting that the owner could offer some help to make the case “go away,” but added that it would be at his complete discretion. Over the five years, the injured family received only 5,000 yuan in compensation from the manufacturer. Emboldened by his guanxi with the vice president, the manufacturer’s owner refused to settle. Matters did not change until five years later, when both the vice president and the president of the court were replaced. The relationship between the vice president and the responsible judge was not only supervisory, but also strong. The vice president treated the responsible judge as his protégé, a relationship reminiscent of the “patron–client” type.43 This explains why the vice president admonished her in front of the litigation party. The vice president expected the junior judge to take care of his business as if it were her own. The responsible judge tried to protect the interests of the owner, but the expectation of the owner exceeded what the law would allow: he did not want to compensate a penny. The pressure on the judge was so intense that she had little means to resist. As it turned out, all the requests of the manufacturer, legal or not, were satisfied. The strength of a supervisory tie is decided more by the institutional makeup of a local court than by the personal relationship between the superior and the subordinate judges. Supervisors who hold broad power foster strong guanxi. They can reward the protégé, if they are pleased, or punish them, if they feel offended. Thus, the subordinates are eager to attend to their supervisors’ needs. “Helping out” a supervisor is a common way for the supervisee to demonstrate her loyalty and therefore strengthen her tie to the patron. The following is another case told by our informant whose colleague was the responsible judge. In a small, low-stakes civil case that took place in a hinterland province, the plaintiff claimed 3,000 yuan in compensation for his stolen motorcycle. He alleged that the bike had been stolen from a parking lot in a residential complex. But he could not produce receipts to prove that he had parked the motorcycle in the lot at the time when the alleged theft had occurred. What he had, however, was a handwritten statement from the management office of the residential complex that the motorcycle was missing. To make matters more interesting, at the hearing, a staff member from the management testified that the statement was fake (Case of the Stolen Motorcycle). This case proceeded just like other routine cases until the responsible judge sought approval for his draft judgment from the court president. The president returned the draft with a terse comment: “This case shall be handled carefully.” As it turned out, the plaintiff was a relative of the political leader in charge of political-legal affairs at the municipal level. This leader was thus the immediate supervisor of the court president, because structurally, a local court in China is no more than a division of the local government. While the judge was unsure about how to proceed with the case, he was summoned to accompany the court president to report it to the political leader. Case reporting occurs for senior local Party state officials in order to coordinate major, influential cases in the region. The judge was surprised that his small case had been selected for reporting. From a legal standpoint, it was a simple contract case in which the plaintiff had failed to produce the evidence in dispute. When the judge shared the case with us, he could not repress his contempt for the political leader and the court president: If I were the leader, I would rather commit suicide. A leader at the municipal level, and the one in charge of political-legal affairs, he summoned the court president and the responsible judge to report on such a trivial case! The sole purpose was to inform us of his expectation. In my eyes, he was worth little more than the 3,000 yuan! The judge continued: “Now I know that it is not easy to be a court president. He looks like a king inside the court. But he was just a nobody in front of the municipal leader. I can’t really describe how my president was chewed out by that leader.” The pressure put on the court president was nonetheless passed on to the judge. Within days, he had a new judgment ready to be endorsed by the president: the defendant would compensate 3,000 yuan for the loss of the motorcycle. In the judgment, the statement of the management office was admitted as authentic, and its rebuttal was ruled invalid for an overtly technical reason: the thirty-day limitation for the submission of evidence had passed when the oral evidence was filed. The judge seemed torn between satisfying his superior and disdaining himself for the judgment. Even though the judge was not keen on promotion, he said: “The court president obviously needed my help when he asked me to report the case to him. I still need this job, and my daughter needs my salary to finish her university degree. What else can I do?” In this case, the judge and his court president acquiesced to the request of their political patron without putting up much resistance. The judge admitted his ruling was weak from a legal standpoint. Yet, that was the ruling he made in the face of the pressure from this strong institutional tie. Another judge commented: “For this kind of case, we create a legal opening not available under normal circumstances.” Through problematic cases like this, ironically, senior court officials and their local political leaders nurture their ties. The political leader owed the court president a favor, and the two grew closer as a result. When they meet next time, they will have more topics to discuss, a situation long desired by the court president. If economic shortages and weak legal infrastructure were the main reasons why guanxi became widespread in the early stage of the reform,44 the influence of guanxi may have declined among the courts located in coastal cities. Legal infrastructures there are more developed. Judges who work in the coastal urban areas are also subjected to greater scrutiny both from within the Party state and from the media and the public. To some extent our data verifies this. We find that in developed areas, both the supervisors and the judges are more cautious in dealing with cases with guanxi. Nonetheless, the influence of supervisory and strong ties remain substantial. A judge in her early thirties told us about the following appellate case that she had handled. Four defendants had been charged with selling pirated cellphones worth more than one million yuan. The trial court determined that a girl in her early twenties had assumed the leading role and sentenced her to three years in jail. The other three defendants were given only suspended sentences for their subordinate roles. The girl was found to have been in charge of the financial transactions and marketing arrangements of the operation. However, one defendant testified that “she is not the boss, but only a wage earner.” The responsible judge—our informant—found from the court hearing that the girl was in charge of the operation as a proxy for the real orchestrator of the crime. However, she refused to disclose his identity. She was the leader among the four. The responsible judge’s division director, connected to the girl, instructed the responsible judge to “consider the case carefully, and to display not just legal skills but also skills in social relationships.” The director had been a mentor to the responsible judge for many years. The responsible judge thus always appreciated the director’s help. Under normal court practices, unless there are defendants at large, the court must designate a leading defendant. In this case, no defendants were at large and none had taken a more prominent role than the girl. When the responsible judge communicated this concern to the director, the director said, “Why not? Shouldn’t doubtful cases be treated leniently?” The responsible