The Institutional Context

Chinese bureaucracy since 1949 has been built on two institutional foundations. The first one is the Leninist party organization, whose organizing principle is political in nature, emphasizing discipline, loyalty, and hierarchy (Selznick, Reference Selznick1952). All government offices are under the strong control of the party-state. The bureaucratic organizations entrench into all arenas and different corners of society, at different administrative levels and in different functional areas.

Chinese bureaucracy in the contemporary era also draws from a second source, which is the historical legacy of the imperial bureaucracy. As scholars have noted, many institutional arrangements, such as personnel management practice and the tiao-kuai dual authority in government structure, resemble the imperial bureaucracy of the past (Brandt, Ma, & Rawski, Reference Brandt, Ma and Rawski2014; Walder, Reference Walder1986; Zhou, Reference Zhou2016). More importantly, traditional social relations, characteristic of what sociologist Fei Xiaotong called ‘the differential mode of association’ (chaxugeju), are alive and thriving, permeating bureaucratic processes across levels and over time.

Both institutional sources have a strong presence in Chinese bureaucracy, in various forms, across different areas, and at different points in time. The Leninist party organization is integrated into formal institutions such as organizational apparatus and disciplinary measures, whereas the traditional legacies find their expressions mostly in informal institutions and social relations. The two are interdependent: the operation of formal institutions is lubricated by social relations, which in turn are often embedded in stable organizational structures. These two are also in contention, as shown by the continuous political campaigns that aim to strengthen formal authorities and rein in informal practices.

The formal institutions of Chinese bureaucracy are both the basis and the result of state-making in the 20th century. Earlier studies proposed various models of totalitarianism and authoritarianism in which bureaucratic organizations served as an organizational weapon. It is through the bureaucratic organization that the authoritarian state exerts its control over different areas. As many have observed, one salient feature of Chinese bureaucracy is its high mobilizational capacities. This results largely from tight organizational links across administrative levels and the strong centralized authority of the party-state.

Over time, scholars came to recognize that the Leninist state is not monothetic (Skilling, Reference Skilling1983); rather, it is infused with informal social relations. Political scientist Lieberthal (Reference Lieberthal, Lieberthal and Lampton1992: 8) proposed the model of fragmented authoritarianism, with the key argument that ‘authority below the very peak of the Chinese political system is fragmented and disjointed’. Implicit in the fragmented authoritarianism model is a rational image of Weberian bureaucracy where higher authorities impose their will and political control through a top-down process (Schurmann, Reference Schurmann1968; Selznick, Reference Selznick1952). Against this rational image, Chinese bureaucracy was seen either as a highly efficient organizational weapon of the state, or as suffering from serious organizational failures (Harding, Reference Harding1981). The model of fragmented authoritarianism sees Chinese bureaucracy as an organization plagued with double pathologies – an authoritarian structure with either excessive centralization of political power or considerable fragmentation, both at odds with the rational design of the organization. In the post-Mao era of decentralization, scholars tried to make sense of the active role of local states in their interactions with state policies and incentive design (Landry, Reference Landry2008; Oi, Reference Oi1995; Walder, Reference Walder1995).

The Organizational Lens

Economist Coase (Reference Coase1937) asked this question: If market transactions are an efficient way of organizing economic activities, why do so many firms exist? Conversely, if formal organizations are the efficient way of organizing, why isn’t the entire national economy organized into one large firm? This question raises fundamental issues about the kinds of transaction costs associated with different ways of organizing and the role of government organizations.

As Pfeffer and Salancik (Reference Pfeffer and Salancik1978) argued, all organizations must adapt to their external environment for resources, and in so doing, adjust their internal structures and strategies accordingly. Institutional theorists Meyer and Rowan (Reference Meyer and Rowan1977) further emphasized the importance of institutional environments, where legal rules and government regulations are the major driving forces. Sociologist Simmel (Reference Simmel1955) pointed out that even antagonistic interactions between two parties may induce isomorphic behaviors toward each other. That is, in response to coercive and competitive pressures, organizations will adopt forms and strategies from their adversaries, thereby inducing isomorphic behaviors.

