Administrative Litigation in China: Assessing the Chief Officials ’ Appearance System

The Chief Officials ’ Appearance System (COAS), introduced in 2015, requires government leaders to appear in court and explain their actions. Unlike other post-2014 legal reforms aimed at reducing political influence in administrative litigation, the COAS uniquely actively involves political officials. This approach is based on the belief that increased participation will help officials to gain a better understanding of public concerns and improve administrative litigation quality. However, few studies have examined the system ’ s effectiveness

In 2015, the revision of the Administrative Litigation Law (ALL) introduced the Chief Officials' Appearance System (xingzheng jiguan fuzeren chuting yingsu zhidu 行政机关负责人出庭应诉制度, COAS hereafter).The revised law requires agency leaders, rather than their legal counsel, to appear in court and defend their administrative actions. 1 The authorities, including the Chinese Communist Party (CCP) Central Committee, the State Council (SC) and the Supreme People's Court (SPC), expected the COAS to "enhance the administrative agencies' legal consciousness, raise the quality of trial and law enforcement, and properly resolve the conflicts between 'mandarins' (i.e. government officials) and citizens," thereby improving the effectiveness of administrative litigation cases (ALCs) in practice. 2dministrative litigation in China has long been seen as a "frail weapon that [has] failed to reduce administrative arbitrariness." 3 This point of view is revealed in previous studies from two perspectives.First, substantially, the ALL allows only limited litigant rights.For example, the ALL stipulates that plaintiffs may only challenge concrete rather than abstract administrative acts and that courts can review only their legality and not propriety. 4Even after its 2015 revision, the ALL's breadth and depth for judicial review improved only modestly. 5Second, in practice, with Chinese courts deeply embedded in local politics, the act of pursuing administrative litigation is often portrayed as an ultimately ineffective endeavour, akin to breaking an egg by throwing it against a stone. 6For example, Kevin O'Brien and Lianjiang Li have described how local officials have pre-empted, derailed or undermined administrative lawsuits, forcing villagers to rally support from sympathetic elites, mobilize collective appeals or stage public protests in order to obtain redress for grievances. 7In sum, scholars opine that ALCs in China are notoriously difficult to file, adjudicate and enforce, owing to the political embeddedness of the Chinese judicial system. 8he most recent reforms since the Fourth Plenum of the 18th Party Congress and the 2015 revision of the ALL are aimed at resolving such political embeddedness, by centralizing the management of courts and experimenting with administrative courts.However, some studies have shown that these reforms have been ineffective at tackling such embeddedness. 9Chao Ma, Chao-Yo Cheng and Haibo He, for instance, explore the impact of designated jurisdiction in ALCs. 10 In 2014, the SPC designated that the railway transport courts (RTCs) would accept and hear ALCs, because they were under the direct administration of the provincial high courts and presumably more independent.Through analysis of 238,000 court decisions in 2015-2019, it was found that only at the primary level were the RTCs more likely to side with citizens.Such effects became statistically insignificant in cases against high-level agencies. 11evertheless, other studies reveal a more positive outcome for administrative litigation.For example, some scholars argue that audacious and reform-minded local courts have strategically used ALCs as hedges against powerful local governments. 12With bottom-up judicial innovation, expanded jurisdiction and coordinated and more enforceable decisions than before, local courts have become regular and strategic participants in local politics and consequently have enhanced their status therein. 13Nonetheless, both groups of studies have empirically engaged with reform measures that are aimed at fending off the political impacts on ALL operations in China.
The COAS, however, employs a rather different logic.By bringing officials back into administrative litigation, it instead aims to increase political officials' participation.The belief that court appearances by chief officials can significantly contribute to the resolution of administrative disputes was based, at least in part, on the recognition of agency leaders' disproportionate influence in shaping and carrying out policies. 14Sadly, not many studies have evaluated whether such a different logic would, in practice, boost the effectiveness of the ALL in China.Most of these studies are based on anecdotal evidence and thus lack explanatory power. 15The current article aims to provide one of the first systematic empirical examinations of the COAS.We chose Beijing as the most-likely case for a systematic analysis.Specifically, we examined 1,551 ALCs that were held in a Beijing court from 2015 to 2018 and conducted in-depth interviews with judges, administrative officials, plaintiffs and lawyers.In addition, we also undertook a brief analysis of national ALCs taken from the China Judicial Politics Database (CJPD) and supplemented this with in-depth interviews in 12 other provinces and data drawn from internal documents.
Contrary to official expectations, our findings reveal mixed results regarding the impact of the COAS.On the one hand, the COAS was not well received by either chief officials or plaintiffs.In Beijing, chief officials appeared in only 62 out of 1,551 relevant cases, mainly to meet evaluation criteria set by their superiors.Moreover, the COAS seemed to generate more grievances than it resolved.On average, plaintiffs were 5.08 times more likely to appeal to higher courts or file second suits when chief officials appeared in court.On the other hand, we observed surprisingly impartial and even strategic responses from the court.Whether chief officials appeared in court or not had no significant effect on the outcome of judicial decisions.Looking more closely at the officials who took the stand in Beijing, we found that outgoing officials lost more frequently than their younger and possibly more promising colleagues.Moreover, the national implementation of the COAS revealed a similar, if not worse, scenario.By and large, we found that the COAS is characterized by a renewed triad in ALCs: indifferent state agencies, increasingly agitated plaintiffs and strategic and empowered courts.
Furthermore, the unintended impacts of the COAS carry certain implications.As evidenced in the subsequent sections, the ineffectiveness of this "Chinese medicine" can be attributed to the transformed conditions it was originally designed for, particularly heightened legal awareness among citizens and the strategic empowerment of the courts in China. 16These altered circumstances, as demonstrated by existing research on Chinese administrative litigation, are the outcome of progressive legal reforms implemented in China over the past four decades.17Still water runs deep, and the perverse impact of the COAS implies that a rule-based approach to dispute resolution would be a more desirable and effective route than a paternalistic one.
The rest of this paper is organized into seven sections.In the next section, we introduce the development of the COAS and propose our hypotheses based on the established discussions around the triad of administrative litigation in China.The third section reports the data and our methodology.Subsequently, we explore the impacts of the COAS by examining the triad involved in the ALCsthat is, the plaintiffs, governments and courts.The article then continues by providing a preliminary examination of the national implementation of the COAS.Finally, we summarize our findings and discuss their implications for effectively resolving administrative disputes in China's transitional context and suggest an outlook for future research.

