2.1 Dual-state framework

“Dual state” was a framework first coined by Ernst Fränkel to describe the form of governance in Nazi Germany during 1933–38. Fränkel’s dual state consists of a “normative state” (Rechtstaat), which refers to “an administrative body endowed with elaborate powers for safeguarding the legal order as expressed in statutes, decisions of the courts and activities of the administrative agencies,” and a “prerogative state,” referring to the “governmental system which exercises unlimited arbitrariness and violence unchecked by any legal guarantees.” ^{Footnote 6} The normative state protected only the “constructive forces” from the Gestapo and its main task was to maintain economic life. ^{Footnote 7} Hence, limited autonomy was granted to the normative state to provide business, trade, and industry with a measure of self-governance. ^{Footnote 8} However, the autonomy of the normative state is vulnerable to encroachment from the prerogative state because the latter has “jurisdiction over jurisdiction” and can draw and redraw the boundaries between them. ^{Footnote 9}

Professor Hualing Fu is perhaps the first constitutional scholar who applied Fränkel’s dual state to China. According to Fu, China’s normative state is built on a “self-defined and self-referencing” legal system ^{Footnote 10} that governs private social and economic affairs; its prerogative state consists of “a zone of exception” that is maintained to “solidif[y] control in areas that are regarded as politically sensitive.” ^{Footnote 11} The normative state, Fu argues, is both resilient and vulnerable to the invasion of the prerogative state despite its “semi-autonomous” status. It is resilient, according to Fu, because it is indispensable for economic development, which is an important source of legitimacy of the Party’s continued rule; it is fragile and vulnerable because it was created by the Party-state for a highly instrumental purpose, it has shallow roots in the Chinese ecosystem, and it has been under periodical siege by a suspicious political system. ^{Footnote 12}

Cora Chan also finds Fränkel’s dual state a valuable concept to portray recent constitutional developments in Hong Kong. However, she argues that this duality may produce a false impression of legal pluralism. Following the Kelsenian legal philosophical tradition, Chan argues that the PRC legal system is fundamentally different from legal pluralism because the normative state is subordinate to the prerogative state and therefore they stand in a monist rather than plural relationship. ^{Footnote 13} Because of the “structural asymmetry” between the two halves of the dual state, Chan contends that “the dual-state is more susceptible to evolving into full-blown, non-bifurcated authoritarianism than into full legality.” ^{Footnote 14}

Eva Pils goes yet one step further. She has not only cast doubts on the claim that authoritarian governance reliant on law is per se valuable but also questioned whether the pursuit of the ideal of rule of law is possible at all in authoritarian systems. ^{Footnote 15} William Hurst presented a dynamic model that was similar to the dual-state framework with a historical dimension. Derived from his inductive analysis and close reading of history, law, and politics in China and Indonesia, Hurst’s model divides legal systems into different “regimes” based on two factors: the level of openness of the polity and the degrees and manner of intervention into the legal system’s handling of specific cases by other state institutions of empowered actors. ^{Footnote 16} According to this model, China was characterized by a “mobilizational legal regime” throughout the Maoist period and thereafter a “hybrid legal regime,” which is divided into a “neotraditional” component for criminal cases and a rule-of-law component for civil dispute resolution. ^{Footnote 17}

2.2 Dual normative system

In my article published in this Journal in 2015, I proposed the framework of a dual normative system. This framework has four components: (1) structural integration of the Party and the state; (2) reserved delegation of authority to the state; (3) bifurcation of state decision-making processes; and (4) cohabitation of the two normative systems: one of the Party and one of the state. According to this framework, the Party normative system regulates the Party-state sphere, where centralization of power and preservation of authoritarianism are dealt with. There, regulation follows Party rules that are enforced by Party disciplines. The state normative system regulates the state-society sphere, where most of the economic and social activities take place. In that domain, regulation is based on state laws sanctioned by the Party.

Unlike the dual-state framework, which perceives the duality in the division of space that is subjected to two different legal systems or regimes, I perceived the duality in the division of labour between two different norm-making authorities: the Party who issues Party rules regulating the Party-state sphere and the state who promulgates laws regulating the state-society sphere. What remains not answered are questions regarding the relation between the Party-state sphere and the state-society sphere: Are the two spheres segregated, overlapping, or completely fused? How do they interact and based on what mechanism(s)?

4.1 Spatial dimension

The spatial dimension of the ARS consists of five scales of administrative divisions (行政区划). The same as in other countries, these scales are used to divide a large territory into a number of units and subunits to make governance practically manageable. Differently from many other countries, however, the relations between the units of difference scales are concentric and highly symmetrical. Currently, China is regulated according to five scales of administrative divisions: ^{Footnote 20}

• Nation/state 国家

• Province 省, including autonomous regions 自治区 and centrally administered metropolis直辖市

• Prefectural city 地级市 ^{Footnote 21}

• County 县 (for rural areas) or district 市辖区 (for urban areas)

• Township 乡镇 (for rural areas) or street-block 街道 (for urban areas)

These five spatial scales constitute the five primary administrative ranks, which are often referred to as “administrative levels 行政层级” in practice. The national level enjoys the jurisdiction of the entire territory of the country. The next level consists of provincial-level territorial units, including 23 provinces, five autonomous regions, and four centrally administered municipalities. Each provincial territorial unit (hereinafter “province”) is further divided into a number of prefectural-level territorial units and each prefectural unit into counties in rural areas and/or districts in urban areas. A county is further divided into townships and a district into street-blocks. Each administrative division can circumscribe the power of its subunits and at the same time its own power is circumscribed by its upper unit. It means that under the national-level authority, no administrative division enjoys autonomy or is free from intervention from above on any issue. Instead, through this chain of command, the national authority can reach every inch of the land without jurisdictional limits based on territorial divide or constitutional restrictions based on subject matters.