In this intellectual context, the significance of Chinese bureaucracy cannot be overstated. The Chinese government, with its regulatory and normative power, provides the most important institutional environment to which all other organizations must adapt in order to survive (Haveman, Jia, Shi, & Wang, Reference Haveman, Jia, Shi and Wang2017). While studies published in organizational research journals, including MOR, mostly focus on the organization of for-profit firms or non-profit non-governmental organizations, many of them connect their theoretical arguments, findings, and implications to the organizational patterns of the Chinese government (Gao, Wang, Wang, & Li, Reference Gao, Huang, Sui and Zheng2023; Li, Tsang, Luo, & Ying, Reference Li, Tsang, Luo and Ying2016; Schuler, Shi, Hoskisson, & Chen, Reference Schuler, Shi, Hoskisson and Chen2017; Wang, Du, & Marquis, Reference Wang, Du and Marquis2016; Yan & Chang, Reference Yan and Chang2018; Zhang, Marquis, & Qiao, Reference Zhang, Marquis and Qiao2016; Zhang, Sun, & Qiao, Reference Zhang, Sun and Qiao2020; Zheng, Kulwant, & Mitchell, Reference Zheng, Kulwant and Mitchell2015).

An organizational lens calls attention to a set of issues about transaction costs, agency problems, and associated incentive design, and other organizational responses in Chinese bureaucracy. The sheer size of the bureaucratic organization, under the central authority at the very top, leads to a long chain of command, involving multiple levels and many related offices and officials. Among various transaction costs, agency problems are especially salient. Such agency problems have induced to distinctive organizational response in Chinese bureaucracy, especially in the prevalence of guanxi and informal institutions in bureaucratic processes. Moreover, interactions between informal institutions and formal authority/institutions lead to organizational dynamics of variable coupling between the central and local governments, across hierarchical levels, and among different arenas.

In the rest of this paper, we will focus on three sets of issues: (1) agency problems and incentive design, (2) guanxi as informal institutions, and (3) variable coupling in the Chinese bureaucracy. Although these issues are often discussed in different literatures, in our view they are closely interrelated, pointing to some salient organizational phenomena in a comparative perspective and calling for the further development of organization theory.

Agency Problems in Chinese Bureaucracy

Agency problems are especially salient and exacerbated in Chinese bureaucracy. One key characteristic is the centralization of authority in Chinese bureaucracy, resulting in a long chain of command involving multiple bureaucratic levels and the dual authority structure of the party and government with multiple principals. For example, public policy on the ‘village road-pavement project’ in the early 2000s was initiated by the Ministry of Transportation in Beijing, which provided earmarked funds for cement road pavement in selected villages nationwide. This policy and related funds were then implemented through the corresponding government offices at provincial, prefectural, and county levels and down to the selected townships and villages. It is mind-boggling to imagine the agency problems involved in this long chain of command. As the policy was implemented downward and finally reached the villages, we observed multiple agents – officials at different levels, local cadres in villages, and different parties participating in the pavement projects – as well as multiple processes – reinterpretation, persuasion, fund diversion, distortion, local mobilization, and so on. The interactions among these multiple actors and processes led to highly unpredictable policy outcomes, with staggering collective debts and village governance undermined (Zhou, Reference Zhou2022: Chapter 9).

Given these agency problems in Chinese bureaucracy, what kinds of institutional practices and mechanisms arise to address these issues? What are the (unintended) consequences?

Incentive Provision: The Institutional Practice

Economic theory of incentives has examined different scenarios regarding incentive provision in public administration (Laffont & Tirole, Reference Laffont and Tirole1993; Tirole, Reference Tirole1994). In the principal-agent framework, Dixit (Reference Dixit2002) proposed a set of criteria for incentive provision in public administration. A key insight in this literature is that there should be weak incentive provision in public administration. The rationale is that public offices are responsible for providing public goods with equal access to all citizens, and their work often involves multiple principals, multiple tasks, and different constituents. Strong incentives along one dimension are likely to induce biased attention in a particular direction, neglecting other tasks and other constituents.

Explicitly or implicitly, incentive problems have also haunted the planned economy in state socialist societies, inducing problems such as shortage and rushed growth (Kornai, Reference Kornai1992). In China, economic incentives were largely abandoned in the de-stratification efforts in the Mao era. Moreover, frequent policy shifts and continuous political campaigns engendered political disillusion and ideological crises, leading to, as Harding (Reference Harding1981) observed, ‘a stiffing conformity among officials, a reluctance to take independent initiatives, a low level of technical and managerial skills, and disillusionment and cynicism at the frequent shifts in official line’, as well as widespread resistance through collective inaction (Zhou, Reference Zhou2022).