judge, however, was reluctant to exonerate the girl from the leading role. If the girl had not been regarded as the leader, then no defendant would have been the leader. As a result, the sentence of the other three defendants would have been elevated to three years in jail, the minimum penalty according to the law. Such an elevation would have violated the principle of “no penalty added in appeal.” After conducting numerous studies and consulting several judges specializing in criminal trials, the judge decided to keep the girl in the leading role, but changed the three years’ imprisonment to a suspended sentence. In order to justify the change, she requested that the girl pay an increased fine, and her counsel did so promptly. When the responsible judge proposed the decision to the director, the director was upset because asking the defendant to pay a higher fine before a formal decision was legally inappropriate. In a routine meeting of the division, the director unexpectedly requested that the responsible judge and her collegiate panel deliberate the case in front of all division members. This was to show that the decision was made transparently. The above case suggests that judges in developed areas are more cautious in handling guanxi-related cases than their hinterland counterparts. The responsible judge did not blindly follow the director’s suggestion of leniency toward controversial cases. To her, the facts indicated that the girl had taken a leading role. On the other hand, the director was also prudent. He did not want his influence on this case to be traced by anyone, so he requested that the collegiate panel deliberate in the presence of the division members to show “due process.” Judging from the outcome, however, the guanxi from the director—supervisory and strong—still affected the responsible judge and determined the outcome. Immediate supervisory relationships remain strong in Chinese courts. One phone call from an immediate supervisor is usually the most impactful to the responsible judge. One interviewee said: If a judge of the Supreme People’s Court (SPC) wants to influence a case at a basic level court, she would not directly contact the responsible judge. Instead, the SPC judge would contact a judge at the high court, who will then contact another judge at the intermediate court, who will eventually contact the responsible judge. In this way, the remote source influences a trial judge through a chain of immediate supervisory guanxi. The chain is so long that the responsible judge may have no idea from where the guanxi source is. This holds true within the hierarchy of a court. A Guangdong judge said, “Court presidents rarely instruct the frontline judges what to do, unless their relationship is close. Rather, they first contact the division director, who will then contact the frontline judge.” This explains why the reach of strong guanxi is long and can overcome umpteenth degrees of separation in China. Favor seekers and their targets, originally disconnected, are bridged through intermediaries to whom both are strongly tied.45 B. Supervisory but Weak By comparison, the scope of weak ties is not as encompassing as that of the strong ones. When guanxi comes from a person in a supervisory position but is weak, the influence becomes moderate when compared with strong and supervisory guanxi. A weak tie is distinguished from a strong tie for its delimited, uniplex nature, i.e., the influence of the person comes more from her office than from her person. At the very least, the target judge needs not to offend the supervisor. That is why the target judge usually will give “face” to her supervisor. One judge said, “I would take it seriously but make sure that my handling follows the law.” A middle-aged Guangdong judge told us that she had been called by a director of another division to help withdraw a case within one day. One of the involved companies had been scheduled to be reviewed by the securities regulatory authorities the following day. This was the last stage before its public offering. It thus needed to clear all pending litigation. Normally, such requests are not approved, since a claim filed can only be withdrawn after a certain period of time. But an early withdrawal is allowed if the court has already undertaken some preliminary examination by taking testimony from the plaintiff. The director therefore suggested to the judge that she take testimony from the company. From a legal perspective, the judge could have just ignored the request, since it was the company that first filed the claim and she had other urgent matters at hand. The judge was not a close friend of the director—the latter had recently been transferred from another court. Nonetheless, it was possible that he could become her immediate supervisor some day because of internal reassignments. The judge set aside all work and conducted the preliminary examination and heard testimony from the company in the morning. She then quickly drafted a decision allowing the company to withdraw and presented it to her supervisor in the afternoon. But she did not inform her supervisor of the call from the division director. She said to us: “I did not want to mention anything. I did not want my supervisor to suspect that I had any interest in it, which I did not. If he approved, that was his decision. If he did not, it would be out of my control.” Luckily, her supervisor was available that day and the decision was issued as requested. The other division director was evidently pleased. The help offered by the judge was procedural: the withdrawal would have been issued without the judge’s help, but only later. The judge helped by doing what was required for an early withdrawal. She set her regular work aside, but her efforts were limited in that she did not urge her supervisor to approve the withdrawal. Frontline judges rarely turn down requests from their seniors. The hierarchical and administrative nature of the courts empowers supervisors who do not have a strong personal relationship with the target. But the intensity of influence subsides when compared with the situation in strong and supervisory guanxi. In some cases we even found resistance from the target judge facing pressure from supervisory, but weak, guanxi. In a trademark transfer case in a coastal city where our informant was on the collegiate panel, the two litigation parties were associated. Prior to the date of the hearing, the parties presented a settlement application, stating that the defendants would pay the 20 million yuan as requested. The judges on the collegiate panel found this litigation suspicious, and questioned if it was a phony litigation—the genuine intention of the parties being to transfer their properties. The judges decided not to approve the settlement application. Before the judgment was announced however, the plaintiff filed a motion to withdraw the case. This move only confirmed the suspicion of the judges that the litigation was phony. However, the motion was accompanied by a request from a vice president of the court, with whom the judge had only a lukewarm relationship. The president wanted her to “carefully consider the withdrawal application.” The responsible judge then found herself in a dilemma: she wanted to deny the withdrawal application and rule on the case; however, the vice president had spoken. Eventually, she employed subtle tactics, which, according to our informant, had been widely used to deflect guanxi requests in her court: the members of the panel were not unanimous. The responsible judge did not follow the opinion suggested by the vice president, but the other members of the collegiate panel held a different but majority opinion. Since a nonunanimous decision would need to be approved by the division director and vice president, the judges’ “agreement to disagree” shrewdly protected them if a formal investigation were conducted—the responsible judge followed the rules, rather than the suggestion of the vice