The Chief Officials' Appearance System
Institutional background According to Article 3 of the ALL (2015), "the person in charge of an administrative agency against which a complaint is filed shall appear in court to respond to the complaint, or, if he or she is unable to appear in court, authorize a relevant employee of the administrative agency to appear in court."This is often referred to as the COAS clause.
Traditionally, administrative agencies have tended to be absent from court hearings in China.Scholars have argued that the absence of accused agencies has impeded effective communication among parties and left plaintiffs dissatisfied. 18In response, between the late 1990s and early 2000s, local courts in the provinces of Shaanxi and Jiangsu recommended that government leaders actively participate in trials.Subsequently, in 2006 and2007, the central authorities, including the CCP Central Committee, the SC and the SPC, acknowledged the beneficial outcomes of this system and decided to implement it nationwide.In 2013, when the National People's Congress (NPC) initiated the revision of the ALL, numerous legislators demanded a summary of the successful local experiences of the COAS and made corresponding revisions to the ALL. 20Consequently, in 2015, the COAS was formally written into law.
The SPC soon followed suit and further developed the COAS through a series of judicial interpretations (JIs). 21These developments can be summarized in four aspects.First, the definition of "chief officials" was expanded to include not only the principal and deputy chiefs but also individuals who manage the implementation of the administrative act or who are in charge of such management work. 22This expansion is a small concession to busy government leaders in the hope that it will promote compliance with the COAS.Second, JIs instructed chief officials to show up in four types of cases: those involving significant public interest, those generating enormous publicity, those potentially leading to mass incidents and those where the court makes such demands in writing. 23hird, when appearing in court, agency leaders are expected to actively participate in the proceedings by, for example, "mak[ing] statements and defences, submit[ting] evidence, debat[ing], offer[ing] final opinions on the case and explain[ing] the regulatory documents serving as the basis" for the case. 24In other words, they cannot remain silent in the courtroom.Last, in cases when chief officials fail to appear, JIs instruct local courts to record officials' absence, issue judicial suggestions to their superiors and disclose such information to the public. 25To summarize, the COAS is a multipurpose system tasked with effectively resolving administrative conflicts, enhancing the awareness and consciousness of administrative officials, and raising the overall level of law enforcement in China.
Moreover, unlike other legal reforms introduced in China since 2014 that mainly aim to empower the judicial system and fend off local protectionism, 26 the COAS instead embodies a drastically distinctive logic.It is deeply rooted in China's paternalistic meritocracy, with competent and virtuous rules put in place to look after the people's interests, and the role of citizens confined to participation and communication. 27By placing chief officials on the stand, the COAS in effect further tilts unbalanced suits between citizens and "mandarins."The expectation that court appearances by chief officials would substantially help to resolve administrative disputes at least partially dwelt on agency leaders' undue influence in policymaking and implementation. 28xisting studies on the COAS have predominantly relied on anecdotal evidence and focus on the appearance rate of officials. 29However, by using anecdotal evidence, these studies have failed to generate a systematic evaluation on the effectiveness of the COAS.The two major studies that have employed large data sets again mainly focus on the appearance rate. 30Nevertheless, as illustrated above, existing studies of the ALL, along with the legislators' expectations, reveal that using the appearance rate is simply insufficient to testify to the effectiveness of the COAS in practice, as administrative litigation itself involves not only the necessary participation of governmental officials but also that of judges and ordinary citizens.Thus, this article aims to provide a more comprehensive and systematic examination of the COAS by looking at not only the appearance rate of officials but also court judgments and considers whether the system can effectively resolve citizens' grievances.