4.2 Institutional dimension

For each administrative territorial unit, governance is undertaken by four state organs: people’s congress, government (the administrative branch), court, and procuratorate. Among these organs, the first two, namely the people’s congress and the government, assume the same rank as that of the territorial unit that they govern whilst the court and the procuratorate assume the rank one level lower. The administrative rank indicates the position of an institution in the entire power structure of the Party-state. According to the Central Bianzhi Commission, the administrative rank “designates the administrative status of the institution concerned; it serves to clarify the administrative chain of command, the extent of binding effects of administrative decisions concerned, and the superordinate-subordinate administrative relations of the institutions concerned.” ^{Footnote 22} Different ranks for an institution mean different capacities to mobilize resources and different degrees of political, economic, and regulatory autonomy vis-à-vis other institutions. ^{Footnote 23}

Within each state organ, tiers of divisions are established to perform specialized regulatory or managerial functions. Each of these functional divisions is ranked at the next lower level than the rank of the state organ concerned. This means that the internal power structure of a state organ is connected to its position in the larger power structure where the state organ is situated. For instance, the State Council, as the administrative branch of the state, enjoys the national rank. Under the State Council are several tiers of division. All first-tier divisions, which are mostly ministries, enjoy the provincial rank, which is one level lower than that of the State Council and equal to the rank of a province (hence the name provincial/ministerial rank 省/部级). The rank of the second-tier divisions is two levels lower than that of the State Council and equal to the rank of a prefecture (hence the name prefectural/departmental rank 地/厅级), and so on.

In order to fine-tune the power relations between a large number of territorial units and between an even larger number of bureaucratic units, further gradation of the ranks is required. To that end, the chief/secondary or chief/deputy gradation is introduced, which splits each of the five primary administrative ranks into two: chief grade and secondary grade. For instance, the national rank is split into chief-national 正国级 and secondary-national 副国级, and the provincial rank into chief-provincial 正省级 and secondary-provincial 副省级, and so on. This device of ranking gradation has a long historical root that goes back more than 2,000 years in early Chinese empires such as the Zhou dynasty (1046–256 BC). ^{Footnote 24} As a result of the chief/secondary gradation, the total number of administrative ranks expands from five to 10 (see Table 1) that are applied in both the institutional and the personal dimensions of the ARS. Its application in the spatial dimension is limited, especially for the larger scales. For instance, there is no territorial unit enjoying the secondary-national rank. However, smaller territorial units, such as what are called “city市,” are finely graded. Depending on which city one is referring to, the administrative rank of a city may range from chief-provincial, secondary-provincial, chief-prefectural, deputy-prefectural to chief-county.

Table 1. Conversion table of ranks based on spatial scales and organizational divisions

4.3 Personal dimension

Administrative ranks are designated not only to public institutions, their divisions, and subdivisions but also to officials who fill these institutions. The importance of personal ranks of officials is best illustrated in a Financial Times op-ed written by two authoritative authors on this matter:

This study finds that the ARS standardizes designated power on a national scale. As Frank Pieke has perceptively noted in his study on cadre training, the ARS allows cadres and institutions across the country to immediately “determine the hierarchical relationship between any two individuals or organizations.” ^{Footnote 26} This is achieved by connecting the personal dimension of the ARS with its spatial and institutional dimensions. Specifically, the rank of a public official corresponds to the rank of the institution that he serves, the latter of which corresponds to the rank of the territorial administrative unit to which the institution belongs. The interconnectivity of various dimensions of the ARS helps the Party to reduce the administrative costs of regulating the power relations among all institutional and individual actors who participate in decision-making in public affairs on a gigantic scale.

Due to this interconnectivity, it is more difficult and costly to change institutional ranks than personal ranks because to raise the rank of an institution entails the raising of personal ranks of all personnel in that institution, which has to be matched by an increase in associated salaries and benefits. It is even more costly and difficult to change the rank of a territorial administrative unit because it affects the ranks of all institutions in that territorial unit as well as their personnel.

On the other hand, thanks to the connectivity between institutional and personal ranks and to the fact that they are separately designated, it is possible to adjust the institutional power relations through the so-called “over-equipping 高配” practice without imposing institution-wide or system-wide change and associated costs. “Over-equipping” happens when an institution of a given rank is “equipped” with a top-leader who has a higher rank. For instance, a state organ of a secondary-prefectural rank can be “equipped” with a leader with the chief-prefectural rank, thus raising the power status of this organ vis-à-vis other institutions. ^{Footnote 27} The ranking status of courts is a typical example of the over-equipping practice, which will be detailed later in Section 6.