president. But the collegiate panel still approved the withdrawal application. The influence of a supervisory tie may be further discounted if the target judge has little incentive to be promoted. In a case in which a married woman sued her village committee for land compensations, the issue was whether she was indeed “married”—she had not obtained a marriage certificate, but had hosted wedding banquets for the relatives of the two families, a ceremony regarded by the locals as equivalent to a marriage ceremony. The vice chief of staff at the court, related to the married woman, had sought out the responsible judge for help. The vice chief of staff had not been a direct supervisor of the judge, nor had his post provided him with any substantial power over the judge. However, the vice chief of staff could be transferred to a position superior to the judge. The responsible judge, a colleague of our informant, could have supported the plaintiff had she collected evidence such as proof that no marriage certificate had been issued or of the amount of money distributed to other villagers. But the vice chief of staff did not provide the responsible judge with material benefits, as had been expected. Moreover, at age fifty-two, the responsible judge was to retire in three years. Without any desire to be promoted or transferred to a better position, she did not offer to help. Instead, she requested that the married woman withdraw the case. C. Nonsupervisory but Strong In a relational society such as China, strong ties are not derived solely from people in supervisory roles. They also come from colleagues, friends, and family members. Nonsupervisory, strong ties hold sway over the decisions of Chinese judges. If supervisory ties represent vertical influences from within the bureaucracy of the court, then strong but nonsupervisory ties represent the horizontal pushing and pulling from outside. Judges are government bureaucrats, but they are also members of what the Chinese call a society of acquaintances. For judges working in rural courts, even though they can buy a lot of previously scarce items on the market, personal networks and the practice of favors and reciprocity remain crucial in getting things done—getting a bank loan or sending children to preferred schools being prime examples. Many judges feel obligated by the strong ties of their personal networks. Judges usually make an effort to help the person who pulls strings, by either repaying a favor owed or bestowing a favor to be reciprocated later (to put someone in renqing debt). What distinguishes nonsupervisory ties from strong, supervisory ties is that the judge has to act without the help, and sometimes the consent, of her superiors. It would be harder and riskier, however, for the judge to engage in illegal activities, since she is subject to the scrutiny of her superiors. Instead, the modus operandi is to maneuver in the grey area of the discretionary power that the judge holds. In a traffic accident case, a man in his mid-seventies crossing the road hastily was hit by a pickup truck (Case of the Jaywalking Man). The traffic police determined that both parties had been negligent, but the pickup truck driver bore major responsibility while the man’s jaywalking act bore minor responsibility. The man’s daughter, the director of an administrative bureau in the region, soon contacted her high school classmate—a middle-ranking judge who shared the case with us. Their relationship was close, as they had kept in touch over the two decades since graduation. Upon being asked by her friend, the judge contacted one of her closest colleagues who was in charge of the initial filings of new cases. This colleague offered astute advice. For example, originally, the plaintiff had intended to file a request to impound the pickup truck. The judge advised the plaintiff to directly file a lawsuit, instead. This would save the man a 3,000 yuan impounding fee. The court formally accepted the case in the afternoon of the day the case was filed. The acceptance decision was made while the plaintiff was still submitting other required filing documents. In addition, the filing judge also “randomly” assigned the case to a judge with whom our interviewed judge had a close relationship. That judge started working on the case as soon as it arrived on her desk, whereas normally it took several days to prepare for a case before a hearing was scheduled. The plaintiff did not hire a lawyer. As the judge indicated to us, it was unnecessary, since the responsible judge had offered detailed legal advice throughout the process. At the end of the trial, the judge ruled in favor of the plaintiff. The judge worked within the rules to deliver a favorable verdict for the plaintiff. The judge awarded the plaintiff 2,000 yuan to cover expenses for consulting medical experts, but this kind of fee is usually excluded from the calculation of damages under Chinese law. Yet, she included it under the heading of “transportation fees.” An item of 3,000 yuan for outpatient medicine costs, again normally inadmissible, was included as “nutritional fees.” The judge also generously included the cost of hiring two helpers to tend to the injured man, while normally the court would allow for the cost of only one helper. The difference of allowing for an extra helper, a discretionary choice that the judge was allowed to make, was more than 3,000 yuan. The judge found the man 20% at fault for his own injury. This, again, deviated from the usual practice whereby a 30% liability is assigned to the jaywalker. Our informant said that the judge had originally allocated only 10% to the man, but eventually changed her mind. A 10% liability for the minor party in a comparable negligence case would have been too exceptional, and would have aroused the attention of the division director. To get the extra 10% in damages would require the tacit approval of the division director, who was only weakly tied to our informant. From the way the judge talked about the case, its outcome was much affected by the strings that she had pulled. The plaintiff received a favorable verdict, given the routine way similar cases had been dealt with. The case was expeditiously processed. Some procedural requirements were generously forsaken, as the trial was scheduled before all documents were formally submitted. The plaintiff had also saved time and money by opting not to impound the defendant’s truck. This had been done based on the advice offered by the responsible judge. Most importantly, the damages had been generously calculated, an area in which the judge exercised much of her discretion. The encompassing effects of strong outside ties show the porousness of Chinese courts as an institution in two ways. First, judges do not set a strong boundary separating work duty from social duty. Second, judges are also able to exercise so much discretion that, to an important extent, it can influence the substantive outcome of adjudication. In general, having close guanxi is crucial, even when all actions take place within the confines of the law. The judge advises the connected litigation party on what key evidence to produce. If the connected party is unable to produce the required evidence, the judge may even exercise her discretion to collect evidence for her. The judge may also evaluate the evidence of the other party and may suggest a good rebuttal strategy for the connected party. In the Case of the Jaywalking Man, for example, the plaintiff did not need to hire a lawyer—no lawyer is more effective than the judge! While the spirit of the law was not followed, no law was explicitly violated. Each of the favors occurred, at the most, in the grey area between a judge’s discretion and explicit legal requirements. Had the case drawn the attention of the division director, or had it been appealed, the responsible judge could have accounted for her decisions by referring to relevant legal provisions and established procedural rules. Each of these