Theory and hypotheses
With the balance between parties interrupted by the appearance of chief officials in line with the COAS, what might happen in ALCs in China?How effective is COAS in resolving administrative conflicts in practice?This study aims to examine the profound impact of the COAS on the triad of ALCsnamely, the plaintiffs, accused administrative agencies and courts.Incorporating findings from previous studies on administrative litigation in China, we propose three hypotheses regarding the abovementioned questions.
First, from the plaintiff's perspective, does a court appearance by a chief official lead to a better resolution of administrative disputes?Advocates of the COAS claim that when chief officials directly engage with ALCs, the plaintiffs should at least be psychologically appeased, as they can finally meet the agency leaders in person. 31Accordingly, plaintiffs should accept court decisions and be less likely to appeal.However, sceptics note that when chief officials appear in court defiantly, ordinary citizens may be further agitated and become more inclined to appeal. 32Thus, we test the following opposing hypotheses.
H1a: When chief officials take the stand, the plaintiff will be less likely to appeal or file a new suit over the same issue.
H1b: When chief officials take the stand, the plaintiff will be more likely to appeal or file a new suit over the same issue.
Second, from the perspective of administrative agencies, do court appearances by chief officials lead to more lawful administrative actions?Advocates argue that the system is instrumental in improving the legal consciousness of both chief officials and administrative staff.When chief officials show up in court, the agency as a whole may recognize the importance of laws and will operate more in compliance than they would otherwise. 33Consequently, the COAS could lead to fewer ALCs over time.Moreover, better administrative performance could bode well for administrative agencies in courts.Plaintiff claims with merit would be settled through negotiations and therefore economize court time.Hence, when unsatisfied plaintiffs file complaints, courts have fewer grounds to affirm their complaints, and the plaintiff success rate declines.However, if the COAS does not function as designed, we might observe few changes in either the number of ALCs or the win rates for administrative agencies. 3430 Li, Huai 2016; Hong and Huang 2021.31 He, Xin 2013.32 Li, Huai 2016.Notably, plaintiffs may also resort to extra-legal means to resolve such conflicts, such as mediation and petitions.See, e.g., Minzner 2006; 2011; Wang, Juan 2012; He, Xin, and Feng 2016; Ng and He 2017.We, however, believe that such practices will not significantly impact our analysis for two reasons.First, we double-checked data from the CJO, and no chief officials appeared in court hearings in mediated cases.Second, since a reform in 2014, lawrelated petitions are no longer accepted by local governments.33 He, Xin 2013.34 Yu, Shaoru 2016.
The China Quarterly H2a: After chief officials appear in court, the number of ALCs will decrease, while the win rate of administrative agencies will increase.
H2b: After chief officials appear in court, there will be no significant changes in either the number of ALCs or the win rate of the administrative agencies.
Finally, from the perspective of the court, do the appearances of agency leaders impact judicial decisions?Advocates of the system argue that it has the potential to enhance court authority.Xin He, for example, contested that the COAS tilted the balance of power towards the courts, as judges took the lead and chief officials served supporting roles. 35With enhanced authority, judges would be at liberty to follow the law to the letter and rule accordingly.Their decisions would not be affected by whether chief officials took the stand.Conversely, others have argued that judges' face-to-face interaction with chief officials, often with higher ranks, would create even greater opportunity for administrative interference. 36This would increase the likelihood of favourable rulings for administrative agencies.
H3a: In cases where chief officials appear in courts, the court's decisions are no different from when chief officials do not take the stand.
H3b: In cases where chief officials appear in courts, the courts will be more likely to rule in favour of administrative agencies.

Research Design
To examine the aforementioned hypotheses, the present study combines a detailed case study of the COAS practices in a Beijing local court as the most-likely case and a preliminary analysis of the national implementation of the COAS.Broadly speaking, we employ quantitative and qualitative data to assess the COAS's impacts on the ALC triad.