A major aspect of economic reform in the post-Mao era was the introduction of incentives by relaxing tight state control over the means of production, first in the form of returning farming decision rights to villagers in rural China, which led to a spectacular increase in agricultural productivity (Perkins, Reference Perkins1988). This positive feedback induced further efforts in the decentralization in urban areas, opening up the private sector and reforming the ownership of collective and of small- and mid-sized state firms.

Incentive provisions have also played an important role in Chinese bureaucracy in the post-Mao era. Walder (Reference Walder1995) highlighted local governments as the headquarters of local industrial firms, with better information and better incentive in managing local development. Oi (Reference Oi1995) discussed the role of local governments in promoting economic developments. Zhou Feizhou (Zhou, Reference Zhou2012) provided a systematic study of the historical evolution of taxation policies, where central and local governments adopted various strategies and behaviors to maximize revenue. As Cheung (Reference Cheung2008) observed, one important propelling force in China’s economic development is the intensive competition among governments at the local (county) level for outside investment and human capital.

This line of arguments is best articulated in Li-an Zhou’s (Reference Zhou2007) tournament model. L. Zhou argues that the central government adopted the tournament model of incentive provision, where the principal sets up well-defined policy targets (e.g., GDP growth rate), and allows chief officials across administrative jurisdictions to compete among one another. Those who perform well are promoted, i.e., advanced to the next round for future competition. The tournament model has several advantages, such as using relative performance evaluation that economizes information and provides strong incentives in career advancement.

While positive incentives have become more salient over time, monitoring and disciplinary measures have never disappeared and also increased their importance. The Chinese government has developed elaborate practices in monitoring and inspecting policy implementation. In the early days of the family planning policy implementation, for example, there were monthly reports, quarterly statistical data collection, mid-year inspection, and annual evaluation and ranking. Indeed, responses to the inspection and evaluation process have become some of the most daunting events in local governments (Lian & Chen, Reference Lian and Chen2023; Zhao, Reference Zhao2010).

Associated with these inspections and evaluations are negative incentives in the form of severe penalties, such as the ‘one-item veto’ (一票否决) policy or similar measures (i.e., the so-called ‘last place elimination system’ 末位淘汰制) that penalize those who fail to meet key task targets. That is, if the key tasks were not met, the performance of the office in other aspects would not be recognized, and the officials would be penalized. The key performance indicator (KPI) contents vary depending on the policy goals set by higher authorities, from family planning, GDP growth to ‘social stability’, or environmental regulation, and so on. As a result, ‘meeting the targets’ or ‘getting things done’ has become a primary driving force for local officials in their everyday work, and the entire bureaucratic machine becomes highly mobilized in response to top-down directives.

Implications for Organization Theory

The emphasis on incentive design in Chinese bureaucracy and the research that tries to make sense of it were influenced by the rise of economic theory of incentives and the new public administration theory (Graeber, Reference Graeber2016) since the late 1970s. No other place than Chinese bureaucracy has experimented with incentive designs in such a variety of ways, on such a large scale, and over such a long period. In a sense, Chinese bureaucracy has offered a variety of field experiments in incentive designs and practices in different arenas and across different localities, with different consequences. We highlight some emerging issues in this area.

First, contrary to the argument for weak incentive provision in public administration, Chinese practice in the form of the tournament model and the ‘one-item veto’ rule provide strong incentives in the bureaucracy. These incentive mechanisms have been argued to account for the active role of local governments in China’s economic growth and urban development. Now that the post-Mao era has gone through the centralization-decentralization-recentralization cycle, it is time to re-assess this set of phenomena: What are the consequences, anticipated or unanticipated, for officials as well as those being affected by their behaviors? How do different incentive mechanisms, disciplinary, monitoring, and evaluation measures, interact? Government performance and organizational response to crises in recent years suggest that the verdict is still out on the proper role of governments in public goods provision.