were due to the strong tie between the connection source and the middle-ranking judge, as well as her strong ties with the judges at the case filing division and the responsible judge. If the Case of the Jaywalking Man shows the influence of strong ties in inland areas, the following will show its abundant existence in coastal cities. One judge with eight years of experience in a Shenzhen court said, “I have friends, fellows from the same hometown, and classmates; so does everyone. They may contact me for procedural facilitation, the judges’ schedule, or substantive favors. But overall, the requests are reasonable and I will try my best. In most situations, no money is involved for compensation.” In other words, although a frontline judge may be willing to resist the requests from her superiors, she still has to deal with ties directly from her own network. Another young judge from Shanghai said, “In some cases, I have to tilt towards a particular party because they are friends of relatives or friends of friends. The cases of renqing will not disappear overnight.” A male judge in his mid-thirties in an intermediate court in Guangdong told us about a rape case handled by a trial court in the region. The victim had invited the two defendants to her rental apartment after a late dinner. The rape occurred when all three were playing with an iPhone on her bed. The second defendant abetted the first defendant in the rape. The parents of the second defendant approached our informant judge through his father, since they were all from the same village. The case was not being handled by the court in which the judge worked. The judge told his father that he could not do anything until the case reached the trial stage. The second defendant hired an experienced lawyer who allegedly advised the family to send 30,000 yuan to the prosecutor. The case was still prosecuted; but the prosecutor recommended a jail sentence of a year and a half. The lenient recommendation suggested that the bribe worked. After the hearing, the responsible judge hinted to the lawyer that a suspended sentence was unlikely. The male judge then sought the help of his colleague at the intermediate court who was close to the responsible judge at the trial court. Both the responsible judge and his colleague said that a suspended sentence was difficult, because it usually required the approval of the division director or even the adjudication committee of the court. However, the responsible judge said that he would propose a suspended sentence and see what happened. It turned out that the proposal was approved—possibly the conditions for the suspended sentence had all been met. To express his gratitude, the male judge sent his colleague and the responsible judge each a bag containing 1,000 yuan (approximately equivalent to 150 dollars, a small gift given the cost of living of the city). In this case, the guanxi was nonsupervisory but strong. The male judge could not reject the request from his father, and thus had to help. However, he could not do much. This is why he told the relatives of the second defendant outright that he could not offer any help until the case reached the trial stage. Even at that point, he could not control the outcome. The only help he could offer was to find the right connection to the responsible judge, who agreed to propose a suspended sentence. If the proposed suspended sentence had not been approved, there would have been little help he could provide. D. Nonsupervisory and Weak The situation is significantly different when guanxi is neither supervisory nor strong. Weak horizontal guanxi from outside the network means little to judges. It barely affects the judges’ career, and they are not obligated to act when solicited by someone outside their personal network. A judge is only willing to act discreetly within the scope of her discretionary power, and she has little reason to defy the law. A colleague and personal friend of a dispatched tribunal director told us about a robbery case in his hinterland court. Three young adults had been involved in the case, and two of them were related to the tribunal director (Case of Three Young Robbers). Since there was no clear division between the leading or subordinate roles among the three, under normal circumstances none would have been granted a suspended sentence. Four judges of different seniorities were involved in the decision making: a junior judge who was responsible for handling the case, a presiding judge who chaired the case’s collegiate panel, the criminal division director, and the vice president of the court overseeing criminal cases. With few exceptions, cases involving suspended sentencing must be reported to, and thus determined by, the adjudication committee.46 The tribunal director wanted to get suspended sentences for the two defendants to whom he was related. Among the four judges, he only knew the presiding judge, but they were not close. His relationship with the junior judge who handled the case and his two senior bosses—the division director and the vice president—was strictly professional. He thus had to work on the three officials while also trying to avoid having the case reported to the adjudication committee, in which the court president usually had the final say. The tribunal director thus held no supervisory power over the judges he tried to influence. By official ranking, a tribunal director was more junior and less powerful than the criminal division director, not to mention the vice president. Of course, they were colleagues, but they had no personal ties. He first treated the senior judge and the responsible judge to a banquet. Right after the banquet, upon the recommendations of the two judges, he hired a lawyer, mainly to argue that the two related defendants had only been subordinates in the robbery. Then, taking advantage of the wedding of the son of the vice president, the tribunal director sent him 5,000 yuan as a gift, while the going rate for the red packet of a regular guest was only 200. Two days after the wedding, the tribunal director revealed his relationship with the two defendants to the vice president. He also expressed his wish to get suspended sentences for the two and to avoid the adjudication committee known to the vice president. Similarly, 3,000 yuan was sent to the criminal division director and 2,000 to his not-so-close friend, the senior judge in charge of the collegiate panel. The latter had initially refused the money and said it would be difficult to recommend suspended sentences for the two defendants, given the facts of the case. The tribunal director nonetheless convinced him to take the money by telling him that both the criminal division director and the vice director had agreed. The eventual decision, however, was disappointing for the tribunal director. The case was eventually decided by the adjudication committee, and only one of the two defendants was granted a suspended sentence. The other defendant related to the tribunal director was sentenced to three years in jail with a penalty of 6,000 yuan. This defendant was eventually granted a suspended sentence in the appellate court. The money spent, to quote the tribunal director, was “beyond imagination.” It was said that the first defendant and his family had been outraged by the decision, since they believed that the three young men had committed the same crime. Weak, nonsupervisory ties contrast starkly with supervisory guanxi. In this case, the tribunal director had weak guanxi with only one judge, and was not personally tied to the other three. Our informant said: If the defendants had direct guanxi with the court president or other major local political leaders, the process would have been much different. In that situation, the responsible judge and the collegiate panel would have suggested suspended sentences. This would have been readily approved by the criminal division director. Even if it had been sent to the adjudication committee, every