Beijing as the most-likely case
We examine the COAS in Beijing as the most-likely case, and its design adheres to the principle of inverse Sinatra inference, which suggests that if the system fails there, it will fail anywhere. 37As the capital city, Beijing enjoys considerable geographical advantages, and its courts have played a key role in China's legal reforms.Specifically, prior studies have noted that the impact of administrative litigation is shaped by the level of economic development, local legal environment and judicial performance in the designated locality. 38As shown in Figure 1, according to economic development, as measured in GDP per capita, judicial transparency, as measured in the number of ALCs published, and legal environment ranking, Beijing constitutes one of the best scenarios for developing administrative litigation (see also Table A1 in the online Appendix).
For ALC disposition, as shown in Figure 2, between 1988 and 2016, the national average plaintiff win rate was 14.8 per cent, fluctuating between a high of 23.1 per cent (in 1992) and a low of 7.7 per cent (in 2012). 39From 2013 to 2016, plaintiffs in Beijing prevailed in ALCs at roughly the same rate (11.9 per cent) as those nationwide (11.7 per cent).35 He, Xin 2013.36 Yu, Shaoru 2016; Hong and Huang 2021.37 Levy 2008.38 Pei 1997; O'Brien and Li 2004; Li, Ji 2013; Ng and He 2017.39 As noted by scholars, a considerable proportion of plaintiffs withdrew their cases, and some obtained the desired judicial relief.Pei 1997 and Peerenboom 2008 therefore estimate that the plaintiffs' actual winning rate may be more than 30%.
The odds of plaintiffs in China prevailing are significantly higher than those of plaintiffs in other countries, such as the US and Japan.
We further limit our analysis to ALCs against district-level governments in Beijing for two reasons.First, since the most recent judicial reform, all ALCs against district-level governments in Beijing are now under the jurisdiction of one court.Analysing these cases exclusively enables us to control court-level variation.Second, we reasonably expect the court appearance of district-level officials to have a greater impact than the appearance of officials at other levels, as higher-level governments enjoy greater authority and greater public trust in general. 40This is consistent with the research design of a most-likely case.
We employed two sets of original data from Beijing: judgments downloaded from China Judgments Online (CJO), the official publishing platform for court documents, and interviews with plaintiffs, judges, government officials and lawyers.First, we manually compiled a database of 1,551 cases filed against district-level governments in Beijing from 2015 to 2018 (COAS Databank).Judges confirmed in interviews that they released all cases online, and our case study is therefore exempted from the missingness issue that has plagued most studies using CJO data. 41Based on logistic models, we assess the effect of the COAS in Beijing on dispute resolution and case outcomes and report the empirical results in the later sections. 42Second, we conducted  semi-structured interviews in Beijing from 2019 to 2022.In total, we interviewed 11 judges, three government officials, one plaintiff and five lawyers.To increase the credibility of our findings, we also triangulated the interview data with related internal documents and news reports.
Examining the national picture In addition, to determine if the significance of our findings, derived from the Beijing case analysis, can be corroborated at the national level, we also look into the relevant national data.We sourced the national data from two different channels: ALCs from the CJPD and interviews in 12 representative provinces, as well as some internal documents.First, we pulled all relevant ALCs from CJPD using syntactic rules.Owing to the unsolvable anti-crawling techniques of the CJO, the CJPD contains approximately 70 per cent of all published cases but is nevertheless one of the most comprehensive databases of Chinese judicial decisions. 43Second, we conducted interviews with judges, officials and plaintiffs in 12 other provinces: Zhejiang, Henan, Guizhou, Sichuan, Jiangsu, Guangdong, Qinghai, Shaanxi, Tianjin, Liaoning, Hebei and Xinjiang. 44These provinces were selected because they are representative of the broader context.Furthermore, we acquired internal documents from our interviewees.We combined these two approaches and are confident that examining the national picture of the COAS provides valuable insights to complement the analysis of the Beijing case.