4.4 Party dimension

The Party dimension of the ARS consists of arrangements that allow the Party to standardize and centralize its control over the state by adding a new layer of power stratification to the ARS. First of all, in every territorial administrative unit, a Party standing committee is established. The same as the people’s congress and the government, the Party standing committee also assumes the same administrative rank as the territorial unit that it governs. However, the relation between them is not equal. Under the principle of “Party leadership,” the Party standing committee leads, directs, and oversees the work of all state organs within its territory through its dispatched offices, including work-committees and “party-groups” that are installed in the decision-making body of each state organ. ^{Footnote 28} The supremacy of the Party authority vis-à-vis state organs is cemented by two arrangements. The first arrangement concerns membership of the Party standing committee and its subcommittees. For instance, members of a provincial Party standing committee typically include a chief Party-secretary, the governor, the first deputy governor, heads of key Party departments and subcommittees, and the heads of the Party from major territorial subunits (the capital city and major prefectures) of the province. Certain state organs do not enjoy membership of the Party standing committee. The most notable example is the judiciary. The head of the highest court of a province is not a member of the provincial Party standing committee. The same arrangement takes place in territorial units of all administrative levels. The exclusion of the judiciary from the Party standing committee places the judiciary at a lower position on the power order and directly restricts the court’s ability to apply the legal order against those actors who enjoy membership of the committee.

The second arrangement concerns the sequencing of membership status in the Party’s decision-making bodies, which introduces an additional layer of power differentiation between those who enjoy membership of the Party decision-making bodies. In other words, not every member is equal. For instance, among members of a Party standing committee, the Party-secretary has a higher-ranking status than the deputy Party-secretary and the deputy Party-secretary than ordinary members of the Party standing committee. For instance, a provincial governor typically holds the position of deputy Party-secretary and plays an assisting role to the chief Party-secretary. Another example is the power relation between the police and the court. During the 2000s, in order to increase the power of the police to control the society, the head of the public security bureau was frequently appointed simultaneously as the head of the Party political-legal committee, to which the leader of the court was only an ordinary member. This arrangement made the police more powerful than the court and police misconducts more difficult to check even though their leaders enjoyed the same administrative rank. ^{Footnote 29}

5.1 Power fields

First, central-local relations. By dividing China’s vast territory into stacked territorial administrative units and assigning different ranks to them, the ARS defines the spatial power relations between the centre and the peripherals. ^{Footnote 30} Studies of Chinese geopolitics have long established that the ARS allows the Party to “strategically reterritorialize as a governing strategy.” ^{Footnote 31} For instance, to facilitate the process of urbanization, the Party “parcel[s] out the authority and power, the autonomy, the status, the importance and the functions” ^{Footnote 32} to the various territorial units through rank adjustments in order to motivate them—a phenomenon referred to as “administrative region economy.” ^{Footnote 33}

Second, Party-state relations. Here, Party-state relations refer not only to the macro-level relation between the Party and the state but also to the meso-level relations between various branches and departments of the Party and between various state organs and their units and subunits. The ARS regulates Party-state relations at both levels as the rank assigned to an institution determines its capacity to mobilize resources and its negotiating power and enforcement capacity vis-à-vis other institutions. Accounts on the potency of the ARS are abundant but the most illustrative case is presented in Abigail Jahiel’s study on the institutional development of the environmental protection agency of the State Council. ^{Footnote 34} Informed by extensive interviews, Jahiel documented in great detail how the regulatory landscape of environmental protection has shifted along the agency’s “rank battle.”

According to Jahiel’s account, when the National Environmental Protection Bureau (EPB) was established in the 1980s, it was given the chief-bureau rank as a third-tier division of the State Council. Its regulatory activities were barely visible. The bureau had left no footprint in making and enforcing national environmental protection policies despite the fact that it was mandated to do so according to the Environment Protection Law passed in 1979. In 1988, the EPB was upgraded to the secondary-ministry rank and became a second-tier division of the State Council solely responsible for environmental protection. Such an upgrade had significantly enlarged its functional domain and institutional resources. ^{Footnote 35} However, pegged to a secondary-ministry rank, the EPB was still one grade lower than other ministries, which limited its capacity to negotiate with the latter and to influence the national policy agenda. ^{Footnote 36} In the 1990s, as the environment began to deteriorate at an alarming pace, the Party-state gradually shifted its policy to balance the need for environmental protection and economic development. Consequently, the EPB was upgraded to a General Bureau and granted a chief-ministry rank in 1998. In 2008, it was further upgraded to the Ministry of Environmental Protection, a proper ministry, which immediately led to an expansion of the staff size by one and a half times. ^{Footnote 37}

In addition, the ARS determines not only the regulatory capacity but also the boundaries of a state organ or agency. It means that state organs or agencies of the same rank are equal and the chain of command does not run between them. This principle is highlighted in the seminal work on policy-making in China by Kenneth Lieberthal and Michel Oksenberg. Drawing on invaluable interviews of governmental informants with work knowledge of public administration in China at high levels, the authors concluded that “units with equal rank have no formal authority over each other” and “only a higher-ranking unit can bring coequals together, elicit decisions, and enforce plans.” ^{Footnote 38} Therefore, rank-based hierarchical interventions become necessary for dealing with domestic cross-border or cross-sector regulatory problems between governments or agencies of equal rank. ^{Footnote 39}

Third, elite politics. It is hardly possible to engage in a proper discussion of elite politics in China on either a theoretical or empirical level without frequent references to the ARS because ranks are the “grammar,” as it were, of Chinese elite politics. Ranks determine an official’s political and social status, one’s scope of power and authority, and the level of prerogatives and privileges that one enjoys. Some officials were so driven by the prospect of a rank promotion that they would pursue radical policies at devastating human costs. A recent study found that the ranks alone can explain 16.83% of the excess death rate during the Great Leap Famine (1959–61) because officials with inferior ranks tended to impose excessive grain procurement targets in order to improve their chance of promotion. ^{Footnote 40} Half a century later, it remains common practice among aspirant officials to ingratiate themselves with their superiors, showering the latter with bribes, gratuitous services, and emotional support, in order to improve their chance of rank promotion. ^{Footnote 41}

5.2 Critical and distinctive features

One should not mistake the ARS for ordinary organizational or professional hierarchies, such as managerial ranks, academic ranks, civil servant ranks, technician ranks, etc., which can be found in all sizeable organizations. The ARS has two significantly distinctive and critical features.

First, multidimensionality. The ARS has multiple dimensions while ordinary organizational or professional hierarchies typically have only one dimension: the personal dimension. Specifically, the ARS regulates not only the power structure within an institution but also the power relations between public institutions, including different branches of the state power and between a political party and these state organs, in a highly integrated and standardized manner, whereas ordinary organizational hierarchies only regulate the internal power structure of an organization, if at all, but not the power relations between different public institutions, certainly not between a political party and these state institutions. Professional ranking systems are even more limited in their application. Being used primarily to regulate standards of professional skills and expertise, they are not only single dimensional but also usually have no indication of the power that one can exert over others. Hence, whilst the ARS is the indispensable regulatory tool of public administration in China, professional ranks are auxiliary and often attached to the former. For instance, when judicial ranks were introduced in China’s courts in the 1990s, they were pegged and attached to administrative ranks instead of replacing the latter. ^{Footnote 42}

Second, interconnectivity. The different dimensions of the ARS are interconnected, which means that the rank of a public post is connected to the rank of the public institution in which the post is located and the rank of the public institution is connected to the rank of the territorial administrative division under which the public institution is established. Both features allow the Party to regulate an enormously large and complex organization with a simple formula whereas, for ordinary organizational or professional hierarchies, given that they have no spatial or institutional dimensions to speak of, their impacts are limited only to the specific organization or profession concerned.

The multidimensionality of the ARS regulating different power fields and its interconnectivity have webbed a multitude of lines of authorities that encompasses all corners and sectors of the Party-state in a single structure. It subjugates all political actors involved, rank by rank, to a single seat of ultimate power: the Party Center and its decision-making bodies. Based on these observations, I contend that the ARS constitutes a formal order of power. In the following sections, I will explain how courts are incorporated into the ARS and its impacts on the legal order.

6.1 Courts embedded in power structures

Within a court, judicial decision-making is stratified by a multi-layer power structure, from court leaders, divisional leaders, to front-line judges. ^{Footnote 43} For a long time, a front-line judge had no power to render a decision without first having it signed off by his or her divisional leader, and then, depending on the stake of the case, also by a court leader. According to the observation of Kwai Ng and Xin He rendered in their study based on extensive fieldwork spanning a decade, judicial decision-making in China’s courts “confirms a well-documented operation principle that permeates every level of organization of the court: administrative rank trumps expertise and collegiality.” ^{Footnote 44}

Over the years, judicial reformers have made a number of attempts to “flatten” the hierarchical decision-making structure but only to find that the ARS is so entrenched that any reform to replace it ends up adding even more levels of hierarchy to it. ^{Footnote 45} The latest round of judicial reforms (2015–16) launched under Xi Jinping’s leadership renewed the earlier efforts to empower front-line judges. It had injected a lot of hope in the reform-minded observers at its inception but became “displaced” and disappointing as the reform proceeded. ^{Footnote 46}

In terms of a court’s external relations, the ARS determines the “administrative status” of a court in the power structure of the Party-state and the extent of autonomy that a court can enjoy vis-à-vis other political actors in the structure. ^{Footnote 47} Since a court had traditionally been considered as a division of the administrative branch, its designated administrative rank is one level lower than that of the government where the court is situated. For instance, a provincial high court is ranked chief-prefectural/department 正地(市)/厅(局)级, the same as a first-tier unit of the provincial government; a prefectural intermediate court is ranked chief-county/bureau 正县处级, and so on. In 1983, as a part of the efforts to rebuild and empower the judiciary, the Party decided to upgrade the rank of courts through a “high-equipping 高配 practice.”

“High-equipping” is a less costly approach to raise the political status of an institution without changing its rank. What it does is to raise or “equip” the top positions of an institution with cadres of a higher rank than the rank of the institution. For example, when the Party sent cadres to fill court posts in 1983, instructions were issued that a court president should be given the secondary rank of the territorial administrative unit where the court sits. Specifically, a provincial high court president would be equipped with a cadre of secondary-provincial rank, even though the court enjoys only the chief-prefectural rank. This upgrade raises the political status of a court president one rank above the leaders of ordinary first-tier units of the executive branch. ^{Footnote 48} The rationale of this practice is that an elevation of the rank of court leaders will bring an elevation of power to courts without actually raising the court’s rank and avoiding the economic costs associated with upgrading the rank of the court.

After taking account of the impact of the high-equipping practices, the ranking status of courts is shown in Table 2.

Table 2. Conversion table of the ARS in different dimensions

Other than being subjugated to the administrative branch of the same territorial unit, a court is also supervised by the appellate court of the territorial unit at the higher administrative level. The extent of supervisory power that a superior court can exert over a lower court in China goes far beyond the kind of appellant power known in Western judicial systems. In the Chinese system, a superior court is entitled to impose judicial policies over a lower court, to review the latter’s cases, to provide judicial guidance in pending cases, to impose or remove the jurisdiction of a case in a lower court, to determine or influence judicial appointments, and to inspect and evaluate the performance of lower courts. ^{Footnote 49}

All the practices discussed above lead to what is called “administrative embeddedness.” According to Ng and He, judicial decision-making in China’s courts shares “a high degree of selfsameness” with other governmental agencies. ^{Footnote 50} This feature makes adjudication in China’s courts distinctive not only from Anglo-American courts but also from the judicial bureaucracy in the continental European systems. This is because, as Ng and He wrote, in a continental European type of judicial bureaucracy, “the law and the rules governing the application of the law are followed as to what has to be done, regardless of who makes the decision,” ^{Footnote 51} whilst in China’s courts, “rules are only applied as to who should be in charge of making the decision” instead of “how the decision should be made or what the decision should be.” ^{Footnote 52} Who has the authority to make what decisions is exactly what the ARS is about.

6.2 Order of power in China’s courts

Two previous studies have pointed out the possibility that there exists an order of power in China’s courts. For instance, in a Special Issue published in this Journal, Juan Wang and Sida Liu found that China’s judiciary is embedded in a relational web that comprises legal professionals, legislative authorities, government authorities, and other social forces. ^{Footnote 53} Both authors then alluded to a coherent power structure that regulates the relational interactions between the actors mentioned.

In the other study, published under the title of “The Power Logic of Justice in China” by Ji Li, the author performed a daunting task to schematize an elusive interactive field. Packaged under the thesis of “power logic,” Ji Li identifies two types of power at work in China’s courts: (1) de jure power, which includes both the laws promulgated by the state and the normative rules issued by the Party, and (2) de facto power, which refers to the ability to engage in collective action, or use brute force or other channels, in particular, informal power of private litigants acquired through political connections. ^{Footnote 54} According to the author, these two types of power constitute the power hierarchy in which courts are situated. ^{Footnote 55} The author believes, quoting Lucian W. Pye consentingly, “[H]igh-ranking officials in China are not constrained by any sense of awe either of the majesty of the law or of a system of ethical virtues. They are constrained by the power of others.” ^{Footnote 56}

In a stylized format, Ji Li develops 15 patterns of triadic power distribution between three actors, namely the plaintiff, the defendant, and the judicial decision-maker, to represent “exhaustively” all variations in judicial behaviour in the real world. ^{Footnote 57} Then, for each pattern, Ji Li shows how the power distribution would determine the ways in which laws would be applied, judicial discretion would be exercised, and resolution methods would be chosen differently by the judicial decision-maker at each stage of the judicial process. ^{Footnote 58} Based on the analysis, the author argues that there is something in common between the trial of, say, Bo Xilai and that of an anonymous villager who sued a governmental agency at a remote corner of the country. ^{Footnote 59} The “power logic of justice” that connects these two cases, Ji Li asserts, is “a function of various configurations of the power status of litigants and judicial decision makers, plus the distributional patterns of information.” ^{Footnote 60} The outcome of this function determines “all major actions of a court” at all stages of litigation in China’s judicial process and the power logic “transcends subject matter areas,” regardless of the nature of the dispute. ^{Footnote 61}

Ji Li’s study not only attests to the postulation that there exists a functional order of power in China’s courts but also lays out how this order of power would interact with the legal order. However, in Ji Li’s “power logic” thesis, power is defined as an abstract concept. The thesis offers no tool to recognize and measure power, which seems incongruous to the broad applicability of this power logic in real life.

7.1 Criminal cases

Unlike in civil and commercial cases, any criminal case has, by definition, a rank-bearing state actor as a litigant. The defendants in most crimes, however, reside at the bottom of the formal power structure of the Party-state and enjoy no administrative rank. For defendants who used to be ranked officials, their rank would have already been stripped from them by the time of prosecution. In addition, each procuratorate in China has a corresponding court in each administrative territorial unit, which share the same administrative rank. Therefore, in criminal cases, the court and the plaintiff are set to be equal in rank. The power relation between the litigants is structurally unequal, with the plaintiff (procuratorate) enjoying more power and privileges than the defendant and the court enjoying little autonomy to dismiss or rule against the procuratorate’s case. This is aligned with the well-acknowledged fact that criminal cases in China have an extremely high conviction rate and courts almost never acquit a defendant unless they are authorized to do so by higher-level authorities. ^{Footnote 63}

The political conditioning of the legal order is also manifested in the protocols of criminal investigations. In China, the primary jurisdictional principle in criminal investigation is territoriality, which means that a case is to be investigated by the investigative authority of the place where the crime has occurred or the place where the suspect or victim has resided. Under this jurisdictional rule, the investigative power is decentralized, which augments investigative resources, motivates local investigative authorities, enables timely investigations, and increases deterrence by improving the chance of detection.

However, duty crimes 职务犯罪 are an exception. Unlike other types of crimes, the investigation of duty crimes is subject to rank jurisdiction. According to the PRC Criminal Law, duty crimes are committed by public officials in the course of discharging or failing to discharge their public duty. Rank jurisdiction requires that a criminal investigation against a ranked official can only be launched by a higher-ranking investigative authority. ^{Footnote 64} And the rank jurisdiction overrides the territorial jurisdiction. For instance, if the investigative agency of Shandong has received a lead about corruption committed in Shandong by a Shandong official, who is a member of the party standing committee of Shandong, the agency has no authority to investigate the case but can only pass the lead to the investigative agency at the central level. ^{Footnote 65} It is not a coincidence but an operational principle that the rank jurisdiction rules on the investigation of duty crimes against a public official mirror the rules that regulate the power to appoint public officials, namely any official of a given rank can only be appointed by an authority of a higher rank. These two decision-making powers are intentionally aligned to ensure that power flows in the same top-down direction as much in the coming to power as in the removal from power of any public official.

If the investigation concludes that an official is criminally liable, the traditional practice is that the investigative agency will then refer the case to the corresponding procuratorate of the same rank, which would then prosecute the case at the corresponding court of the same rank. This process has been modified for cases against high-ranking officials for practical reasons in recent decades. In the past, a criminal case against an official appointed by the Party Center 中管干部 must be prosecuted by the highest-ranking procuratorate, i.e. the Supreme People’s Procuratorate (SPP), at the highest-ranking court, namely the Supreme People’s Court (SPC). Beginning from the 1990s, the intensification of both corruption and anti-corruption activities has led to a number of practical problems. First, processing many high-profile cases at the highest-ranking judicial institutions would inevitably bring a lot of unwanted attention to these institutions and place them under critical public scrutiny. Second, it would overwhelm these institutions in terms of case-load. Third, if the SPC became the first-instance trial court, the defendant would lose the opportunity to appeal since there is no court above the SPC. To overcome or avoid these issues, arrangements have been made to allow the SPP to relegate the prosecution to a procuratorate at a lower level, usually at the prefectural rank, through an individual-case-based special authorization. The tasked procuratorate will then bring the case to the court where it is located. The prosecution and trial will be supervised from Beijing through the back channels. ^{Footnote 66}

It is worth noting that, in contrast to duty crimes, the investigation of political crimes committed by political dissidents is also decentralized and follows the territorial principle in the same way as ordinary crimes. In fact, a considerable volume of political offences are prosecuted under non-political crimes, such as “picking quarrels and provoking troubles” or disturbing the public order, which are handled by the investigative authorities at the grass-roots level who are not required to seek approval from higher authorities for preliminary investigations. The fact that the investigation of political offences is decentralized and treated the same as non-political offences may be surprising to those observers and analysts who believe in a bifurcation of the “prerogative state” and a “normative state,” where the legal norms that are observed in the latter are discarded in the former in order to protect the Party’s prerogatives.

Under the framework proposed in this article, the picture is different. When criminal laws are implemented by law enforcement agencies and/or applied in courts, the legal order has already been relativized in favour of the state due to the structured power relations between the litigants and the court. Therefore, it is not necessary to separate politics from non-political offences and create an “extra-legal” space to repress them. It is certainly against the Party’s interest to centralize the handling of political offences because it would slow down the investigation, reduce the chance of detection, and unnecessarily place the central authorities under the spotlight. On the contrary, it is the duty crimes committed by the privileged political elites that require centralized control from the Party because these officials enjoy prerogatives, which, as dictated by the order of power, can be taken away only by those with more prerogatives.

7.2 Administrative and “constitutional” cases

Known as a component of the “small constitution,” ^{Footnote 67} the PRC Administrative Litigation Law (ALL) was passed in 1989 and makes it possible for citizens to sue governmental agencies in courts. ^{Footnote 68} In the two decades that followed, the number of administrative litigation cases grew very slowly and unevenly between rural and urban regions. ^{Footnote 69} Governments won many more cases than they lost. ^{Footnote 70} In recent years, the ALL has been revitalized under the leadership of Xi Jinping. As a component of his rule-of-law political-legal project, ^{Footnote 71} the ALL was amended in 2014, which expanded the scope of administrative conducts and agencies that can be sued in courts. ^{Footnote 72}

Administrative cases provide the best opportunity to study the impact of the order of power because the defendant in each case is by definition a ranked governmental institution and the plaintiff is almost always a private person or entity who assumes no position in the ARS. Unlike in criminal cases, where the ranks of the procuratorate and the court are always equal, the ranking relation between the court and the defendant varies. First of all, Party institutions of any level cannot be sued at courts. This means that a court cannot rule against a decision or a conduct made by a Party institution at any level on the grounds of law. This “judicial immunity” protects the supremacy of the authority of the Party as a whole with the assurance that its authority cannot be challenged in terms of legality in a court and/or ruled against by a court.

Second, according to the rank jurisdictional rules stipulated in the ALL (2017), grass-roots-level basic courts can only take cases against functional departments of a government at the province level and below. Complaints against governments of the county level and above as well as ministries or other functional departments of the State Council have to be brought to an intermediate court. Given that the defendants in most administrative cases are local governments and their departments, the rank jurisdictional rules are designed to allow a great number of administrative cases heard by courts that outrank the defendant. Only by positioning their ranking relation in this way would the court be able to enjoy a measure of autonomy to rule against the defendant and make the administrative litigation minimally meaningful. Recent studies find a strong correlation between the winning rate of the defendants in administrative cases and its ranking position vis-à-vis courts. For instance, Ma Chao et al. studied 240,000 administrative court decisions and found that courts are more likely to rule in favour of the plaintiff when the court outranks the defendant, regardless of whether the court is an ordinary court or a specialized administrative court. ^{Footnote 73} In a different study, Hui Zhou et al. analyzed 70,000 administrative litigation cases and found that a one-unit increase in the defendant’s rank relative to the court’s decreases the plaintiff’s win odds by 42.99%. ^{Footnote 74} Similar patterns are also observed by Xiang Miao and Fan Liangcong in a study of 1,150 administrative litigation cases concerning house demolition in Zhejiang province in 2013. ^{Footnote 75}

Third, in cases where a court faces a higher-ranking defendant, a number of strategies are adopted by the court to avoid clashes between the ranking order and the legal order. For instance, when a court finds that a higher-ranking defendant has overtly violated the law and evidently infringed upon the rights of the plaintiff, the court can render declarative, non-enforceable court orders, which appears to have upheld the law but does not offer any substantive relief to the plaintiff. More specifically, a court can, for example, rule against a defendant by declaring that the disputed action taken by the defendant is unlawful and at the same time reject all remedial claims of the plaintiff, or kick the case back to the administrative complaint system, or, in the better scenario, order the defendant to recompense the plaintiff according to the same compensation scheme that was rejected by the plaintiff prior to the litigation. ^{Footnote 76} Yu Xiaohong calls this kind of court practice a “respectful concession to core interests [of governmental defendants]” as one strategic approach to comply with the political order. ^{Footnote 77}

The legal order would clash with the ranking order when a court intends to rule merely on the grounds of law against a defendant which has a higher rank than itself. In such circumstances, courts would take measures to first realign the ranking order before upholding the law. Such measures include jurisdictional rearrangements, such as the invocation of foreign jurisdiction or rank-lifting. Both measures are designed to break the chain of dependency between the original court and the defendant by transferring the case to a different court. ^{Footnote 78} Courts can also seek instructions from the appellate court, which is necessarily two ranks higher than itself, to compensate for its rank deficiency. This instruction-seeking process is easy to miss because it usually takes place behind the courtrooms and is not recorded in the court judgment. It is therefore worthwhile to share in this space the behind-the-scenes rank fight in the Henan Seeds case, which was once celebrated as a “quasi-constitutional” case and sparked hope for the introduction of judicial review. ^{Footnote 79}

This case started with a judgment from Luoyang Intermediate Court (rank 5) in 2003. The dispute concerned the terms of price of agricultural seeds in a sales contract between two companies. The court rejected the claim of the defendant on the grounds that the local statute that the defendant relied on is inferior in authority to the state law, which favoured the plaintiff’s claim. The ruling caused fury from the local and provincial law-makers, namely Luoyang People’s Congress (rank 6) and Henan Provincial People’s Congress (rank 8). Being furious at the ruling of Luoyang Intermediate Court, which is inferior in ranking, the Henan Provincial People’s Congress circulated a decision that annulled the court ruling and demanded that the judges who were involved in the making of the ruling must be punished. ^{Footnote 80} Under pressure, the Party-group of the Luoyang Intermediate Court suspended Judge Li Huijuan who heard the case and the deputy-head of her division. ^{Footnote 81} At the same time, the defendant appealed the case to Henan High Court (rank 7), which in turn referred it up to the SPC (rank 9). The SPC issued a reply that supported the initial ruling. It was only then that Henan High Court ruled to sustain the ruling of the first-instance court. ^{Footnote 82}

7.3 Civil and commercial cases

In criminal and administrative cases, courts always face a litigant who has a ranked position in the ARS: the procuratorate in criminal cases and a government or governmental agency in administrative cases. The direct participation of another state power in these disputes immediately places a constraint upon courts and limits the scope of autonomy that the court can enjoy. In civil and commercial cases, however, litigants are either private actors without a position in the ARS or ranked individuals or entities who engage in private affairs and are treated, at least in the letters of the law, as private actors in the disputes concerned. Hence, civil and commercial cases are where courts enjoy maximal judicial autonomy. ^{Footnote 83} The increased judicial autonomy in these cases also opens the space for judges to rule in favour of any party who is willing and able to provide sufficient affective and/or economic incentives. Such corruption opportunities are presented to all judges who participate in the decision-making process at different stages of the life of a case. However, the distribution of opportunities and resources to engage in corruption with judicial decision-makers is highly uneven between litigants and clearly favours those with power or associated with those in power. ^{Footnote 84}

Equally, the distribution of opportunities for corruption is also uneven between judges of different ranking statuses and the distributional patterns are fully aligned with the order of power. As explained in Section 6, the decision-making power in China’s courts is highly concentrated in the hands of leader judges. ^{Footnote 85} However, leader judges handle cases only indirectly through supervising the work of and giving instructions to front-line judges. Front-line judges enjoy sole decision-making power in cases that are not significant enough to attract the attention of their superiors. In addition, since leader judges do not attend court hearings on a regular basis and have to base their decisions largely on the information provided by front-line judges in oral or written case summaries, front-line judges can influence the outcomes of these cases by choosing to provide only information that favours their proposed rulings. Hence, according to my previous research on judicial corruption in China’s courts, a large amount of misconduct of front-line judges occurred in the fact-finding process during the trial, such as admitting or excluding evidence without giving the parties equal opportunities to contest it, tampering with evidence, and obstructing access to evidence by violating the discovery procedure or manipulating the forensic examination results. Such misconduct involves blatant violations of ethical and procedural rules and is susceptible to a higher chance of exposure and detection. Moreover, the number of bribes that front-line judges receive is limited because of the comparatively low stake of the cases that they handle and because their decision-making power can be curtailed by leader judges.

In contrast, leader judges deal with high-stake and high-economic-value cases. The greater the stake and value involved in a case, the higher the level of supervision and approval required. It is indisputably evident that judges of higher ranking receive bigger sums of bribes simply because the power that they hold has greater exchange value. For instance, the president of Anhui High Court Zhang Jian who fell in 2021 was convicted of bribe-taking a total of more than RMB 71 million and Xi Xiaoming who used to be a first-rank justice and SPC vice president took more than RMB 100 million in bribes.

Furthermore, since leader judges participate in decision-making only indirectly and briefly through supervision and sending instructions to front-line judges, it allows them to handle a much greater number of cases of interest than they could manage if they were required to participate fully in the judicial process. This arrangement has, at the same time, increased the leader judges’ capacity to engage in more corruption in a limited span of time. For example, Zhang Jiahui, who fell from the position of vice president of Hainan High Court in 2020, had taken bribes totalling RMB 437 million from 37 bribers in exchange for her “help” in at least 59 cases in a five-year period. And the higher Zhang had been promoted on the ranking ladder, the more bribes she received.

In summary, in China’s courts, when a plaintiff brings a case to court, a micro power field is formed. The outcome of the case is no longer simply a matter of law but also a matter of the power dynamics between all actors involved in the case. By employing the rank-based jurisdictional rules, the Party can adjust the measure of autonomy that courts can enjoy in cases of different subject matters, thus modifying the weight of the political order vis-à-vis that of the legal order in decision-making and aligning the two orders through a number of supplementary measures. For instance, in criminal cases, the power relation between the litigants is structurally unequal because the procuratorate is a ranked state institution and the defendant either has no ranking status or has already had it stripped off him at the time of prosecution. And since the court and procuratorate are pegged and are given the same rank, the court enjoys little autonomy to dismiss the procuratorate’s case or acquit the defendant. My analysis also shows that the power dynamic between litigants changes in duty crimes that involve ranked officials. However, this change takes place at the investigation stage, not after. In the investigation of duty crimes, the ranking status of the offender dictates who could launch an investigation. The rank-based jurisdictional rule has effectively centralized the investigative power of duty crimes and offers protection for the prerogative enjoyed by political elites.

Similarly, in administrative cases, rank jurisdictional rules are introduced to require plaintiffs to bring cases against higher-ranking governmental institutions to higher-ranking courts, thus narrowing the rank gap between the court and the defendant so that the court can enjoy a measure, however limited, of autonomy to make the litigation minimally meaningful. In cases where courts are outranked by defendants, which is largely a reluctant solution to reduce the increasing case-load of higher courts and to contain contentions that arise from problematic administrative practices at the grass-roots level as much as possible, a number of supplementary measures are at the courts’ disposal to make courts appear to be upholding the law without undermining the authority of the governmental defendant and/or infringing upon its core interest.

Among all subject matters, it is in civil and commercial cases that courts enjoy the most autonomy because both litigants appear in courts as private individuals. Such autonomy also affords litigants and judges the most opportunities to engage in corruption. Corruption certainly does not take place in a power vacuum. As I have shown above, the distribution of the opportunities and resources for corruption is highly uneven not only between the litigants but also between judicial decision-makers. The distributional patterns in both groups are fully aligned with the order of power. For litigants, those who are closer to power have more opportunities and resources to influence judges either through political pressure or economic incentives or both. For judges, those with higher ranks have greater opportunities to engage in corruption and reap more rewards. This is because, as I have concluded in my earlier study, judicial corruption in China’s courts is not a simple aggregation of a few isolated deviant behaviours but a by-product of the routine operation of a judicial decision-making mechanism that is based on an order of power. ^{Footnote 86}