member would have endorsed the suggestion tacitly. The suspended sentences would have been lawfully covered. Once again, in addition to the strength of guanxi, whether it is supervisory, or in this case, potentially supervisory is key. Indeed, the junior responsible judge was the only official not monetarily compensated, though he had directly handled the case. This was because he was junior enough to become the subordinate of the dispatched tribunal director in the future. E. Summary One informant summarized her viewpoint on guanxi as follows: “Guanxi helps, but it must be rock strong [Ying in Chinese]! With shallow personal relationships or indirect guanxi, one can exert little impact on a case. The so-called guanxi may serve as an introduction; but to affect the case outcome, you have to resort to money.” What the judge describes as “rock strong” is the guanxi that we have analyzed as either supervisory, in which the judge feels pressured and must take it seriously because it is from those who govern her, or strong and direct, in which the judge feels trusted and obligated. Another judge in his early fifties said: I feel both obligated and obliged for requests from my supervisors. If I cannot satisfy the requests, it seems that I did not do my job well. But if I do, it is good for me and my career. For requests from my strongly tied friends, I feel emotionally bad if I cannot help. Guanxi is a private, informal process.47 This is because obtaining influence from guanxi is illegal, or at least publicly disapproved. Therefore, both the favor seekers and the favor givers must know and trust each other so as to eliminate concerns about potential risks. In reality, the risk of exposure probably is not high, since a large number of people are involved in such activities and evidence of bribery is hard to come by. But as we will show, the potential risk of exposure does exist, and exposure taints judges’ reputation and career prospects. Moreover, the guanxi must be strong enough to the extent that the favor givers feel obligated to overcome the perception of committing illegal and socially disapproved activities. As a result, the trust and obligation between the guanxi source and the target judge becomes key in understanding guanxi’s influence. The stronger the guanxi between the guanxi source and the target judge, the more influence the guanxi has, because the target judge feels trusted, obligated, and also protected by favoring someone strongly tied. By the same token, the more directly supervisory the guanxi between the guanxi source and the target judge, the more influential the guanxi is. Compared with strong guanxi, which may only reward the target judge through long-term reciprocal benefits, supervisory guanxi can both reward and punish the target judge. It thus may be more powerful. At the same time, the supervisory relationship reinforces the trust and obligation because the target judge feels better protected by her supervisor. On the other hand, weak guanxi’s influence is weak because of a lack of trust and obligation. That is why the influence is the strongest in the category of supervisory and strong guanxi. The influence declines in the category of nonsupervisory but strong ties. It declines further in the supervisory but weak guanxi, and the influence is lowest for nonsupervisory and weak guanxi. IV. Guanxi and Money In the existing literature, guanxi is commonly treated as a precursor of corruption; in most cadre promotion cases documented by Wang, for example, money is involved and transferred through guanxi.48 Our data, however, show the intricate relationship between guanxi pulling and bribery. Strong guanxi, characterized by trust and obligation, discourages the use of money. Favors are not repaid immediately, but they are tallied in long-term obligations. Moreover, guanxi becomes strong when it spans different personal spheres between the favor seeker and favor giver. The offering of money, a rather impersonal medium of exchange, undermines the personal quality of guanxi.49 This seemingly paradoxical finding explains why the most influential guanxi is often the most elusive and untraceable. The vice president in the Case of the Collapsed Kiln did not offer his subordinates any money or favors in exchange. He did not need to. It was understood that he would take care of them in the future. Target judges also expect nonmonetary benefits in the future. Likewise, money is not necessary when the guanxi involved is strong but nonsupervisory. In the Case of the Jaywalking Man, the plaintiff only invited the three involved judges—her classmate, the judge at the case filing division, and the responsible judge—to a banquet. Strong ties rarely require monetary exchange—the ties are too multiplex and resilient to require an immediate and exact repayment. Again, favors will be repaid in the long run. But when the ties are weak, the favor is returned immediately.50 This is true even for a supervisor seeking help from weak guanxi. At the least, he is expected to offer some nonmonetary compensation. A gift or a banquet serves to repay the renqing, or personal debt. This was why the judge refused to help when the vice chief of staff, without offering gifts, sought help in the married woman case. The judge expected an expression of gratitude, since the guanxi was not strong enough to make the request obligatory. The gift was also an acknowledgment that the supervisor was “in debt” to the frontline judge who had offered help. Similarly, when guanxi is neither supervisory nor strong, the weak connection mostly just paves the way for monetary transactions. In the Case of the Jaywalking Man for example, if the judge had gotten the approval of the division director, the guanxi source would have needed money to open doors, since the division director was outside the judge’s trusted personal network. With weak or no guanxi, money serves as a risky substitute for the affectively charged relationships created by gifts and reciprocal favors.51 This is an area where the practice of guanxi shades into the crimes of corruption and bribery. As Li demonstrates, to get things done, a tactic for auctioneers is to first form close relationships with the judges.52 In this way, personal guanxi strengthens the trust and makes it safer to offer money. With only weak guanxi, the effect of money is nevertheless limited. The offer of money has little impact on judges’ actions. Judges are in a higher position of power relative to the litigants. Judges might vaguely respond, “I’ll see what I can do.” The offering of money may make the pleading more earnest, but it never reaches a point where the judge feels obliged or obligated. This distinguishes this type of guanxi from the strong affectively charged relationships. As seen in the Case of Three Young Robbers, the judge who sought help offered a substantial amount of money to the three senior judges. Yet he did not get the desired decision, leading him to spend more money in the appeal process. Judges have little incentive to put in effort, and their actions are unpredictable. V. The Limits of Guanxi The influence of guanxi on judges is not without limits. Although they want to and may favor their connected parties, judges are also constrained. Due to the structure of litigation, most cases have at least two parties, and both parties may have guanxi. The judges involved thus have to balance them. In a hinterland case between a state-owned enterprise (SOE) responsible for a water reservoir and irrigation project and a local credit union, the credit union wanted the SOE to repay a loan before the stipulated term. The SOE had guanxi with the court president, who instructed the responsible judge to deal with the case promptly. The credit union had guanxi with the division director, who instructed the responsible judge to take his time. The two sides had been clashing when the legal representative of the SOE accused the division director of being arrogant and uncooperative. The