The COAS and Plaintiffs
Have chief officials' appearances in court helped to resolve administrative disputes?Using Models 1-3, we test H1 in this section and report the empirical results in Table 1.The independent variable is whether chief officials appeared in court, and the dependent variable is whether the plaintiff appealed or filed a new suit over the same issue.Models 4-6 test H3, and the dependent variable is whether the court decided in favour of the plaintiffs.We report only the results for Models 3 and 6 with both time and district fixed effects controlled.
As shown under Model 3, when government officials appeared in court, the plaintiffs were 5.08 times more likely to appeal or file a new suit over the same issue (1/exp (-1.626) = 5.08).This result is statistically significant at the 1 per cent nominal level.Contrary to our expectation, there seems to be a reproductive function of the COAS even in the best-case scenario.It seems ironic that a system introduced to resolve disputes has ended up generating more of them.Through interviews with judges, lawyers, government officials and plaintiffs, such a "reproductive" function can be understood by the "mismatched" expectations at both ends: pragmatic plaintiffs and indifferent and sometimes overconfident officials.
First, plaintiffs have grown increasingly pragmatic in ALCs.On the one hand, they have cherished the opportunity to communicate face-to-face with chief officials, expecting it to help address their grievances.They have felt treated with greater respect and consequently expect more favourable results.Judges noted that the plaintiffs "were very excited and thought that they would eventually obtain favourable results.Oftentimes plaintiffs were more than eager to communicate with chief officials … some even approached chief officials immediately after trials, and some officials were willing to explain to them again about government policies." 45n the other hand, plaintiffs' enthusiasm for meeting with chief officials is irrelevant to the decision of appeal.As one plaintiff noted in an interview, "I prefer higher-level government officials to appear in courts because they can resolve the problem more efficiently and they would be exempted from individual interests … If the agency can honour my claims, I'm willing to withdraw." 46 The China Quarterly judge further confirmed this pragmatism: "Currently, [the plaintiffs] no longer consider it an honour to shake hands with leaders … what they want is real benefits, such as better compensation in land acquisition cases." 47Another judge even mentioned that in one case, "the plaintiff and his lawyer refused to meet with chief officials in courts.They feared that they would lose the case if officials did take the stand." 48The judge had to make a concerted effort to persuade the plaintiff to meet the officials.
Second, chief officials seldom actively participated in trials.Their indifferent and sometimes overconfident behaviour risked further angering the aggrieved plaintiffs.It has been widely reported that chief officials remain silent most of the time.As one judge noted, the chief officials "were voiceless … many remained silent throughout the trial … most of the time, they simply read a prepared statement at the end of the trial." 49The chief officials' caution is understandable: "most district heads don't have backgrounds in law.They fear that once they say something wrong, it will be recorded and used against them in the courtroom." 50ome officials made statements that only further agitated the plaintiffs.In one interview, the official stated: Generally speaking, we have done a fair job in terms of "the administrative rule of law."We are thoroughly cautious about the entire administrative process, whether in granting permissions or imposing penalties.You can see from the overall outcomes of our ALCswe rarely make mistakes and seldom lose cases, but some plaintiffs are very stubborn.They do not trust us but simply insist that we are wrong … taking the stand is mostly to satisfy the emotional needs of the plaintiffs, as we all know they trust their leaders. 51en pragmatic plaintiffs met with indifferent or even overconfident officials, the COAS generated more administrative disputes, which was the opposite of what the policy was designed to do.

The COAS and Administrative Agencies
Have the appearances of officials in court induced any changes in administrative behaviour?No evidence confirmed that the COAS visibly impacted administrative behaviour between 2015 and 2018, based on Figure 3.The figure shows the number of ALCs and plaintiff win rates in the 16 districts of Beijing, and the vertical dashed lines are the time of each court appearance by chief officials.Specifically, we found that no consistent trend existed in terms of the number of ALCs or the local plaintiff win rates across the 16 districts after chief officials appeared in court.We found a slight, short-term decline in ALCs and a decrease in the plaintiff win rate in three districts, Shunyi 顺义, Dongcheng 东城 and Changping 昌平, but only for one quarter.Consistent with H2b, the COAS did not induce any consistent or significant changes in either the number of ALCs or the win rate of administrative agencies.This lack of impact is due in part to government officials' overall indifference towards the system in both the timing and frequency of court appearances.First, most chief officials appeared in courts at the end of the year when the annual cadre evaluation was underway.As supported by Figure 4, 73.6 per cent of chief officials took the stand in the fourth quarter, which was significantly higher than the proportions in the other three quarters.In one interview, the government official explained that "chief officials are under the pressure of the annual evaluation, and taking the stand is one of the evaluation indicators … [chief officials] are so busy at work.Thus, we sometimes have to remind them to complete this task before the end of the year." 52econd, the frequency of chief officials' court appearances is consistent with the minimum standards set for cadre evaluations.According to a 2018 internal document, "if the annual number of ALCs is higher than five and lower than ten, the chief officials shall appear in courts at least twice a year; if the annual number of ALCs is higher than ten, the chief officials shall appear in courts at least three times." 53Figure 5 shows the number of ALCs and the frequency of chief officials' court appearances in Daxing 大兴 district from 2010 to 2019.The number of appearances was exactly two or three times, although the overall number of ALCs fluctuated considerably in the given period.