Second, bureaucratic practice in China raises issues about the boundary of organizations and the conceptualization of principal-agent relationships. While most studies of Chinese bureaucracy focus on a specific organizational setting – a township government, or an office in the public health area – the larger issue is how to conceptualize the boundaries of the organization, i.e., the relationship between the bureaucratic office under study and its higher authorities. One conceptualization is to see the specific office under study as part of the larger bureaucratic organization, where the relationship between the two is intraorganizational, with authority relationships, information asymmetry, and incentive design evaluated within organizational boundaries. Alternatively, one can treat the higher authorities, including the central government, as the organizational environment to which the bureaucratic office under study must adapt. These two conceptualizations capture different aspects of Chinese bureaucracy, with different implications for interpreting principal-agent relationships.

For example, if one treats the entire bureaucracy as an organization, with the central government as the principal, then unauthorized discretion at the local level can be seen as agency problems writ large. However, if one conceptualizes local governments (e.g., a county, a township, a state-owned firm) as the boundary of the organization, with the local authority as the principal and lower-level officials as the agent, then we can interpret official behaviors as the principal and the agent coordinating their behaviors in response to the external environment imposed by the higher authority (including the central government). In this view, collusive behaviors among officials at local levels may be interpreted as local initiative and adaptive flexibility, rather than agency problems.

Finally, several studies question the empirical evidence of the tournament model and emphasize instead the importance of patronage networks (Jiang, Reference Jiang2018; Shih, Adolph, & Liu, Reference Shih, Adolph and Liu2012). Earlier findings based on the tournament model are questioned upon empirical scrutiny (Miebe, Reference Miebe2024). Despite these objections, incentive designs and measures for lateral competition among officials are widely observed and adopted, such as KPI, the practice of rank-ordering performance along certain designated tasks (GDP growth, foreign investment, fertility rate, environmental regulation indicators, etc.). How do we make sense of the fact that theory, empirical observations, and quantitative analyses are at odds with one another in this case?

This observed tension provides an occasion for theoretical development. To explore its implications, we propose this hypothesis: the tournament principle in incentive design is adopted and becomes popular because of its role in lowering the transaction costs of signaling and evaluation as well as increasing the strength of incentive in a centralized bureaucratic system. In organizational practice, however, informal, social relations in the bureaucratic process overtake these design principles. That is, those with decision power make use of formal institutions (the incentive design) to sponsor their protégés (Lu & Zhou, Reference Lu and Zhou2023). Empirically, then, we observe both the active adoption of the rules of tournament competition and the widespread practice of patronage networks. Seen in this light, there is decoupling between symbolic compliance with the rational organizational design and actual behaviors in implementation, resulting in the dualism of the tournament principles in formality but patronage network in practice.

The preceding discussion on agency problems and organizational responses points to several important mechanisms – the role of informal institutions such as patronage networks, and loose coupling among policymaking and implementation, which we will discuss below.

Intellectual Background: Informal Institutions Through a Comparative Lens

Informal social relations are always present in organizations, and their relationships with formal authority vary with organizational settings, leading to their different roles in different contexts. Before we focus on the Chinese setting, let us first consider broader issues of social relations in organizations.

Research on social relations was a central focus in classical, sociological studies of formal organizations, from small groups within bureaucracies (Blau, Reference Blau1963), industrial organizations (Gouldner, Reference Gouldner1964), to government bureaucracies (Merton, Reference Merton1952; Scott & Davis, Reference Scott and Davis2007). A major theme emerging in these early studies was the prevalence of informal relations that significantly deviate from the rational image of the Weberian bureaucracy. Such social interactions create shared experiences and understanding, hence expectations and norms, which may give rise to organizational culture (Kreps, Reference Kreps, Alt and Shepsle1990), but they may also run counter to formal institutions and undermine authority relationships in organizations. In either case, they are often seen as part of the cultural toolkit in problem-solving (Swidler, Reference Swidler1986).

Turning to the role of social relations in China, one main theme is about the nature of guanxi (Chen et al., Reference Chen, Chen and Huang2013). Specifically, is guanxi a concept unique to Chinese culture, infused with distinctive cultural elements (Li, Zhou, Zhou, & Yang, Reference Li, Zhou, Zhou and Yang2019), or is it a more general network characteristic that finds different expressions in other cultural contexts (Burt & Burzynska, Reference Burt and Burzynska2017)? What kinds of behavioral patterns are captured in the guanxi concept?

In a recent book devoted to guanxi, Bian (Reference Bian2019) provides the following definition: ‘Guanxi is simply a connection between two individuals, but importantly it is a personalized connection, a subjectively close connection, and a potentially resourceful connection’. In his further elaboration, Bian explains that the connection may be kin or non-kin based, and that only a portion of kin or non-kin ties are elevated to personalized, close ties. ‘At its basic level, guanxi refers to a dyadic, particular, and sentimental tie that has the potential for facilitating the exchange of favors between the two parties connected by the tie’. Bian points out that the dyadic relationship can be extended to other actors, leading to a guanxi network.

To take a fresh look at the guanxi issue, a historical detour is helpful here. While social relations in general and guanxi specifically are well recognized today, this was not the case in the early days of the Mao era. The earlier paradigms of research on China, similar to those of Soviet studies, were characterized by models of authoritarianism and totalitarianism. In these theoretical images, there was little room for informal social relations in socialist China.

Mayfair Yang’s (Reference Yang1994) work was an early effort to call attention to guanxi in the state socialist society of China. In retrospect, her work was part of the larger paradigm shift in the 1980s when scholars’ close observations found considerable deviations from the images of totalitarianism and authoritarianism (Gold, Reference Gold1985; Shirk, Reference Shirk1982). Models of new traditionalism (Walder, Reference Walder1986) and fragmented authoritarianism developed theoretical arguments on the importance of informal social relations infused in authority relationships in the workplace and within and between government offices. These efforts led to the recognition of the salience of informal social relations, often based on bureaucratic apparatus, as an important organizing mechanism.

Earlier research on guanxi was especially sensitive to the political settings in which guanxi evolved and connected to the socialist hierarchy (Walder, Reference Walder1986; Yan, Reference Yan1996; Yang, Reference Yang1994). In recent years, research on guanxi in sociology and organization studies has shifted to interpersonal, or interorganziational, relationships, focusing heavily on instrumental ties in resource exchange, as shown in studies at MOR cited before.

We want to bring politics back to the study of guanxi, as the two are intricately intertwined in Chinese bureaucracy. In so doing, we highlight (1) the role of guanxi as political resources and (2) guanxi as the cultural script that frames interaction patterns in the political arena. On this basis, we consider their implications for organization theory.

Guanxi in and Around Chinese Bureaucracy

Similar to the early organizational sociologists who discovered the limits of Weberian bureaucracy, scholars studying bureaucratic behaviors of Chinese governments found a variety of institutional practices that are at odds with the images of the Leninist party-state.

For example, as Wang, Liu, and Sun (Reference Wang, Liu and Sun1997) and Sun and Guo (Reference Sun and Guo2000) show, in the procurement of grains and tax collection in rural areas, township government cadres used social relations, such as kin relations and friendship, to persuade reluctant villagers to pay agricultural taxes. The authors labeled this pattern as the ‘biantong’ (adaptive flexibility) strategy that uses informal social relations to accomplish the goals of formal institutions (i.e., tax collection).

As it turns out, this behavioral pattern is widely practiced in different areas and different localities. As many studies have shown, informal social relations are rampant in all kinds of bureaucratic settings, just like guanxi is everywhere in Chinese society. Ying’s (Reference Ying2001) study shows that social relations were used to put pressure on those who tried to challenge local authorities. Guanxi is often used by local authorities to overcome the resistance of villagers and local residents in policy implementation (Deng & O’Brien, Reference Deng and O’Brien2013; Lee & Zhang, Reference Lee and Zhang2013). The extension of relational repression is the other side of biantong, that is, the selective use of informal relations to accomplish official tasks.

Moreover, guanxi is intertwined with bureaucratic behaviors and response strategies based on stable organizations and authority relationships therein, leading to the class of collusive behaviors among local officials (Zhou, Reference Zhou2010). In other words, guanxi is not merely interpersonal but is based on stable sponsorship and patronage ties in organizations. Whereas guanxi has been mostly seen as personalized exchanges for private favors, it has been widely used in organizational practice for the exchange of favors for official tasks across offices, such as the transfer of resources across areas to meet policy targets. Social relations are activated in response to government inspections, coordination among offices, mobilization for resources to get things done, and in resolving local problems (Ai, Reference Ai2011; Wu, Reference Wu2007).

Significance for Organization Theory

Informal social relations are integral parts of organizations, but their prevalence in Chinese bureaucracy is striking given that, over the years, Chinese governments have undergone a series of reforms establishing elaborate rules and regulations of bureaucratic behaviors. Research on guanxi in Chinese bureaucracy has considerably expanded our understanding of the scope and significance of social relations and has raised important issues for organization theory: What are the sources of informal institutions in Chinese bureaucracy? What are the implications of the salient role of informal institutions for organization theory?

The preceding discussion contrasts sharply with the foundation of formal rules and routines in the Weberian bureaucracy, as part of the rationalization processes in the West, where the Roman law-based legal rational authority has provided the institutional foundation. Organization theory developed in this intellectual context has emphasized the importance of institutionalization processes that shape corporate as well as individual behaviors (Meyer & Rowan, Reference Meyer and Rowan1977). In contrast, informal institutions are often treated as a residual factor and are relegated to an overlooked corner in organization theory in this context. In other words, organization theory developed in the Western intellectual environment has not given adequate attention to alternative organizational mechanisms and practices embedded in different cultural scripts.

Here belies the opportunity for the development of organization theory. Research on Chinese bureaucracy highlights the prevailing role of informal institutions, where the interplay between formal and informal institutions is central to the logic of governance in China (Zhou, Reference Zhou2019). In a political system of centralized authority like China’s, ironically, the extent of the centralization of authority creates equally strong pressures for flexibility in local adaptation. Unlike the legal-rational mode of organizing in the Weberian bureaucracy, in China such flexibility occurs not in changes in formal rules, procedures, or policy targets, but in the form of informal practices. That is, while formal institutions remain stable and rigid, informal institutions and practice permeate to allow considerable variations in selective implementation, interpretation and flexible adaptation at local levels, as documented in a large literature cited above.

The recognition of the role of informal institutions in Chinese bureaucracy points to alternative mechanisms of organizing in a comparative perspective. Bureaucratic practices reveal the endogenous perpetuation and reproduction of guanxi, which is embedded in those cultural patterns characteristic of ‘the mode of differential association’. That is, as sociologist Fei Xiaotong put it: ‘[The Chinese pattern of social interaction] is like the circles that appear on the surface of a lake when a rock is thrown into it. Everyone stands at the center of the circles produced by his or her own social influence. Everyone’s circles are interrelated’ (Fei, 1992[Reference Fei1948]). In this picture, strengths of social relations vary with the social distance between the ego-center and the specific person/group, thereby presenting a distinctive organizing mechanism. By resorting to guanxi for local problem-solving – such as social order maintenance, tax collection, crisis management, and getting things done – these bureaucratic practices give legitimacy to informal social institutions and reinforce awareness and identity based on such social institutions. As a result, they wittingly or unwittingly reinforce existing social relations that may be used for private interests and collective mobilization on other occasions (Feng, Reference Feng2010; Wang, Reference Wang2024).

The prevalence of informal institutions in Chinese bureaucracy points to an alternative model of institutionalization. In China, guanxi has become a social institution; that is, the informal, social relations are institutionalized in the everyday practice in Chinese bureaucracy and serves as the cultural script by which collusive behaviors and coordination take place. Instead of the formal institutions gradually replacing the informal, both formal and informal institutions coexist and coevolve, inducing distinctive organizational dynamics.

From Rational Design to Loose-Coupling

Research on Chinese bureaucracy provides an ideal setting to examine these questions. Contrary to the image of a tightly coupled Leninist party organization, a large literature on bureaucratic behaviors has demonstrated different aspects of loose coupling in Chinese bureaucracy: between policymaking and implementation (flexible, selective, and collusive), between symbolic compliance and actual behavior (formalism), between formal structure and informal behavior (extrabudgetary resource transfer), between rules and practice (collusion), and considerable variations over time.

Loose coupling phenomena manifest themselves across a variety of organizations, but these observations in Chinese bureaucracy are especially intriguing: Why is a centralized organization at the same time characterized by widespread loose coupling in its practice? What are the mechanisms and processes that sustain and reinforce loose coupling practices? Research on Chinese bureaucracy provides many clues for us to put together to solve this puzzle.

One salient aspect of the loosely coupled system is the relationship between centralization in policymaking and flexibility in local implementation. The fundamental tension highlights the defining characteristic of governance in China: The state has a tendency to centralize decision rights upward through the political control of both resources and personnel. Nevertheless, given the vast territories and the huge diversity therein, the top-down policies cannot fit diverse situations across localities. There have been recurrent struggles between central and local governments, in different forms and at different times, regarding decision rights over resource allocation, personnel management, and other issues. These tensions and ensuing crises often force the central authority to allow or give acquiescence to flexible adaptation at local levels. The expression that described the Chinese emperor’s strategy in the implementation of taxation policies – ‘stay firm in principles; be flexible in implementation’ (yuanzeshang burangbu, shishishang bujianchi 原则上不让步, 实施上不坚持) (Wang, Reference Wang1981[1948]) – vividly captures the dialectic relationship between policymaking and policy implementation in China’s governance. These tensions and compromises create the momentum and space for loose coupling between the central and local governments. Indeed, the so-called ‘two-track politics’ (双轨政治) in history as described in Fei (Reference Fei2012) is a prototypical loose-coupling structure where the bureaucratic track stopped at the county-level, beyond which rural institutions took over, with the two tracks institutionally intertwined (through personnel and cultural values) but organizationally loosely coupled (Zhou, Reference Zhou2019).

Another aspect is the relationship between symbolic compliance and behavioral autonomy. Local officials engage in symbolic compliance in order to protect their behavioral autonomy and local adaptation (Kejun, Reference Kejun2014). Symbolic compliance reinforces the value system that legitimizes and sustains formal authority. Ironically, efforts by the centralized authority to strengthen formal institutions also provide the impetus for local officials to cultivate social relations and signal symbolic compliance as strategies for self-protection. In so doing, symbolic compliance alleviates the top-down pressures that intrude into the autonomy of local authority.

We want to distinguish loose coupling from decentralization. Cycles of centralization and decentralization have long characterized in China’s governance practice, which involve the reallocation of decision rights at different levels of the bureaucracy and policy reorientation. In contrast, loose coupling represents stable institutional arrangements that preserve the identities of the different elements or parts of the organization. One may have cycles of centralization and decentralization in administrative policies and advocacies, but loose coupling is more or less preserved throughout those fundamental institutional practices and in informal institutions. In other words, loose coupling provides both the stable basis for cycles of centralization and decentralization, and the very impetus that induces shifts across different phases in these cycles.

An important development in theorizing about loose coupling in organizations is the emphasis on the dialectic relationships between loose-coupling and tight-coupling. Orton and Karl (Reference Orton and Karl1990: 204–05) wrote:

A loosely coupled system is not static; rather, the very notion of loose coupling implies that it can be mobilized into a tightly coupled system temporarily. Tight coupling in one part of the system can occur only if there is loose coupling in another part of the system (Weick, Reference Weick1976: 10). Tight coupling in certification (professionalization) may lead to loose coupling in inspection (bureaucratization), as in the case of professional organizations (e.g., hospitals, universities, etc.).

Structurally, loose coupling in Chinese bureaucracy takes places across levels of the hierarchical structure or among administrative jurisdictions as well as along the dual tiao-kuai lines. We expect that tight coupling along the vertical, functional authority leads to more efforts at loose coupling by the territorial authority in the lateral territorially based jurisdictions. Research shows that the acceleration of downward pressures (cengceng jiama 层层加码) is often accompanied by greater extent of biantong (adaptive flexibility) and collusion (Chen, Zhang, & Hu, Reference Chen, Zhang and Hu2015; Zhe & Chen, Reference Zhe and Chen2011).

To extend this analogy, tight coupling in one period of time may lead to loose coupling in another period. There are times when different parts are tightly coupled through campaign-style mobilization, resource mobilization, and incentive design. The case of environmental regulation inspection over the 5-year plan cycle illustrates this point well (Zhou, Lian, Ortolano, & Yinyu, Reference Zhou, Lian, Ortolano and Yinyu2013): Unexpected progress in meeting targets in the early years by local bureaus led to considerable loose coupling in their performance evaluation and incentive allocation in the subsequent years. These observations caution us to look into those behavioral practices, those carefully hidden areas, and the dynamic processes for the evidence of loose coupling.

From Loose Coupling to Variable Coupling

Following Weick’s advocacy of dialectics in organizational coupling, we propose an alternative model of Chinese bureaucracy as a variable-coupling system to make sense of the observed organizational processes and behavioral patterns. That is, both tight coupling and loose coupling may coexist but in different parts of the system or they become salient or hidden at different points in time. In this sense, loose coupling implies variable coupling. The extent of coupling among different parts of Chinese bureaucracy varies over time and across areas and shifts in predictable ways driven by recognizable organizational mechanisms.

We argue that loose coupling in Chinese bureaucracy is built on stable institutional processes and mechanisms – it is prevalent today, it had existed in the Mao era, and it was deeply rooted in the processes of state building in Chinese history. Indeed, the turbulent political campaigns in the People’s Republic of China were partly the top leaders’ efforts to combat bureaucratic phenomena associated with loose coupling.

Underpinning these relationships is a set of organizational mechanisms that generate the dialectics in loose coupling. We consider some of these aspects. First, organizational attention allocation is an important regulatory mechanism here. Attention to one activity is attention away from other activities, this is especially so in a system of centralized authority, as different parts must respond to the directives at the very top. As a result, mobilization to focus on one target induces loose coupling on others. Consider campaign-style mobilization, which is in effect the mobilization of attention of different offices for the designated policy target (i.e., family planning, GDP growth, etc.). But the more intense the mobilization in one area, the greater the extent of loose coupling in other areas. As a result, mobilizing attention to one area incurs the cost of generating loose coupling in others.

Second, one unintended consequence of campaign-style mobilization is the cumulation of organizational slack. As essential work must be maintained even when political campaigns direct attention to other areas, it is not surprising that an office is likely to add extra personnel in response, just like the storage of spare parts in the face of frequent breakdowns. In other words, organizational slack and redundant resources provide a loosely coupled system and they serve as buffers to the top-down mobilizational pressures, a similar pattern observed in the shortage economy in socialist societies (Kornai, Reference Kornai1980).

Third, another mechanism is so-called formalism (xingshi zhuyi 形式主义), that is, lower-level offices show symbolic compliance but decouple their behaviors from higher-level pressures. Considerable attention is given to the recurrent theme of formalism and bureaucratism in Chinese bureaucracy. In many ways, this is one response strategy that produces a loosely coupled system in that the top-down pressures are deflected by formal gestures. In this sense, symbolic compliance and local autonomy coexist and reinforce each other. From the loose coupling perspective, formalism may be interpreted as a form of symbolic compliance that allows behavioral loose coupling and, in this sense, it is a form of strategic response as the weapon of the weak.

The variable coupling phenomena have broad implications in other organizational contexts. As Mahoney and Kathleen (Reference Mahoney, Kathleen, Mahoney and Thelen2009: 4) observed, policy implementation often generates a process of what they called ‘variable compliance’. Such informal practices are significant because ‘institutional change often occurs precisely when problems of rule interpretation and enforcement open up space for actors to implement existing rules in new ways’. Moreover, variable compliance may produce institutional change, or it may facilitate institutional durability through behavioral flexibility under authoritarian rule.

As we can see, the image portrayed in the model of fragmented authoritarianism fits the variable coupling model well: both capture the variable relationships between the central and local governments, and across levels of bureaucracy. Moving from the image of fragmented authoritarianism to that of variable coupling allows us to situate Chinese bureaucracy in a larger comparative perspective and subject it to organizational analysis. We can make sense of these paradoxical bureaucratic phenomena in light of stable institutional arrangements and organizational processes based on both formal and informal institutions.

The observed variable coupling is not based on rational design; rather, they are often haphazard responses to unexpected crises and policy failures. Therefore, we need to pay attention to the actual behaviors of bureaucrats in action rather than follow those theoretical models that focus on formal institutions or official rhetoric. To understand variable coupling phenomena, we need to consider those longer processes and dynamics that involve interactions among those multiple parties, how these interactions lead to shifts in policies over different stages. We need theories that can better capture the dialectic aspects of the organizational processes, explicate how the interaction of multiple processes generates observed patterns, and pay attention to the unintended consequences, especially what are the benefits and costs, from whose perspective, in these dynamics.