division director gave in as soon as he realized that the SOE had guanxi with the court president. Eventually, the court president orchestrated a settlement. In a criminal case resulting from a traffic accident in a hinterland court, a couple had been knocked off their motorcycle by a car. The man had only been slightly injured, but the woman died. It turned out that both the car driver and the injured husband were related to the judge. In this type of case, the defendant usually wants a suspended sentence, and the relatives of the deceased want economic compensation. The judge walked a fine line. He spoke to both sides politely and thoughtfully, explaining to them the fine points of the legal procedures. As to the decision, the judge managed to satisfy the request of the man for economic compensation, but also informed him that this meant no jail time for the defendant. Both sides were satisfied. Did guanxi influence the outcome of this case? Yes, if it is judged by the way the judge treated the litigation parties. No, if we are judging by the outcome. The case would have been decided more or less the same without guanxi. As demonstrated by Kwai Ng and Xin He, the major concern of the judge was to settle a case.53 As a result, he managed to get a criminal reconciliation for both parties, even if no guanxi was involved. In this case, the guanxi of one cancels out that of the other. This is because if one party changes his or her mind, the judge is caught between the two connections. In other words, the court is constrained when favoring one litigation party due to the presence of multiple connections. The second limit is the law. In most situations, judges do not commit what the law explicitly disallows. They are pragmatic. Rarely do they risk their careers for the sake of the connected parties, however strong the bonds may be. They are also cautious when engaging in unusual practices that may draw the attention of their superiors. In the Case of the Jaywalking Man, the allocation of responsibility for the plaintiff was set at 20% rather than 10%, since the latter percentage would have been far out of line with established practice. This was done not to draw the attention of the division director. Even in the Case of the Stolen Motorcycle, under the pressure of a strong supervisory tie, the judge still presented a legal pretext (on a technical procedural ground) for his unjust decision. Similarly, in the Case of the Collapsed Kiln, the judges only deferred the case indefinitely, attributing the delay to procedural reasons. With all the supervisory and strong guanxi from the vice president, the judge still did not render a judgment against the plaintiff. Hence, even for a case of strong supervisory ties, in which judges have more room to maneuver and sometimes are emboldened by the protection of their supervisors, they remain cautious. They want to find themselves a way out, in the event a case were to be reviewed by upper-level courts or governments. Self-protection is a top concern. In the Case of the Stolen Motorcycle for example, three men and three women confronted the responsible judge at lunchtime right after the judgment had been rendered. The judge who described the case to us recalled: They were related to the defendant, the old superintendent taking care of the motorcycles. They yelled, “On what basis did you make the judgment? On what basis did you decide that the motorcycle was in the parking lot? The judgment made our father sick to the stomach. He fell ill because of this judgment. Unable to pay the medical bills, we will take him to your court instead of the hospital! Is not the court a place for reasoning?” The responsible judge explained that they could appeal, and asked the protestors to calm down. As they became more agitated, it appeared that they were about to physically assault the judge. One of them said, “Is it still a time for law? At this moment, only economic compensation and the fist can talk.” With sweat running down his forehead, the responsible judge took out a statute book and explained the meaning of the statute of limitations, as well as the procedure to appeal, to the agitated men and women. The judge answered all of the questions the six raised. Thirty minutes later, the petitioners left the court. The judge’s explanation in no way addressed the defendant’s and his family’s grievances. However, he did have the law to fall back on, and he was not in an indefensible position. It would have been difficult for him to do so without appealing to the statute of limitations. Certainly, in the eyes of the legally trained, the judge’s appeal to legal technicality appears suspicious. That is why when determining what to do for the connected party, a judge usually takes into consideration the other party’s level of knowledge and resources. Whether the other party is legally represented, for example, is key. A competent lawyer is able to identify practices that appear to be legally justified but are in fact unusual, and suggest his client either appeal or petition. In civil cases, the judge also evaluates the likely reactions of the party without guanxi. The situation can backfire if a favorable decision to the connected party leads to escalated petitions and protests from the losing party. Perhaps that was why the judges in the Case of the Collapsed Kiln were hesitant to rule against the plaintiff. This risk assessment plays a more prominent role in civil than in criminal cases. In criminal cases, as long as the court does not acquit a defendant, the prosecutor usually will not protest. In the Case of Three Young Robbers, for example, the judge was determined to get suspended sentences for the two connected defendants, despite an objection to the judgment by the first defendant. The third risk is the concern for social instability. Since promoting social stability has become a key criterion in assessing the performance of both judges and courts,54 judges are mindful of the possibilities of protests or petitions when making decisions. As shown in the Case of the Stolen Motorcycle, a protest of the decision was hard to deal with. To favor the connected party is not worth it if it leads to these kinds of incidents. The following example illustrates this further. The plaintiff sued its tenant, a sixty-two-year-old retired worker, for unpaid rents in a rural court (Case of the Poor Retired Worker). The court decided that the tenant would pay approximately 10,000 yuan for the rent and 1,000 yuan for the litigation fees. Nothing was wrong with the decision itself except that the lawyer representing the plaintiff had been a vice president at the local People’s Congress. The defendant petitioned against the decision on the basis that the court had allowed the vice president to represent the case. Our informant said: The defendant said that the People’s Congress was to supervise the work of the court, and this vice president often visited the court to assess the performance of the court. “How could the court allow such a person to be the lawyer representing the case? How could the court be neutral?” To make matters worse, the defendant had a mentally disabled son and a mother in her late eighties. Economically, her family was in a dreadful situation. They were left with only “four empty walls” as is said in a Chinese idiom. In petitions, the defendant took her aging mother and disabled son to the court. Her acts eventually captured the attention of a major political leader in the region, who asked the court to resolve her problem by all means. The court then (1) persuaded the plaintiff to forfeit half of the rent and to compensate the other half from the fund of special remedies; (2) forfeited the litigation fees; and (3) provided more than 50,000 yuan from the court’s remedial fund to support the livelihood of the defendant. Furthermore, the court had already given 1,000 yuan to the defendant at the time of her petitions. The only beneficiary in this case was the vice president of the