The COAS and the Courts
Has the COAS impacted judicial decisions in any meaningful way?As shown in Table 1, the system has had no significant impact on the rulings in ALCs.Consistent with H3a, the Beijing court ruled in an impartial fashion and did not discriminate between cases with and without the presence of local leaders.
What explains this surprising finding?A closer investigation suggests that strategies adopted by local courts offset the possible impacts of the COAS.First, both the case records and our interviews indicate that when chief officials took the stand, court leaders assumed the roles of presiding judges. 54As shown in Figure 6, in cases where chief officials testified, presidents of the court heard 61.3 per cent of relevant ALCs, and division chiefs heard 25.8 per cent.This might be owing to a courteous reaction from the court, but it strengthened the psychological grounds for judges to make decisions impartially.
Second, courts were more likely to favour young and promising government leaders than older, outgoing officials.In 62 cases in which chief officials showed up in court, 33 officials were involved.The numbers of cases and officials were too small for multivariate analysis; we therefore carried out a descriptive bivariate analysis.Figures 7 and 8 plot chief officials' tenure and age against judgments in the cases.On average, district-level chief officials in Beijing were 46 years old and had been in office for 22 months.Interestingly, there was an adverse relationship between chief officials' tenure 52 Ibid.53 Art. 9, Daxing Government, 2008  or age and case outcomes.Young and promising chief officials generally won more cases than their outgoing colleagues.To some extent, the strategic consideration of Chinese judges resembles the "strategic defection" of Argentinian judges.Gretchen Helmke notes that Argentinian judges tended to rule against the government when it began to lose power. 55The judges' lack of institutional security incentivized them to distance themselves from the outgoing government. 56Similarly, in our interview, one court leader commented, "When deciding cases, we must consider the political, legal and social effects … Most of our work relies on the support from the governments, and we need to ensure their support, not to make them 'lose face'." 57In that sense, young and promising chief officials may be able to ensure better cooperative terms with the court.The China Quarterly

A Comparative Picture
The Beijing case shows that contrary to the authorities' original intent, the COAS produces a perverse impact on the triad involved in an administrative dispute.This case offers detailed and systematic evidence of the mediocre performance of the COAS.However, there remain two   Initially, we extracted relevant ALCs from the CJPD by utilizing syntactic rules.In total, we identified 28,805 judgments where chief officials appeared in court, and 146 of them involved city-level leaders, who are equivalent to district-level leaders in Beijing. 58Figures 9 and 10 report the proportion and disposition of these cases, respectively.Broadly speaking, government leaders across China rarely attended court hearings.On average, chief officials attended 929 cases, while city-level officials in each province attended five between 2015 and 2018.Beijing has witnessed the most court appearances by higher-level officials.Quite interestingly, a significant number of plaintiffs obtained favourable outcomes in the COAS cases.The average win rate for plaintiffs in the COAS cases was 43.1 per cent at all levels and 41.9 per cent at the city level, well above the aforementioned national rate of 12.9 per cent.We leave systematic or even causal examination of this issue for future multivariate analysis.
Additionally, we conducted extensive interviews in 12 other provinces (see Figures 9 and 10, highlighted in bolded italics).Generally, the interviews provided a similar but dimmer scenario concerning the implementation of the COAS.First, when chief officials appeared in court, the courtrooms were equally characterized by pragmatic plaintiffs and silent officials.As a judge in Qinghai province noted, "plaintiffs were the so-called nail households (dingzihu 钉子户), ones who refused 58 In China's administrative hierarchy, Beijing is a municipality directly under the central government and its districts are at the same level as prefectural cities.
The China Quarterly to relocate and came to court only to claim better compensation.They wouldn't be pleased by simply meeting the officials in person.Instead, sometimes they even interrogated the officials and we had to interrupt and stop them." 59Such pragmatic accounts were also reported by judges from Jiangsu and Liaoning. 60Moreover, government officials seldom actively attended trials.A report in Zhejiang mentioned several instances of officials' inactive participation, including remaining silent when questioned, playing with their phones in the courtroom, or leaving the court in the middle of a hearing. 61The High People's Court in Henan conducted a survey with 300 administrative judges and reported that only 26 per cent agreed that the "COAS effectively resolved administrative disputes," 23 per cent observed "no visible effect" and 45 per cent noted that "although the COAS was not effective in resolving the disputes, it did soothe the plaintiffs' emotions." 62Second, from local court reports, we found that chief officials often appeared in court only when required to do so for their performance evaluation.For example, in Qinghai province, the administration only recommended that officials appear, and they seldom did. 63In Liaoning province, a judge tried to persuade the provincial administration to install a Suzhou-style cadre evaluation mechanism to force agency leaders to appear in court: "it was impossible, [he ridiculed,] to force officials to take the stand without a gun." 64 In contrast, in provinces where the number of court appearances was included in the cadre evaluation scheme, such as Guangdong and Zhejiang, some chief officials offered to attend court by the end of the year. 65When the weight of court appearances in cadre evaluations was reduced, however, as reported by Zhejiang courts, the motivation to appear diminished significantly. 66Third, on a national level, we observed a greater variety of activist/strategic responses to the lack of official appearances at local courts.For instance, many courts endeavoured to persuade local Party committees, people's congresses or local governments to issue specific directives mandating chief officials' court appearances.One Liaoning judge even used his close personal connections with the agency leader to encourage him to be the first mover. 67More audacious judges capitalized on theatrical court hearings when officials were present.One judge from Zhejiang commented that "sometimes I would intentionally tolerate plaintiffs' emotional expressions to exert extra pressure on administrative agencies." 68Another Qinghai judge used the court hearing as a platform to disseminate legal knowledge and he often directly interrogated chief officials in order to send a clear signal that he was going to rule against the agency.According to him, "it had proven effective to facilitate the after-hearing coordination.After the hearing, the agencies would be more than happy to negotiate with plaintiffs." 69ncluding Remarks Incorporated into the ALL in 2015, the COAS is a unique Chinese practice designed to increase the participation of political officials in administrative litigation.Legislators initially hoped that the COAS would effectively resolve such disputes by leveraging the undue influence of agency leaders in policymaking and implementation.However, contrary to official expectations, this study finds that the COAS reproduces the administrative grievances that it is tasked with substantially resolving.Moreover, when chief officials appear in court, administrative litigation is characterized by a renewed triad: apathetic state agencies, increasingly agitated plaintiffs and strategically empowered courts in Beijing and beyond.
First, government officials' court appearances are often marked by their indifferent and sometimes even antagonizing attitude towards plaintiffs.In such cases, the officials' court appearances do not lead to better litigation outcomes nor fewer disputes further down the line.Second, the COAS generates more controversies than it solves: when government officials appear in court, plaintiffs file more appeals or pursue a second litigation.Finally, our findings corroborate He's argument, which posits that courts benefit from the COAS and are likely to be the only immediate beneficiaries. 70The courts rule impartially and display no significant differences in their rulings whether chief officials appear in court or not.Our subsequent descriptive analysis highlights courts' strategic considerations: young, promising government officials fare better than their older colleagues.Additionally, we observed a greater diversity of positive and strategic behaviour among local courts on a national scale than before.
What are the implications of such a renewed triad in effectively addressing administrative grievances in China during this transitional period?First, as discussed above, scholars have dismissed administrative litigation in China as no more than a "frail weapon" because of the political constraints courts face when adjudicating cases of "citizens versus mandarins." 71The stealthy empowerment of the courts manifested in our study showcases the lifting of such political constraints to some extent.This results from the cumulative effects of several reform measures implemented over the last four decades and the strategic behaviour of judges specifically introduced to carry out the COAS. 72The courts appoint their most prestigious and high-ranked judges to preside over cases in which government officials take the stand, both to display their deference to officials and to ensure their equal standing with government agencies.Although officials' court appearances did not seem to make a difference overall, judges were more likely to rule against the government when the officials were nearing the end of their term.In a less amiable legal environment, judges took the initiative to consult upper-level governments or strategically used the theatrical court hearing to exert extra pressure on local governments.Although still deeply embedded in the political system, empowered courts may deliver impartial decisions and potentially resolve administrative disputes more efficiently than before.
The second implication is the plaintiffs' enhanced legal consciousness.Contrary to the expectations of some scholars, the mere appearance of government officials in the courtroom did not appease aggrieved citizens.In contrast, we observed adverse effects of officials' court appearances, as citizens were motivated to take further legal action.As summarized by the interviewed judges, this was partly owing to citizens' improved legal knowledge and the idea of social justicethat is, their awareness of their own rights. 73To some extent, the reproductive feature of the COAS resembles the "mismatched discourse" Xin He and Yuqing Feng identified in China's petition system. 74The mismatch between the "legal terms" employed by petitioners and the "channelling discourse" of petition officials is injurious to petitioners' experience and incentivizes them to make new claims.In sum, both the citizen's enhanced legal consciousness and the stealthy empowerment of the courts in China imply that a rule-based approach of resolving disputes seems more desirable and probable than a paternalistic one.
Having explored the implications of the COAS for administrative litigation and legal reforms in China, we want to reflect on how related research might proceed in the future.This paper presents a detailed examination of ALCs in a Beijing court from 2015 to 2018, supplemented by a rough approximation of the overall situation in China.With more comprehensive national data, future research could expand on this study and explore regional variations in the efficacy of the COAS.Furthermore, to fully understand the interactions between officials and citizens, future research should look beyond administrative litigation, which is often placed at the end of the conflict spectrum, and look into the effectiveness of other dispute resolution mechanisms.Since the turn of the century, China has pushed for a diverse set of conflict resolution mechanisms, including petitions, mediation and arbitration.If in-person appeals to the leaders no longer work in the courtroom, we could reasonably expect a similar trend in the implementation of other dispute resolution mechanisms.Supplementary material.To view supplementary material for this article, please visit https://doi.org/10.1017/S0305741024000018

Figure 1 .
Figure 1.Regional Distribution of Published ALCs, GDP per Capita, and Legal Environment Ranking in 2016 Sources: Data regarding the number of ALCs were retrieved from China Judgements Online (https://wenshu.court.gov.cn).Data on GDP per capita were collected from the China Statistical Yearbook 2017.Data on legal environment ranking were collected from the 2017 "Report of the NERI index of marketization of China's provinces."Notes: Circle size represents the legal environment ranking: the higher the ranking, the larger the circle.Beijing is ranked second out of 31 provinces.

Figure 2 .
Figure 2. Number of ALCs and Plaintiff Win Rate in China and Beijing, 1988-2018 Source: Zhongguo falü nianjian (Law Yearbook of China), various years.Notes: Only in four years (2013-2016) did the authorities report the win rates of plaintiffs in ALCs in Beijing.Please refer to Appendix B for the calculation of the plaintiff win rate and the original data.Since 2017, the Law Yearbook of China no longer reports the case disposition of ALCs.

Figure 3 .
Figure 3. Number of ALCs and Plaintiff Win Rate by District in Beijing, 2015-2018 Source: COAS Databank.Notes: The bars here represent the number of ALCs; the lines represent the plaintiff win rate; and the vertical dashed lines represent the timing of each court appearance by chief officials.

Figure 5 .
Figure 5. Frequency of Chief Officials' Court Appearances in Daxing District, 2010-2019 Sources: Internal documents provided by the interviewee.

Figure 6 .
Figure 6.Rank of the Presiding Judges Source: COAS Databank.

Figure 7 .
Figure 7. Scatter Plot of Chief Officials' Tenures and Case Outcomes Sources: Case outcome data are from COAS Databank; chief officials' tenure and age data are from official government websites.
unanswered questions: first, to what extent do the patterns observed in the Beijing case apply to its implementation on a national scale?Second, how can we comprehend the implications of these patterns in effectively addressing administrative grievances in China during its transition?We explore the first question in this section and the other in the conclusion.

Figure 8 .
Figure 8. Scatter Plot of Chief Officials' Ages and Case Outcomes Sources: Case outcome data are from COAS Databank; chief officials' tenure and age data are from official government websites.

Figure 9 .
Figure 9. COAS Cases Involving City-level Officials, 2015-2018 Source: CJPD.Notes: The dark bars represent the number of COAS cases that plaintiffs won, and the light bars represent the number of COAS cases that the relevant public agency won.The vertical dashed line marks the average number of COAS cases across the 31 provinces.The data tabs show the total number of COAS cases for each province.The plaintiff win rate is in parentheses.

Figure 10 .
Figure 10.All COAS Cases, 2015-2018 Source: CJPD.Notes: The dark bars represent the number of COAS cases that plaintiffs won, and the light bars represent the number of COAS cases that the relevant public agency won.The vertical dashed line marks the average number of COAS cases across the 31 provinces.The data tabs show the total number of COAS cases for each province.The plaintiff win rate is in parentheses.
A 43Liebman et al. 2020.44SeeTable E1 in the online Appendix for further details.45 Interview with judges, Beijing, December 2019.46 Interview with plaintiff, Beijing, April 2022.

Table 1 .
Regression Results Robust standard errors are in parentheses.*, ** and *** represent significance at the 10%, 5% and 1% nominal levels, respectively.Control variables include plaintiff characteristics (whether the plaintiff is an individual or an organization, the number of plaintiffs, whether they are represented by a lawyer), defendant characteristics (whether there are co-defendants, whether they are represented by a lawyer), judge characteristics (whether the court head is involved), and case characteristics (number of laws cited, type of case).Please see the online Appendix D for the full table of regression results.
Notes: 47 Interview with judge, Beijing, December 2019.48 Interview with judges, Beijing, March 2022.49 Interview with judge, Beijing, December 2019.50 Interview with judge, Beijing, March 2022.51 Interview with government official, Beijing, January 2020.