local People’s Congress. He had received the fees for representing the plaintiff, but the court paid a heavy price for this guanxi. Conclusion Our Article demonstrates that guanxi is key to understanding decision making in Chinese courts. Through outlining the informal ordering based on guanxi, as opposed to formal ordering based on laws and rules, this Article deepens the understanding of the operation of the Chinese courts and their decision making. In particular, the hierarchical nature of the Chinese court makes the supervisory guanxi forcible in affecting the decision-making process. Compared with strong ties, it may reward as well as punish the target judge—a supervisor can not only promote a supervisee, but also make her life miserable. This dimension is so important that the variations between courts in terms of different levels of economic development and legal professionalism are marginalized. In courts with a high level of economic development and legal professionalism, we only see that judges are more discreet. The overall patterns, however, are consistent with those found in hinterland areas. In addition to providing empirical evidence about how guanxi operates in Chinese courts, this Article develops a framework to make sense of the varying influence across different types of guanxi. Contrary to the conventional wisdom of viewing guanxi as either cultural or institutional, we argue that the guanxi practice is both cultural and institutional. The interactions of the two variables make the four types of social ties different from each other in their effects. The cultural thesis suggests that ties differ in strength. Ties in one’s inner circle are more powerful than weak ties in the outermost circle. The institutional thesis suggests that structural positions matter. A direct hierarchical relationship between the favor seeker and favor giver creates an extra dimension of power absent in horizontal relationships. The supervisory or hierarchical characteristics of guanxi force most judges to cooperate, if not heed to the needs of their political patrons or simply their supervisors. The horizontal aspect also helps us understand the outcome of the influence of guanxi: the influence of strong and direct ties is compelling while weak ties are almost futile. Our framework, and especially the supervisory dimension of guanxi, may provide a clue to the long-standing debates on the relative efficacy of social ties.55 Echoing the emphasis on institutional contexts in the study of social ties,56 our analysis suggests that one must first look at the institutional setting. The hierarchical nature of Chinese courts determines the significant influence of supervisory guanxi. The illegal and socially disapproved nature of guanxi also means that strong ties are much more consequential than weak ties to influencing judicial decisions. This study also sheds light on why the public and the media in China perceive corruption as rampant, even though few cases are ever officially reported. According to the Supreme People’s Court’s annual working report of 2015, courts at all levels found only 863 persons guilty of disciplinary violations or violations of the law against abusing adjudicatory or enforcement powers. Of these, only 138 were pursued for criminal responsibility. This number is negligible compared to the 190,000 Chinese judges nationwide. While underreporting may be an issue contributing to the unbelievably low numbers, our analysis shows that guanxi runs through the “pores” of laws and regulations.57 Most favors are given within the scope of discretion and reciprocity in strong guanxi is long term. There is virtually no time limit on repayment.58 Long term and multiplex, strong ties make any concrete offer and acceptance of bribery superfluous. Our study shows that people with weak ties are most likely to resort to offering bribes. If corruption is viewed as “the abuse of public office for private gains,” as most scholars have in the literature,59 the prevalence of guanxi and its influence on the porous judicial system explains the widespread existence of judicial corruption in China. While many connections operate within the ambit of laws, they nonetheless amount to “the abuse of public office for private gains.” Echoing macro-quantitative analyses, which show that corruption is correlated with the ineffectiveness of the legal system, this Article provides micro-level data to demonstrate the dynamics of their interaction. Resonating with Li’s documentation of how several types of officials/judges engage in corruptive activities, this study illustrates how these officials’ behavior is enmeshed with guanxi.60 Since the hierarchical nature of the Chinese court system is shared by other branches of Chinese bureaucracy and state-owned work units, our findings provide insight into the operation of guanxi in these places more generally. Because the limits of guanxi in the setting of Chinese courts, including two countervailing parties and the concerns for the law and social stability, are less a concern for other types of bureaucracies, guanxi’s influences there may be more rampant. By the same token, our findings are also helpful in understanding corruption in countries where the structure of the government is hierarchical and client patronage is the norm. This Article also sheds light on China’s judicial reform and efforts to promote rule of law. Obviously, rule of law efforts have not drastically reduced the prevalence of guanxi. Guanxi thrives when rule of law is underdeveloped and power remains concentrated. In the foreseeable future, there is little doubt that the influence of guanxi will persist. Our framework and analysis points out that the supervisory connection constitutes a major driving force in steering judicial decisions away from strictly legal consideration. The bureaucratic nature of the courts, the incomplete rules, and the vast discretion enjoyed by judges are a hotbed for connections to operate. This is especially true in areas where patron clientelism is the norm. The operation of connections is also hard to detect because it is often covered by a legalized mantle. To minimize the influence of guanxi, China should further reform the bureaucratic nature of the courts. As experiences in urban courts demonstrate, this should help to reduce the impact of supervisory guanxi. It should also allow more room for lawyers to operate in this process which deters judges from maximizing the interests of their guanxi. The law, instead of guanxi, should take center stage in judicial decision making. This research was supported by a NSF grant (award number: 1252067) and a GRF grant from the Hong Kong government. We are also grateful for the comments of the anonymous reviewer of the American Journal of Comparative Law. Special thanks go to the Chinese judges who kindly agreed to be interviewed. Footnotes 1. Xiaotong Fei, From the Soil: The Foundation of Chinese Society (Gary G. Hamilton & Wang Zheng trans., Univ. of California Press 1992) (1949). 2. Jean-Jacques Rousseau, Social Contract (H.J. Tozer trans., Wordsworth Editions, Ltd. 1998) (1762). 3. Thomas Gold, Doug Guthrie & David Wank, An Introduction to the Study of Guanxi, in Social Connections in China: Institutions, Culture, and the Changing Nature of Guanxi3 (Thomas Gold, Doug Guthrie & David Wank eds., 2002); Zhai Xuewei (翟学伟), Renqing, Mianzi Yu Quanli De Zaishengchan (人情、面子与权力的再生产) [The Reproduction of Favor, Face, and Power] (2005); Bian Yanjie (边燕杰), Guanxi Shehuixue: Lilun Yu Yanjiu (关系社会学: 理论与研究) [The Sociology of Guanxi: Theory and Study] (2011); Huang Guangguo (黄光国), Mianzi— Zhongguoren De Quanli Youxi (面子—中国人的权力游戏) [Face—Chinese Game of Power] (2004). 4. Feng Junqi (冯军旗), Zhongxian Ganbu (中县干部) [Cadres of Zhong County] 171 (2010) (unpublished Ph.D. dissertation, Peking University, Beijing, China) (on file with authors). 5. Id. at 172. 6. Wang Xiaoqiang (王小强), Touguo Mianzi De Guanxian Kongzhi—Yi Duofang Shengcunxing Zhihui Boyi Wei Jichu (透过面子的关系案控制—以多方生存性智慧博弈为基础) [Controlling “Guanxi” Cases Through Faces—Based on Game-of-Survival Wisdom of Multiple Parties], 12 Xiangjiang Falü Pinglun (湘江法律评论) [Xiangjiang L. Rev.] 249, 265 (2015). 7. Mark Granovetter, The Strength of Weak Ties, 78 Am. J. Soc. 1360 (1973); Jeremy Boissevain, A Village in Malta (1980); Edward C. Banfield, Political Influence (1961). 8. Doug Guthrie, Information Asymmetries and the Problem of Perception: The Significance of Structural Position in Assessing the Importance of Guanxi in China, inSocial Connections in China: Institutions, Culture, and the Changing Nature of Guanxi, supra note 3, at 37, 40. 9. Doug Guthrie, Dragon in a Three-Piece Suit: The Emergence of Capitalism in China (1999); David L. Wank, Commodifying Communism: Business, Trust, and Politics in a Chinese City (1999); Catherine R. Xin & Jone L. Pearce, Guanxi: Connections as Substitutes for Formal Institutional Support, 39 Acad. Mgmt. J. 1641 (1996); Lisa A. Keister, Guanxi in Business Groups, in Social Connections in China: Institutions, Culture, and the Changing Nature of Guanxi, supra note 3, at 77; Bian, supra note 3. 10. Mayfair Mei-hui Yang, Gifts, Favors, and Banquets: The Art of Social Relationships in China (1994). 11. Yangjie Bian, Guanxi and the Allocation of Urban Jobs in China, 140 China Q. 971, 999 (1994); Yanjie Bian, Bring Strong Ties Back in: Indirect Ties, Networks Bridges, and Job Searches in China, 62 Am. Soc. Rev. 266, 285 (1997) [hereinafter Bian, Bring Strong Ties Back in]; Amy Hanser, Youth Job Searches in Urban China: The Use of Social Connections in a Changing Labor Market, inSocial Connections in China: Institutions, Culture, and the Changing Nature of Guanxi, supra note 3, at 137; Feng, supra note 4. 12. Ethan Michelson & Benjamin L. Read, Public Attitudes Toward Official Justice in Beijing and Rural China, inChinese Justice: Civil Dispute Resolution in Contemporary China 169 (Margaret Y.K. Woo & Mary E. Gallagher eds., 2011). 13. Ling Li, The “Production” of Corruption in China’s Courts: Judicial Politics and Decision Making in a One-Party State, 37 Law & Soc. Inquiry 848, 877 (2012). 14. See Judicial Independence in China: Lessons for Global Rule of Law Promotion (Randall Peerenboom ed., 2009), a book devoted to exploring extralegal influences on China’s judiciary, in which guanxi is conspicuously missing. 15. Gold, Guthrie & Wank, supra note 3, at 3; Pitman B. Potter, Guanxi and the PRC Legal System: From Contradiction to Complementarity, inSocial Connections in China: Institutions, Culture, and the Changing Nature of Guanxi, supra note 3, at 179. 16. Xiaobo Lu, Cadres and Corruption: The Organizational Involution of the Chinese Communist Party (2000); Jing Vivian Zhan, Filling the Gap of Formal Institutions: The Effects of Guanxi Network on Corruption in Reform-Era China, 58 Crime L. & Soc. Change 93 (2012); Peng Wang, Extra-Legal Protection in China: How Guanxi Distorts China’s Legal System and Facilitates the Rise of Unlawful Protectors, 54 Brit. J. Criminology 809 (2014). 17. Wang, supra note 6. 18. Id. at 252. 19. Id. at 254. 20. Id. Bian, Bring Strong Ties Back in, supra note 11. 21. Gold, Guthrie & Wank, supra note 3. 22. Guthrie, supra note 9; Wank, supra note 9; Xin & Pearce, supra note 9; Keister, supra note 9; Josephine Smart & Alan Smart, Personal Relations and Divergent Economies: A Case Study of Hong Kong Investment in South China, 15 Int’l J. Urb. & Regional Res. 216, 233 (1991). 23. Andrew G. Walder, Communist Neo-traditionalism: Work and Authority in Chinese Industry (1988). 24. Fei, supra note 1. 25. Id. at 20. 26. Wang, supra note 6, at 252. 27. See also Wu Tiejun (吴铁钧), “Mianzi” De Dingyi Jiqi Gongneng De Yanjiu Zongshu (“面子”的定义及其功能的研究综述) [A Review on the Study of the Concept of Mianzi and Its Function], 27 Xinli Kexue (心理科学) [Psych. Sci.] 927, 930 (2004). 28. Bian, Bring Strong Ties Back in, supra note 11; Yang, supra note 10. 29. Bian, Bring Strong Ties Back in, supra note 11, at 972. 30. Id. 31. Gold, Guthrie & Wank, supra note 3. 32. Ai Jiahui, Zhongguo Fayuan Zuidahua Shenme? (中国法院最大化什么?) [What Is the Maximization of Chinese Courts?], 3 Falü He Shehui Kexue (法律和社会科学) [Law & Soc. Sci.] 98, 151 (2007). 33. Wang, supra note 6. 34. Id. at 257. See also Xin He, Court Finance and Court Reactions to Judicial Reforms: A Tale of Two Chinese Courts, 31 Law & Pol. 463 (2009). 35. Granovetter, supra note 7. 36. Id.; Peter V. Marsden & Karen E. Campbell, Measuring Tie Strength, 63 Soc. Forces 482, 501 (1984). 37. Bruce Kapferer, Norms and the Manipulation of Relationships in a Work Context, inSocial Networks in Urban Situations: Analyses of Personal Relationships in Central African Towns 181 (J. Clyde Mitchell ed., 1969). 38. Bian, Bring Strong Ties Back in, supra note 11. 39. Lois M. Verbrugge, Multiplexity in Adult Friendships, 57 Soc. Forces 1286 (1979). 40. Wang, supra note 6. 41. Walder, supra note 23, at 170–75. 42. Granovetter, supra note 7; Mark Granovetter, Getting a Job (1974); Mark Granovetter, Afterword to the Second Edition of Getting a Job, supra (2d ed. 1995). 43. Walder, supra note 23, at 170. 44. Gold, Guthrie & Wank, supra note 3, at 14–16. 45. Bian, Bring Strong Ties Back in, supra note 11. 46. Xin He, Black Hole of Responsibility: The Adjudication Committee’s Role in a Chinese Court, 46 Law & Soc’y Rev. 681, 712 (2012). 47. Bian, Bring Strong Ties Back in, supra note 11, at 371. 48. Lu, supra note 16; Ling Li, Performing Bribery in China: Guanxi-Practice, Corruption with a Human Face, 20 J. Contemp. China 1, 20 (2011); Li, supra note 13; Zhan, supra note 16; Wang, supra note 16. 49. Georg Simmel, The Philosophy of Money (1990). 50. Bian, Bring Strong Ties Back in, supra note 11, at 372. 51. Yang, supra note 10, at 167. 52. Li, supra note 48. 53. Kwai Hang Ng & Xin He, The Logics of Legal Commensuration: Criminal Reconciliation and Negotiated Justice in China, 122 Am. J. Sociol. 1104 (2017). 54. Yang Su & Xin He, Street as Courtroom: State Accommodation of Labor Protests in South China, 44 Law & Soc’y Rev. 157 (2010); Benjamin L. Liebman, A Populist Threat to China’s Courts?, inChinese Justice: Civil Dispute Resolution in Contemporary China, supra note 12, at 269; Xin He, Maintaining Stability by Law: Protest-Supported Litigation and Social Change in China, 39 Law & Soc. Inquiry 849 (2014). 55. Granovetter,supra note 42; Nan Lin, Walter M. Ensel & John C. Vaughn, Social Resources and Strength of Ties: Structural Factors in Occupational Status Attainment, 46 Am. Soc. Rev. 393, 405 (1981); Peter V. Marsden & Jeanne S. Hurlbert, Social Resources and Mobility Outcomes, 63 Soc. Forces 482, 501 (1988); Bern Wegener, Job Mobility and Social Ties: Social Resources, Prior Job, and Status Attainment, 56 Am. Soc. Rev. 60, 71 (1991); David Krackhardt, The Strength of Strong Ties: The Importance of Philos in Organizations, in Networks and Organizations: Structure, Form, and Action 216 (N. Nohria & Robert G. Eccles eds., 1992). 56. Mark Granovetter, The Strength of Weak Ties: A Network Theory Revisited, inSocial Structure and Network Analysis (Peter V. Marsden & Nan Lin eds., 1982); Yanjie Bian, Chinese Occupational Prestige: A Comparative Analysis, 11 Int’l Soc. 161, 186 (1996). 57. According to Wang, more than twenty-one laws and regulations are promulgated to prevent guanxi from influencing judicial decisions in China. But most of them do not seem to work. See Wang, supra note 6, at 255–56. For similar views, see Liu Lianjun (刘练军), Ruhe Kongzhi Faguan (如何控制法官) [How to Control Judges], 4 Dongfang Faxue (东方法学) [Oriental L.] 108, 123 (2012). 58. Yang, supra note 10. 59. See Li, supra note 48, at 13. 60. Li, supra note 13. © The Author(s) . Published by Oxford University Press on behalf of the American Society of Comparative Law. All rights reserved. For permissions, please e-mail: firstname.lastname@example.org.
American Journal of Comparative Law – Oxford University Press
Published: Dec 1, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera