Flexibility and Authority: Resolving Labor Disputes in A County Government in Western China

Abstract

Existing literature regards flexibility and authority as key characteristics of informal justice. We further contend that the combination of the two is crucial for informal justice to be effective. We investigate the process of dispute resolution by a Chinese labor agency. Following the life cycles of a sample of 810 labor disputes, we find that this informal justice forum was efficient and effective, made possible by the combination of flexibility and authority. Flexibility means that the agency attracts certain types of cases that are usually screened out of the formal legal system and that agency officials use “informal,” hence flexible, techniques. Authority means that the administrative agency possesses additional powers over the disputants; hence, the disputants are under pressure to follow its suggestions and decisions. A comparative analysis of various cases of informal justice reinforces the importance of combining flexibility and authority. We further demonstrate that flexibility without authority is insufficient and that some informal justice forums are effective because they enjoy both.

Over the last decade, China has witnessed a resurgence of mediation in dispute resolution. Institutionally, many agencies, such as government bureaus, community organizations, and courts, have been called upon to proactively mediate disputes, resulting in “the Grand Mediation.” Although adjudication and arbitration continue to be utilized, mediation has been aggressively pursued, and has become the preferred method of the Chinese regime. According to official statistics, 63 percent of the 1,332,000 labor disputes in 2012 were settled by mediation (Labournet.com.cn 2012).

Is mediation an effective means of dispute resolution? Answers vary based on different perspectives and data. Fieldwork from rural China, for example, Xiong (2014), has shown that judicial mediation remains effective and important. Focusing on medical disputes, Ding (2015) contends that mediation is “a dose to cure ‘medical chaos’.” Huang (2015) suggests that securities dispute mediation has the potential to become a successful forum. Conversely, the literature is critical of this revived mediation. Minzner (2011) dubs this resurgence a “turn against law,” which in the long run undermines the legal spirit. A recent study (He 2017) demonstrates that, in divorce cases, judicial mediation from the perspective of stability concerns has undermined legal principles. Some (Ng and He 2014) argue that there are internal contradictions in judicial mediation. When judges conduct mediation, role conflicts are widely registered (Chen 2015; Xian 2015). Judges, especially those who receive formal legal education, seem resistant to the renewed mediation (Fu and Cullen 2011). As a result, many call for a “rethinking” of the mediation campaign (Zhang 2015; Zhao 2015a, 2015b).

However, few studies have systematically assessed the effectiveness of mediation. We all know that from the prereform era to the mid-1980s, mediation was the most important form of dispute resolution for the so-called “internal conflicts among the people” in China (Lubman 1999). It then declined in the 1990s. The political ideology, personnel, organizational capacity, and supporting systems that made mediation effective in the prereform period have either vanished or eroded (Halegua 2005). Will the revived mediation be different?

Drawing on archival records from a commonly used forum known as the Labor Inspection Brigade (LIB) in Western China, supplemented with interviews of relevant officials and complainants, and secondary literature, this article examines the case flow, assesses its effectiveness, or lack thereof, and explores institutional conditions. Given the vast size of China and its immense regional variation, data from a single county over only 2 years do not provide a comprehensive and accurate picture for the country as a whole. It will, however, allow us to track the trajectories of a group of labor disputes in China from beginning to end, including both those that reach a formal legal forum and those that do not. We try to determine how many of these are resolved and by which channels, the extent to which they are resolved, and how their nature affects their final resolutions. We also assess the level of their effectiveness. Comparing our findings with other dispute resolution forums, we examine how the effectiveness of a particular forum is related to its social and political setting.

We argue that the combination of flexibility and authority is crucial in order for informal justice to be effective. The existing literature has long noticed that flexibility and authority are the characteristics of informal or popular justice (Abel 1982a; Merry and Milner 1993). Its emergence is a response to the complex and rigid formality of the formal court system. Authority is also regarded as important in informal or popular justice (Abel 1982a; Merry 1992). The existing studies, when evaluating the effectiveness of informal justice, tend to discuss the two factors separately and discretely. Rarely have they been put together to assess their effectiveness. For example, Abel (1982a) argues that authority is crucial to the effectiveness of informal justice forums. In a classic study, Silbey and Merry (1986) contend that the strategies of successful mediators fall into four categories: control of the process, control of the substantive issues, presentation of self and the program, and activation of norms. Implicitly, these strategies echo our argument that flexibility and authority are crucial to effective dispute resolution. Both control of the process and control of the substantive issues occur only when flexibility permits. The rigidity of procedural rules narrows the room to manipulate either the substantive issues or procedural processes. The self-introduction of the mediator, according to Silbey and Merry (1986: 12), claims either expert knowledge or legal authority (emphasis added). Finally, only in a flexible institutional forum can mediators reach beyond legal norms, and activate moral and therapeutic norms.

By studying dispute channels provided by a noncourt government agency, we look into situations where the disputes may otherwise not be heard. It is in this sense that we use the term “effectiveness,” or, occasionally, “success.” To the extent that a large number of disputes enjoy third-party resolution, we regard the system to be more “effective” than others. This success is no easy feat, as the majority of the disputes end up without any recourse at all, as the dispute pyramid implies (Felstiner et al. 1980-1981). From the standpoint of complainants, the concept of “effectiveness” is straightforward. Another dimension of effectiveness, however, is from the governing point of view. The government agency at the center of this study has been created as a watchdog to ensure fair labor practices. It exists in the context of the overarching concern for social stability, a daunting challenge for governance. Those small claims are not ignored, but somehow obtain a resolution, will comfort an otherwise grievous labor force. Small or even trivial disputes, when unattended, could escalate into large troubles. Therefore, it is “effective” governance if the system has a way to deal with small claims. Our concept of effectiveness or success, however, is silent on the resolution quality. Is the agency's intervention appropriate according to the merit of the case and literally by the labor law? Are there disputes with no merit to start with, hence any third-party intervention is waste of resources? These are important questions, but they are outside the scope of this study. Here, we focus on access to justice, as the complainants are mostly underprivileged migrant workers.

1. Informal Justice's Effectiveness

Alternatives to the formal court system have long existed. In theorized concepts such as “informal justice” (Abel 1982a) and “popular justice” (Merry 1992; Merry and Milner 1993), the driving spirit tends to be “unofficial (dissociated from state power), noncoercive (dependent on rhetoric rather than force), nonbureaucratic, decentralized, relatively undifferentiated, and nonprofessional; substantive and procedural rules are imprecise, unwritten, democratic, flexible, ad hoc, and particularistic” (Abel 1982a: 2). According to Merry and Milner (1993: 32), “popular justice is a process for making decisions and compelling compliance to a set of rules that is relatively informal in ritual and decorum, nonprofessional in language and personnel, local in scope, and limited in jurisdiction.” Although alternative dispute resolution (ADR), informal justice, and popular justice have different emphases, underlying their conceptions and practices are the flexible procedural rules on both the acceptance and processing of disputes in a nonprofessional context.

Flexibility is thus a built-in feature of informal justice. Contrary to complexity, the characteristic attributed to the formal court system, flexibility has been regarded as critical to informal justice. Indeed, the contemporary advocacy of informal justice is built on two interrelated criticisms of the court system in common law jurisdictions. One is its adversarial mode of handling the contesting parties; the other is its complexity. The adversarial mode leads to overcautious measures to protect individual rights, giving rise to elaborate procedures and intricate channels (Kagan 2003). Inefficacy, including case overload, gridlock, and delays, results. The legal system becomes inaccessible to the poor and disadvantaged. In the United Kingdom, the influential Woolf Report focused on “improving access to justice, reducing the cost of litigation, and removing unnecessary complexity” (Zuckerman 1996).

But are these informal forums an effective remedy to the ills documented in the formal court system? Do they provide just and speedy relief to disputants? Is flexibility alone enough? According to Delgado (1988), at the height of the ADR movements of the 1980s, most assessments seemed positive. Later research, however, suggests that the forums might in fact serve the state better than the citizen users (Hofrichter 1982, Harrington 1985).

1.1 Indispensable Authority

In the existing literature, few have suggested that authority is crucial in evaluating both the success and failure of informal justice forums. One exception is Abel (1982b). Rejecting the nostalgia inspired by the demise of traditional forums, Abel asks, “what is the common bond between idealized images of nineteenth-century America, Eastern civilizations, and African tribal societies? It is respect for authority.” He adds, “Although such institutions are presented as forums for dispute settlement, they performed that role because they were, preeminently, loci of authority” (Abel 1982b: 275). He elaborates upon this idea with his analysis of the failure of several neighborhood justice forums. Unlike traditional townships and villages, the “neighborhood” in contemporary American cities is not an organic unit, and thus it lacks cultural infrastructure that generates community authority. Implied in his argument is that, without authority, flexibility alone seems inadequate to promise an effective informal justice forum. This is consistent with Silbey and Merry's (1986) study of mediators strategies. For successful mediators, the key strategy is to exert authority in a flexible setting.

As the rise of the contemporary nation–state, the pervasiveness of the state form structure across the world has effectively destroyed most traditional sources of authority such as customs, religion, and personal loyalty. An inference can be drawn from this observation that informal justice forums would be difficult to build outside the state. In other words, these forums may have a better chance of success and survival if and only if they are associated with the state in a meaningful way. Following this line of reasoning, one might expect mediations and arbitrations organized within the government's administrative agencies to have a better chance of being effective.

This assertion has been vindicated by the experiences of the revolutionary socialist states in which state authority permeated informal justice forums. In Castro's Cuba, initially, popular tribunals had popularly-elected lay judges, public discussion and critique of offenders, and much discretion for judges. As the state became more concerned with planning and order, popular tribunals were replaced by more sedate, professionalized, bureaucratized, and formalized courts (Salas 1983). In the Soviet Union, popular tribunals gradually became more tied to state law (Henry 1983). Similarly in China, from the 1960s to the late 1970s, local mediation centers became places for enunciating and applying state policy, rather than undermining state authority (Lubman 1967). Over time, popular justice in China has been entrenched in existing relations of state power.

1.2 How Will the Chinese Case Fare?

The focus here is on a Chinese administrative agency, the LIB, functioning as an informal dispute forum for labor grievances. Its nature is mainly administrative, but it is also involved as a third party in dispute resolution—that is, a noncourt agency may dispense “courtness,” defined by Shapiro (1981) as a triad involving two disputants calling upon a third party for assistance.

In light of the two factors—flexibility and authority—will a Chinese government-related forum be effective? The existing literature offers, at best, conflicting evidence on the issue. Such a forum could be ineffective as some evidence suggests that it lacks either or both of these factors. First, Chinese government agencies may not be flexible. They cannot escape the trappings of bureaucracy, and thus work slowly and inefficiently (Gallagher 2006; Lee 2007). Second, the government's authority may be limited. As separation of powers is neither norm nor practice, one may argue that Chinese government enjoys monopolized authority. But this does not mean that each branch of government enjoys absolute authority. Furthermore, the government is widely reported to be corrupt, and collusion abounds between local officials and business owners (Lee 2007). To maintain an investment environment, the government also heeds the interests of employers. The government rarely applies “legal sanctions when handling workers' complaints about employers' breaches of the law as they are afraid that this may hamper business” (Zhuang and Chen 2015: 389). Third, due to stability concerns that plague the regime, the agency only focuses on bigger cases that threaten to escalate into street protests, ignoring the smaller ones (Su and He 2010). Many minor but routine cases may not be taken seriously, and thus are poorly handled. In other words, the concern for stability may trump both flexibility and authority and become the decisive factor in determining the effectiveness of the forum.

But other evidence in the existing literature suggests exactly the opposite: government agencies indeed enjoy both flexibility and authority and thus may be able to effectively resolve disputes. First, many scholars argue that their handling of labor disputes is flexible and pragmatic (Thireau and Linshan 2008, Su and He 2010, He et al. 2013, Lee 2007, Zhuang and Chen 2015). This is especially so when social protest is involved (Chen and Xu 2011; Su and He 2010; Zou et al. 2015). As the Hu Jintao era, government agencies have been “service oriented.” The government tends to soften in labor dispute resolution, and this has been vindicated even inside the courts (Halegua 2008; Su and He 2010). Also, the Grand Mediation movement, while trying to resolve complicated and influential disputes by coordinating various government branches, has enhanced this flexibility (Hand 2011; Wang and Minzner 2015). In addition to flexibility, government agencies also enjoy a high level of authority, with administrative powers of the agencies as assets. As the regulator of the sector, the agency has the powers and the means to control employers. More generally, the authority of Chinese governments is high and people respect their decisions (Lai et al. 2010). When authority is combined with flexibility, it encourages more effective dispute settlement and resolution.

2. Labor Mediation and the LIB

With the death of Mao and the ensuing reforms, China employed laws as a means of social control and had moved toward formal legality and rights: numerous laws and regulations were put into effect, lawyers were trained, and a formal court system was established, along with other institutional reforms. Indeed, during most of the period in which Xiao Yang was the president of the Supreme People's Court (1999–2008), the judiciary emphasized formalism and initiated Western-style judicial reforms.

In terms of labor disputes, a four-stage system was set up in the mid-1990s: voluntary mediation, mandatory arbitration, civil lawsuits, and appeals (Zhuang and Chen 2015). The number of disputes channeled through labor arbitration and the courts had been increasing steadily. However, these legal venues were soon jammed. With the rapid economic development and transition in China, labor disputes skyrocketed. For example, from 1995 to 2006, the number of labor arbitrators doubled, but the number of labor disputes jumped nearly 50-fold (Ministry of Labor and Social Security 1996–2006). The legal complexity of the formal legal system only led to sluggishness and delays in the settlement process. In Guangdong province, it normally took a year for workers to proceed through the legal channel; 30 months to settle a labor dispute was typical, owing to delays caused by bureaucratic ineptitude (Li 2008). The inefficiencies and slow processes of legal proceedings often frustrated disgruntled workers, forcing them to either follow the petition route, or take to the streets (Chen and Xu 2010; He et al. 2013; Su and He 2010).

According to Di and Wu (2009: 240), “the dramatically increasing number of negative social incidents and the rapidly growing caseloads” pushed for the revival of mediation. Zhou Yongkang, then China's “law tsar,” and other leaders, believed that the Western-style rule of law reforms had gone too far (Wang and Minzner 2015). They believed that the reformed judiciary did not resolve disputes, even though many cases were now being adjudicated with a higher degree of procedural fairness. They also believed that the skyrocketing number of petitions (上访) and collective incidents was a result of the judiciary's incompetence.

The turn-about came in the early 2000s. In an opinion issued in 2002, the Communist Party endorsed the positions of the Ministry of Justice and the Supreme People's Court on strengthening mediation (Di and Wu 2009: 240). In October 2006, the Central Committee of the Chinese Communist Party issued its “Decisions concerning major questions in the building of a socialist harmonious society.” It had further called for the strengthening of mediation, in order to resolve disputes at the grassroots level and nip conflicts in the bud. To maintain social stability, the paramount concern of the regime, the policy response has been to renew its reliance on mediation across all the sections of society. The purpose is to resolve disputes as quickly and as early as possible, to prevent them from escalating into incidents that could threaten social stability (cf. Hand 2011; Minzner 2011). Mediation thus again became an overarching theme in every section of social organizations and governments.

This trend was also conspicuous in other bureaucratic agencies involved in dispute resolution. In December 2007, the new Law on Mediation and Arbitration of Labor Disputes was promulgated, which reinforced the government's role in mediating labor disputes. According to the law, “triple mediations” (i.e., civil, administrative, and judicial) should be conducted before arbitration and adjudication.

Only in this context can one understand the role of the LIB in resolving labor disputes. The LIB is one subsidiary of the labor bureau—the government branch that regulates labor relations. The responsibilities of the LIB are mainly to, through inspections, oversee companies, contractors, and small business owners on issues regarding safety, the environment, and employment contracts. Dealing with labor-related complaints had primarily fallen within the domain of another subsidiary—the arbitration committee. As the revival of the mediation policy, however, the LIB has become one of the leading offices in handling labor disputes; it promotes its role in resolving labor disputes through various channels. Posters proclaiming that the LIB provides free and efficient means to collect unpaid wages and to regulate harmful working environments have been hung not only in the offices of the LIB, but also at construction sites and in the streets. These posters provide a hotline for potential disputants to report their grievances. With its changed role, the LIB hopes to eliminate any possibility that minor disputes are transformed into social protest, suicides, or petitions, all of which are categorized as social instability incidents.

Procedurally, the LIB is informal and flexible. Although visitors must state their identity, they do not have to submit written complaints. Indeed, many complaints are not filed in person, but rather by telephone or e-mail. LIB officials usually accept every complaint submitted. They do not have to match them to different categories; collective cases or cases with significant influence are filed in a special folder. Nor is there any need for the complainant to provide any evidence. Although the officials ask what happened, they take the complainant at his or her word. According to our interview with an LIB official, “most workers there are trustworthy. They have no incentive to come here and say something baseless.” Once a complaint is filed, the officials respond to the complainant, usually after contacting the party toward whom the complaint was targeted. Of course, there is no guarantee that every complaint will be addressed.

The LIB has thus become an informal channel, in addition (and parallel) to the formal channel. Officially, the formal resolution process in labor disputes consists of three methods: mediation conducted by a state enterprise's labor dispute mediation committee; arbitration conducted by the arbitration center in the labor bureau; and litigation (i.e., filing suits in court). Legally, only when one party is unsatisfied with the decision of the arbitration center can a suit be filed in court.

In achieving the goal of social stability or social control, the LIB stands with the existing formal legal channel. Despite apparent differences in the duties and responsibilities of the two branches, they often overlap. The LIB thus often intervenes in ongoing conflicts, or acts as the mediator. There is also much crossover from one body to the other, and a case may be referred back to the LIB when a plaintiff in the arbitration committee refuses to carry on after realizing the process is too protracted and expensive. For example, the LIB, where we conducted our fieldwork, has taken on far more cases than the formal channel. In 2005, it registered more than 400 cases, whereas the county courts took less than 50 labor cases. Of course, the cases the LIB registered might have been smaller and more casual, whereas the cases with which the court dealt could have been more serious. Nonetheless, the changed role of the LIB is a response to the perceived failure of the formal legal system.

3. Methods

Studying dispute resolution in the Chinese government presents tremendous methodological challenges. For the LIB, labor issues, especially those related to social protest, collective action, and petitioning, remain sensitive. Any mishandling may lead to incidents which could threaten social stability. Both the LIB and its officials are thus wary of any inquiry from researchers outside of Mainland China. This is also why the existing literature relies primarily on sporadic interviews, anecdotes, survey data, and news reports.

To overcome these challenges, we have adopted unconventional methods. Having gained the trust of several LIB officials from our previous projects, we were welcome to conduct fieldwork investigations in County Z during 2011 and 2012. During our empirical work, we stressed that our research was anonymous and that any details that could leak the identity of the informants would be removed, so as to ensure the authenticity of our data. We were allowed to sit beside the hosting officials and silently observe how they handled disputes. We also had casual talks with the complainants at their convenience, both before and after they arrived at the office. A total of 21 petitioners agreed to provide information for this study. In addition, we managed to interview the hosting officials after their sessions had ended, so as to understand their approaches toward the disputes. Five officials agreed to such interviews, and the interviews lasted between 30 min and 1 h. Lastly, we have also interviewed seven employers—the targets of the complaints. They explained to us the reasons why they had either not paid, or had decided to pay, and when this had been requested by the LIB officials.

County Z is located in Sichuan Province in China's Western heartland, far from the booming coastal cities leading China' economic development. Over 75 percent of its 780,000-plus population is rural, suggesting that agriculture has remained the pillar of the local economy. According to official statistics, County Z's GDP per capita was only 9975 yuan (1200 USD) in 2005, about one third of the more developed coastal areas. The region's underdevelopment has led the central government to mount the “Grand Western Development” campaign to attract high-profile investment projects. At the time of our fieldwork, the county seat was experiencing a construction boom. The construction workers were farmers who had come from villages both within the county and from surrounding areas.

Although the construction sector provided migrant workers with job opportunities, it was poorly regulated. Stories of unpaid wages were routine. The situation of wage earners here was particularly precarious, compared with their counterparts in coastal regions. Migrant workers went to the county government office with wage disputes almost daily.

During our fieldwork in County Z, the formerly intimidating government compound was open to laborers. They passed the guards without being stopped, and proceeded straight to the fourth floor where the LIB was located. Some of their quests for justice eventually might reach the Arbitration Committee and even the courthouse, but the with an LIB official was the starting point. The first encounter typically lasted between 15 min and half hour. Based on our observations, if the amount in dispute was tiny and straightforward, officials would take action directly, without recording anything. Only when the officials felt a written record was necessary would they then require the complainants to provide written documentation, and usually ask them to fill out a standardized form. The form included the identity and contact information of the complainants and the other parties, as well as a brief outline of the disputed issue. If the complainants were illiterate or had poor writing skills, the officials would fill out the forms for them while simultaneously raising questions. One incident, instead of one complainant, constituted a form. As a result, one form might include a batch of complainants against a single employer. According to some LIB officials, the recorded disputes were only 60 percent of the total disputes reported to the LIB. At year's end, these forms, collected in the daily log of complaints (台账) of the LIB, were bound and archived. As the LIB is an informal dispute resolution agency, it is understandable that some information was missing from the daily log. Laborers would occasionally not fill in all the blanks, or the dispute would be swiftly resolved, rendering the recording of information unnecessary. Some disputes might have reached beyond the capabilities of the LIB, and were thus unrecorded.

Toward the end of our fieldwork investigations, the local officials kindly agreed to make available to us the daily log of complaints in 2004 and 2005. There were 810 disputes in total. In the article, we refer to these 810 disputes registered in the LIB as “the daily log sample” (see Tables 1 and 2). With all pages running continuously and consistent with the content at the beginning, it was clear that no cases had been removed due to their sensitivity or other concerns. Six years had passed since the disputes had been filed, which may have desensitized the records, and this could have been why the officials felt comfortable granting access. If we were to have chosen the dates, we could have chosen more recent years to conduct our research. Nonetheless, the 6-year gap between the recorded disputes and our fieldwork investigation should not have affected our overall analysis, because the national policy on dispute resolution has been oriented toward mediation since the early 2000s. If anything, the tendency toward outcome-based settlements was only more pronounced in 2011. The outcome would have been more effective, had the disputes been handled in 2011.

Table 1. Profile of Disputes Registered by LIB of County Z

Note: Data source—the daily log sample.

Table 2. Initial Processing of Labor Disputes by LIB

		Frequency	Percentage
Initial processing	Phone the petition target party	469	57.90
	Visit the company	49	6.05
	Ask the target party to explain	85	10.49
	Transfer case to other agency	39	4.81
	Legal channels suggested	10	1.23
	Did not get processed	19	2.35
	No information	139	17.16
	Total	810	100.00
Result of initial processing	Complete at LIB	379	46.79
	Transferred to other agency	39	4.81
	Next step suggested	207	25.56
	Not processed or pending	158	19.50
	Total	810	100.00

Note: Data source—the daily log sample.

More valuable for the sake of this project was that most of the records had kept the contact information of the complainants. This allowed us to conduct follow-up interviews for each dispute. Of course, some contact information was outdated, as migrant workers may have migrated to other locales. Nonetheless, we were successful in reconstructing the entire life cycle of 225 disputes. We will refer to these 225 disputes as the “interview sample” (see Tables 3-5). Although the data are incomplete, they offer a rare peek into the systematic evolution of labor disputes, and how (in)effective the LIB was in handling them.

Table 3. Subsequent Trajectories and Outcomes of Labor Disputes

Note: Data source—the interview sample.

Tracing how these disputes are handled offers a systematic picture of how they evolve. Aided by two research assistants, we used the complainants' phone numbers registered in the daily log to conduct interviews to find out how these cases had fared, and how they had ended by the year 2012. This interval of several years between the occurrence of the disputes and the interview ensured that most disputes had been settled by the time of the interview. After briefly introducing ourselves and the goal of the research, we asked the complainants to relate their stories. In 225 cases, the respondents kindly accepted our interviews. There were also cases in which the complainants forgot the details. Regardless, we retrospectively recorded the trajectories of these disputes' life courses by asking the following questions:¹ Was a second government agency involved? Did the complainant use street protest? Did the case ultimately end up in arbitration or in court? What was the outcome? Was the grievance addressed, and in what way? We then ran statistical analyses on the data collected.

In our multivariate analysis, we run regression models on two dependent variables. One is a dichotomous variable in the life cycle of the dispute, whether a case has ever entered the legal channel—defined as either appearing in court and a government-run arbitration committee. Of the 225 petitions in the interview sample, 25 (11 court cases and 14 arbitration cases), or 11.11 percent are such petitions (see Table 3). In the other dependent variable, we measure the outcome of the dispute—whether the petitioner ended up obtaining monetary compensations, recorded as “1” if yes, and “0” if no. For both dependent variables, logistic regression models are appropriate.

Among the key independent variables, we include one dichotomous measure to indicate whether the dispute had been processed by the LIB. This measure is important in predicting both of the above dependent variables: whether the LIB's process reduced the likelihood of a dispute moving upward toward the formal legal channels; and whether such processing by the LIB helped the petitioners satisfy their demands. In the models predicting the dispute outcome, another key independent variable is whether it entered the legal system. We are curious if entering the legal system had a positive or negative effect on achieving the petitioner' goal.

We investigate these relationships in the context of controlling for other factors that may obscure them. We use the following measures as the control variables in our models.

• • Petitioner' gender, coding male as 1, female as 0

• • Type of petition, coding those that involved wages and employment deposit as 1, the others as 0

• • Whether the petitioner was a local resident (i.e., from within the county); coding yes as 1, no as 0

• • Whether the petitioner was an urban resident or a peasant, coding urban as 1, peasant as 0

• • Size of the company being targeted by the petition, coding large company as 1, the others as 0

• • Number of petitioners involved

• • Whether the dispute involved any collective actions, coding yes as 1, no as 0

• • The amount of the money claimed by the petitioner.

Typically, a quantitative research paper bases its argument on its regression models. However, our article is different. Although we present statistical tables and multivariate regression models, it is important to emphasize the mixed-methods nature of our research: it provides both quantitative and qualitative angles to investigate the process of a set of Chinese labor disputes. Not only do we examine the dispute process from “outside in” by interviewing disputants on their personal experience and the process as they perceive but also we investigate the same process from “inside out,” via our in-depth ethnographic fieldwork inside the government agency. Thus, our thesis about flexibility and authority rests on both quantitative data and ethnographic narratives.

The former, with descriptive statistics and regression models, presents a collective profile of the disputes that were initiated in and processed by the LIB, and assesses the results of these disputes. The message in this part of the analysis suggests a certain level of effectiveness. Unlike the situation depicted in the classic dispute pyramid, the majority of the disputes entering the LIB in County Z were processed, and many of them ended with an outcome favorable to the petitioner. The latter, with interview data with the government officials and the participatory observations of the government action, explains why the disputes turned out as they did. In sum, although we use regression models as an important step to build our thesis, our thesis does not rest solely on the quantitative findings; it also incorporates the findings our ethnography of the LIB.

4. Trajectories and Outcomes

Our research design allows us to track the lifecycle of the disputes beginning with their inception, their registration at the County Z LIB in 2004 and 2005. Using the daily log sample, Table 1 presents the profiles of the disputes in terms of their characteristics such as the issues of complaints, the numbers of individuals involved, the monetary value involved and the types of employers.

4.1 Small but Significant Disputes

Let us first look at the nature of these disputes. Other than those lacking details, all were labor related: 784 of 810 records showed that they were either related to wages, job deposits,² retirement, welfare, work injury, or layoff severance. This is apparently because LIB officials would not register disputes if the complaint was unrelated to employment. More than three quarters (77 percent) of the disputes concerned wage or job deposit issues. As will be shown, issues surrounding wage arrears and the unfair withholding of job deposits were more straightforward than the others.

These disputes involved small amounts of money. Twenty-two percent of the 810 disputes involved sums equal to 1000 RMB (approximately $120 USD as of 2005 rates) or less. However, 1000 RMB was significant, as it was equal to roughly 1–2 months of wages, or the deposit for migrant workers in the construction business. Twenty-three percent was between 1000 and 10,000 RMB. This amount was equivalent to the wages of a migrant worker covering a period of several months to 1 year. Eight percent had a stake greater than 100,001 RMB (or $12,048), which usually involved multiple individuals. Of the 810 cases in the daily log sample, 61 percent involved one individual complainant. The rest were registered on behalf of a collective: 18 percent had been filed by two individuals, 17 percent involved between 10 and 100 individuals, and 26 cases (3 percent) had been filed on behalf of 100 or more workers. Those with higher amounts in dispute usually involved multiple complainants. As the daily log only recorded those relatively complicated disputes, leaving many straightforward disputes unrecorded, the LIB handled primarily small disputes.

All disputes had been filed by employees. Approximately three quarters of the complainants were migrant workers, the rest being local residents of the city. Among the targets of the complaints, 9 percent were ge-ti-hu (individual households). Legally speaking, the LIB had no power against this group because they were natural persons rather than employers: the LIB can only inspect employers. Another 40 percent were mid-size private businesses. State-owned businesses only accounted for 6 percent of the sample.

4.2 The Role of the LIB

Again based on the daily log sample, Table 2 reports the actions taken by the LIB once a disputant came to the office for petition. In the table, we only report the initial actions, such as making a phone call, visiting an employer, transferring the case to another agency and so on. The practice of keeping the daily logs provides more detailed information on these initial responses than does the processing in the subsequent stage. Therefore, for the further actions in the later stages of the dispute, see our follow-up interview reports.

Upon reaching the LIB, disputes were processed immediately by the official on duty, often on the same day. The forms for 80 percent of the 810 cases in our daily log sample had been marked with at least one specific means of processing, including making phone calls, paying visits, interviewing the target, transferring the case, or suggesting other channels. Fifty-eight percent of the targeted employers had received a phone call inquiry, and 10 percent had been asked to come to the LIB to explain the situation. In 49 cases, an official had visited the company to investigate. After the initial processing, only 19.50 percent of disputes are still marked as pending or unprocessed; 46.79 percent are marked complete, and 25.56 percent of them are suggested for a second step.

From the standpoint of the disputants, 46.79 percent indicate a high level of responsiveness of the government to their claims. If we refer back to the dispute pyramid, we know that due to the lack of recourse, the majority of disputes end up being lumped or otherwise set aside. A tiny proportion of them are processed by the public authorities including the court. The percentage varies, of course, but 46.79 percent are by no means insignificant. Indeed, they are remarkably high. This is especially so because most legal systems tend to ignore these small claims.

4.3 The Dispute Pyramid

To follow the entire life cycle of the disputes, we successfully interviewed 225 petitioners 5 years after the initial petition. Table 3 uses this interview sample to report actions taken after the initial stage.

Our findings confirm the established pattern of the dispute pyramid—only a tiny proportion of disputes reached the “apex” of the formal legal process (Felstiner et al. 1980-1981; Michelson 2007). Contrary to traditional wisdom, however, our data suggest that many other disputants did not simply dump their cases, nor were their grievances silenced without resolution. Instead, the government agency intervened with various third-party efforts. Ultimately, many of the outcomes favored the initial disputants.

The data from 225 follow-up interviews help reveal what happened after the initial processing by the LIB: 147 disputes, or 65 percent, ended without a second agency. This result is fairly consistent with the data from the daily log sample, which reports that 46.79 percent of the disputes were deemed “complete” by the LIB after the initial processing. The higher percentage here may indicate some of these cases reached the conclusion because the petitioner stopped perusing the recourse midway. Apart from the five cases lacking details, the rest all went to a second government agency: the arrears clearance office (4), the letter and petition office (3), media organizations (1), the arbitration committee (14), the court (11), another branch of the labor bureau (9), or an unspecified agency (31). As the formal legal channel includes the arbitration committee and the court, of the 225 disputes, 25, or 12 percent, reached them. The complainants in 15 of 225 disputes reported that they retained a lawyer and 8 of 215 disputes reported protesting.

Only a minority of the disputes resorted to the formal legal channel (12 percent), yet most of “the other cases” obtained third-party interventions in one way or another. These interventions came from the LIB or other government agencies. As reported in the last panel of Table 2, such interventions produced positive relief for about half of the disputes (46 percent).

A clear pattern emerges: although most disputes did not achieve the status of a legal case, they did not just disappear during the dispute resolution process. They were processed with the assistance of government agencies, whose simple interventions tipped the balance in favor of the employee, thus enabling the employee to obtain a certain amount of compensation from the employer.

4.4 Which Types of Disputes Were Transformed into Litigation?

If the majority of disputes were processed by a noncourt agency in their initial stages, how did such processing affect their chances of becoming formal legal cases? In other words, did the noncourt agency function as a springboard to the court, or as a dispute-settling occasion itself which ended the dispute? The above crossable analysis has partially answered these questions. Many disputes ended in the LIB or other government agencies, without entering the formal channels.

A more complete answer, however, requires a multivariate context. In order to ascertain the effect of the initial processing, we control for the other characteristics of the disputes, including the dispute type, the claim size, the petitioner's socioeconomic status, the type of targeted party, and the number of petitioners involved. These characteristics potentially have their own effects on whether a dispute will reach the legal channels. We run two regression models in Table 4 to investigate. Model 1 is a fuller model in terms of sample size with 225 cases; we also introduce a control variable indicating the size of the petition claim, which results in some missing cases. In Model 2, the sample becomes 153. In both models, however, the message is the same: if the case had been processed by the LIB first, it may have ended there with a resolution: it would have been less likely to have escalated into become a court case.

Table 4. Coefficients of Logistic Regression Predicting Likelihood of Dispute Entering Legal Channels

^**p < 0.05; ^***p < 0.01.

Note: Data source—the interview sample.

Other factors being equal, there are two significant effects, shown in the statistical results. The first is the dispute type—specifically, whether the dispute involves issues of wages or job deposits. Take Model 2 as an example. This factor has a strong negative effect. That is, wage or deposit-related disputes were less likely to develop into litigation, with a negative coefficient, −1.767. The corresponding odds ratio, as shown in the table, is 0.171(i.e., exp[−1.767]), indicating that the likelihood of this type of dispute is a fraction of that of the other types. The likelihood for a wage or deposit-related dispute to go to channels, measured in terms of odds, decreased by 82.9 percent, in comparison with the other types of disputes.

The second was the effect of the initial processing of the LIB. Once they were processed by the agency, the likelihood of disputes later being litigated decreased by a negative coefficient −1.913, with the odds ratio being 0.148. That means processing by the LIB will decrease the odds of entering the legal channels by 85.2 percent).

4.5 Which Elements Facilitated Positive Outcomes?

As shown in Table 3, most disputes received by the LIB were processed, and about half of the 225 disputes in our interview sample obtained a positive outcome. Why can “winning reward” be an indicator of effectiveness of the dispute resolution? The reason lies in the lopsided power relationship between migrant laborers and the business. The migrant workers were routinely mistreated, with the wage arrears being the most severe problem. Previous research has amply documented the plight of the disadvantaged (Lee 2012). In particular, He et al. (2013) and (Zhuang and Chen 2015) detail the length for the workers to get their wages back.

Our further multivariate regression analysis was to ascertain which elements were crucial to achieving the above outcomes.

In the logistic regression models in Table 5, we employ a set of independent variables to estimate the likelihood of the disputes achieving monetary compensation, as a positive outcome for the initial petitioner. With other factors controlled, three findings stand out. The first is the effect of government assistance. The LIB's processing increased the likelihood of redress by a factor of between 4.368 (see the odds ratio in Model 3) and 5.241 (see the odds ratio in Model 2). This finding is consistent with the previous discussion.

Table 5. Coefficients of Logistic Regression Predicting Monetary Reward in Dispute Resolution

^**p < 0.05; ^***p < 0.01.

Note: Data source—the interview sample.

Second, “target being large company” is a negative significant factor in predicting the likelihood of monetary awards, by an odds ratio of between 0.385 (in Model 2) and 0.420 (in Model 3). This suggests that large companies might be more capable of thwarting workers' claims. The claims against them might therefore have less positive outcomes.

Third, entering the legal channels in fact would decrease the likelihood of being compensated. As shown in Table 5, in the first two models (Models 1 and 2) the effect is negative and significant, with the odd ratios between 0.305 and 0.229. This indicates that if a dispute could not be resolved by the LIB and become a legal case, its prospects of a favorable resolution were worse than the other cases. This is perhaps because disputes that failed to be resolved by the LIB were the more difficult ones to begin with. Indeed, after controlling for the dispute type, the negative effect of going through legal channels is less pronounced (odds ratio being 0.614 in Model 3).

Finally, the data also show that although the involvement of a lawyer had a positive effect, it was not significant. This contradicts previous studies which have found that lawyers are important for case outcomes (He and Su 2013). A plausible explanation is that as the LIB intervention was effective, it restricted lawyers' room. Of course, one should not read too much into this, as less than 7 percent of the petitioners in the interview sample had legal representation (see Table 3).

These patterns challenge the conventional wisdom of dispute resolution in two ways. First, those disputes that did not reach the “apex” of the pyramid—the formal legal channel—do not fail per se; they do obtain a form of redress. Second, the government agency plays a crucial role in achieving this. This factor is so prominent that the effectiveness of the formal means—lawyers and courts—is overshadowed: it seems that they do not advance the chances of gaining relief.

5. Flexibility and Authority

The above analysis finds that the government agency plays a crucial role in resolving disputes. Compared with the labor arbitration and litigation, the agency is effective at resolving a large portion of disputes that would otherwise be denied access to be heard, let alone a positive resolution. For many migrant workers, going to court is out of the question. The daunting image of a rigid formal legal channel and the costs involved deter them. As suggested by He et al. (2013), these workers are “beneath the law.” Furthermore, the outcomes of the disputes are not insignificant. More than 60 percent of complaints are complete after the initial processing. Only 7 percent were not processed. After the LIB's intervention, only 5 percent went to court and 4 percent involved social protest. More than half of the interview sample received a positive outcome. Usually there are two reasons for a complete complaint: one is a satisfactory solution seen by the complainant (Felstiner et al. 1980-1981). The other would be a rather neutral outcome—the complainant may not be satisfied, but sees no way forward. They may stop pursuing further because they are dissuaded by LIB officials, or due to other concerns. Nonetheless, all the cases are “complete” when there is no longer impetus to escalate and linger. The LIB has at least pacified the complainants. It is also far more efficient than arbitration or litigation as most disputes reaching the LIB are processed within hours.

This process is far from perfect, of course. As will be detailed, LIB officials often worked with enterprise bosses to keep the level of workers' compensation lower than the standard of the labor contract law. Nonetheless, a less satisfactory but immediate outcome, when acceptable (even reluctantly) to the complainant, is better than an unrealistic promise from the law and the court. This is also why three quarters of migrants consider government mediation to be the best way to resolve disputes (Zhuang and Chen 2015: 398).

How does the LIB achieve this? Although the LIB is vested to inspect labor relations, its real power is limited. For example, although the Statutes of Labor Inspections stipulate that the LIB can request an employer to pay wage balances or even fines, the LIB cannot actually enforce that request. It has to rely on the courts to enforce its decisions (SPC 1998). According to several LIB officials interviewed, this arrangement renders enforcement unrealistic because most wage claimants cannot afford the time it takes to initiate the court proceedings. The interviewed officials also admitted that they are, in a way, toothless tigers, especially in cases involving enterprises familiar with the relevant laws. All these beg the question: why then are many disputes resolved by the LIB?

Our fieldwork investigations, including interviews with officials and participation observation, suggest that the effectiveness of government involvement lies in the nature of the LIB as an informal dispute resolution institution: flexibility and authority. As an informal dispute resolution institution, it does not enjoy the kind of power that can directly determine the outcome of a dispute; it is informal in nature. However, this frees it from the rigidity of the formal institution, and thus enables it to address disputes effectively, especially for certain types of disputes. It also allows the LIB to transmit its power as a regulator to authority in dispute resolution.

First, the flexible nature of the LIB attracts specific genres of dispute. As shown in the daily log, most registered disputes concern small amounts; they are straightforward and predominantly involve migrant workers. The issue is simple. In many instances, the employers did not disagree with workers' requests; they just did not pay as promised when a claim was initially made. For example, a young man working for a small restaurant complained that his employer—his brother-in-law—had failed to fulfill a promise of payment before a certain date. The employer, upon receiving an inquiry from the LIB official, explained that he had never refused to pay, but needed one more week because he had recently been short on cash. The case was thus resolved through a phone call. These so-called disputes can be thought of as diverging opinions of the employer and the employee. In the LIB daily logs, few cases involved complex legal disputes. We did not see, for example, any disputes related to the medical expenses or retirement benefits of large groups of employees in state-owned enterprises. Apparently, such complicated disputes might have been taken directly to the labor arbitration committees, the courts, or local political leaders.

Second, in part due to their limited power, LIB officials employ flexible means and techniques. Precisely because the LIB is a toothless tiger, its officials are careful in communicating with both the employers and employees. According to our observations, the officials were usually friendly toward complainants. Patiently listening to the laborers ensured that they understood the causes of complaints. After listening, they would then suggest a solution. In cases they believed to be credible, they immediately offered to call the employer. In their interaction with the employers, the most commonly used technique was to elaborate upon the employees' predicament, and ask the employers to place themselves in the employees' shoes. “Persuading with emotion and reason” (以情动人，以理服人) was crucial in communicating with the employers, according to an official interviewed. The following was a telephone conversation between an LIB official and an employer:

Boss Tang, … your laborers have come to my office. Can you reply to them by the 18th? One should just pay the wages due. It is already the eve of the Spring Festival. They all have to return to their hometowns. You do not want to always be bothered by this, right? As they have come to us, we are in the position to convey the message. OK? (Wang 2009: 86–87)

In this conversation, the official insisted that “one should just pay the wages due” and noted that the Spring Festival was approaching, a time when most migrant workers needed money to visit their families. He also implied that once the issue was resolved, the boss would be free to do other things. Of course, he also showed his authority as an official—“we are in a position to convey the message.” For many disputes, this technique was effective. With a nudge from the LIB officials, most minor labor disputes, such as this, were resolved.

Third, the LIB's flexibility also means that their dispute resolution is pragmatic. For many labor complainants, this is the most attractive aspect of the LIB. Although the court may offer a full set of compensation to the plaintiff, the court decision may not be enforced, and the legal process itself takes much longer. Many migrant workers, eager to get paid, did not have the time, expertise, or money to go through the formal process. The ultimate goal of the LIB officials was to get the wage arrears back for the migrant workers. One official cast serious doubt on the formal legal channel:

Following the legislated procedure cannot solve these problems. Tens of unpaid migrant workers are staying at office. What do you do? Who pays their lunch boxes today? These are questions that need imminent solutions. It could take years if we follow the normal procedure, notifying the companies through written documents, asking them to be consulted by appointment, and making changes according to our suggestions!

In a survey of the labor authorities' flexible means to defuse labor disputes in Guangdong province, Zhuang and Chen (2015: 398) echo:

Migrant workers have a strong preference for mediation … as it can avoid prolonged litigation or arbitration and can bring about quick compensation, although that can be a smaller sum than they expected. … About 87 percent of the respondents expected a more active role from the government in mediating labor disputes, and approximately three-quarters believed that mediation by the government was the best way to resolve labor disputes.

Several legal requirements are compromised. The LIB officials try to find a balance between the requests of the migrant workers and their employers. This is consistent with Zhuang and Chen (2015), who found that the labor bureaus in the Pearl River Delta usually depressed workers' claims for compensation, or would limit the amount of compensation required by the law, so as not to drive away investors. As a result, the money paid was sometimes less than the agreed amount, or was transferred later than the agreed time (He et al. 2013). Our interviews with the LIB officials and migrants (He et al. 2013) demonstrate that for claims involving disruptive actions such as protests or threats to suicide, “migrant workers only received 50 to 70 percent of the requested amount. Many migrants were certainly unhappy about the result… Partial payment is better than nothing, and they do need to take the money home…” For government officials, however, as long as it was acceptable to the other parties, it was a good deal. In their own words, “to get both parties to accept is a triumph (抹平就是水平).”

This also explains why so few lawyers were involved in the process. Getting a lawyer entailed money and time, but could not guarantee that the outcome would be enforced (Michelson 2006). What the LIB offers is raw justice, but timely raw justice is better than late justice, let alone an unenforceable court judgment obtained after a lengthy and expensive legal battle.

Finally and most important, the LIB is not bound by rigid procedural rules. Unlike the judges, the LIB officials rarely worried about evidence. In their own words, migrant workers were always trustworthy. If they had decided to visit them or make a phone call for help, they must have suffered some grievance. The LIB officials thus were not bound by the rigidity of evidence rules, so they could act efficiently and effectively to resolve disputes. Furthermore, the LIB has taken up many disputes outside of its jurisdiction. Labor law has a clear definition of what constitutes the employment relationship. For many migrants who did not even have a labor contract with their employer, common in the construction business (Pun et al. 2010), their claims, legally speaking, should not have been regarded as labor cases. In situations when the employers had disappeared, the migrant workers could only go after the owner of the project. If the LIB had been a formal court and had strictly followed the law, there would have been no legal basis for accepting the disputes. Similarly, if the claims had been against the ge-ti-hu, over which the LIB has no jurisdiction, the legally appropriate way would have been to guide the complainants to other channels. According to both our observations and the data, however, in many situations the LIB officials did not even bother to ask the nature of the complained business before they made the call. Sometimes, regardless of the absence of a legal relationship, they would request the developer, the contractor, the project managers, the group leaders and the representatives of the migrant workers to convene at the LIB office and work out a solution. An LIB official said:

Call all these related parties to sit together! How much do you owe? This migrant worker brother said 500 yuan was owed. The boss responded, “how can it be that much? Only 200 yuan.” One said 500 and the other said 200. We will mediate them in one day!

In addition to flexibility, the other characteristic critical to the success of the LIB is power. A toothless tiger is still a tiger, with a scary face and powerful claws, as it is also the regulator of the business. In the court and arbitration committee, the judges and the arbitrators have the formal power to determine the outcome of cases, but they are not the regulators of the employers. This line is clear and hard to transgress. The LIB, however, has become both regulator and adjudicator in handling disputes. Although the LIB's formal power to issue administrative penalties is limited, it has the power to regulate employers' behavior. For example, the LIB is empowered to inspect the safety of the working environment. It can also inspect whether or not labor contracts comply with legal requirements. For instance, in 2002, a migrant worker from a construction company committed suicide by jumping off a rooftop. Lacking the administrative power to impose a penalty for this incident, the LIB imposed a fine of 3000 RMB on the company for noncompliant behavior in labor contracts (Wang 2009: 94).

This form of the authority of the LIB transformed from its power as a regulator makes a difference. Another LIB official said:

Those who are legally obedient and follow the rules usually pay as requested. For those who are diehard and resistant, they just ignore you. What can you do? … We coordinate, mediate. We still have some authority. Oh! Why do you owe others' wages? One would feel scared to hear this, right?

Although the statement suggests that the LIB enjoys some level of authority, it states that “those who are diehard and resistant, they just ignore you.” Indeed, the authority vested in the LIB is neither thorough nor absolute. Local governments are also beholden to employers because they need to develop the local economy. They are often at the mercy of big bosses and cannot do much more than inspect construction sites as a show. Nonetheless, government officials were still able to wield some level of authority and garner some respect from employers. To avoid future hassles, employers sought to maintain good relationships with the LIB. At the very least, they did not want to offend it (Li 2013). One employer said:

The government can always find ways to deal with us. No enterprises can follow the law 100 percent. They may find safety problems in your construction site, they may find inappropriate employment relationships in your company, they may find that you employed underage workers, they may find you are evading taxes…This agency may not have the specific power to penalize us; but they can always coordinate with other agencies. At the end of the day, as long as their requests are within the range, we cooperate.

When the employers were not available, it was also common for the LIB to seek the help of the developers, who often showed their respect to the LIB.

In an incident in which a construction contractor did not pay the wage arrears of migrant workers for 1.7 million yuan, and more than 50 migrant workers had protested at the government compound on the eve of the Spring Festival, the LIB and the Construction Bureau urged the developer to help. This occurred even though the developer did not have any legal relationship with the migrant workers. The developer eventually agreed to pay 800,000 yuan. The boss explained:

In this business, the developer commonly gives face to the government. The government invites us to invest here. We cannot leave them alone when they (the government) are in trouble. After all, they are very supportive of us. Whenever we need help, they are quite enthusiastic and helpful. To pay this 800,000 yuan is completely baseless, but we have to!

The role of the LIB as both regulator and adjudicator thus places it in a powerful but subtle position and enables it to make suggestions. Its authority as a regulator has been transmitted to the area of dispute resolution. The pressure of its “inquiries” or “suggestions” is constant and powerful. It is a variant of “bargaining in the shadow of the law” (Mnookin and Kornhauser 1979). What is distinctive about the LIB is that the shadow it creates for itself is not so much a legal shadow, but an administrative one. Formally the LIB is a regulator and informally it is a dispute resolver. Even though it does not have legal teeth, it does have administrative teeth, and those teeth are arguably more powerful, or at least more petty and available. Indeed, many employers are unclear about the exact jurisdiction of the LIB. There is, to use economic terminology, an information asymmetry between LIB officials and employers: although the LIB officials know the boundaries of their power, employers may not. In a country and culture in which officials and governments are generally respected, a “suggestion” made by an official is more powerful than might be expected elsewhere. For disputes that are clear-cut, it makes no sense for employers to resist. Most respect the officials and thus are willing to respond quickly.

This authority also works well on the employees, and especially the migrant workers. Once a portion of the requested amount is paid, the officials usually persuade the migrant workers to be satisfied with what they have been given. The following words of officials are telling (He et al. 2013: 725):

“Please take this portion now, and come back for the rest later.”

“You know, in many places the migrant workers can only get the traveling expenses, you are much better already.”

This is why most complainants, though with only a portion of the requested amount paid, stop pursuing further. LIB officials' education and persuasion make a difference. Obviously without possessing a relatively higher level of authority over the claimants, the LIB officials could not have achieved this.

6. Beyond the LIB

Flexibility and authority are critical for the LIB to be effective. Is this finding limited to the case study, or can it also be applied to other informal justice forums? A useful theoretical construct has to survive beyond its original context. With this in mind, this section looks to instances of effective mediation or informal justice beyond the LIB in China. The aim here is to highlight a common dynamic: to identify analogies that can make sense of details that otherwise would seem accidental. Although remaining attentive to variation stemming from a host of factors, we can begin to assess the generality of the findings from China's LIB, and avoid wrapping a conclusion around a single case.

The first example, the fate of China's People's Mediation Committee (the PMC) before and after the reforms, illustrates the crucial role of authority in the effectiveness of informal justice forums. In Mao's China, the PMC was forceful. This was because the mediation system, including the PMC, was politicized (Lubman 1967). The mediators were both activists and cadres, collaborating with the police. Following state policy and political standards, they helped repress the enemies of the people, and “educate” or “persuade” bad elements to reform (Lubman 1967: 1348). They combined the power to persuade with the power to sanction. They might have lacked explicit coercive power, but they could trigger more authoritative actions leading to criticisms and greater sanctions than could the judges. “[m]ediators are integral parts of the mechanism of Party rule” (Lubman 1967: 1349). In a word, the authority of the PMC before the reforms was formidable. In the reform period, however, the PMC's authority declined. Although 7.4 million cases were handled by the PMC in 1990, this dropped to 5 million by 2000 and fell to just 3.1 million cases in 2002—roughly 40 percent of the 1990 figure. One reason for the decline, according to Halegua (2005), is the lack of authority for the institution of the PMC and the mediators (see also Di and Wu 2009: 242, Wu 2014: 121–122). The state has put more emphasis on rule of law and court professionalism, and has thus sidelined the PMC. Disputants have “less respect for the authority and abilities of an old-aged mediator with less education and perhaps poorer understanding of law than they themselves have” (Halegua 2005: 719–720).

If the shrunken effectiveness of the PMC reveals the importance of authority in an informal justice forum, the fate of judicial mediation in China suggests that its inflexibility eclipses its effectiveness. As the Grand Mediation policy was put into place in the mid-2000s, judicial mediation has been required for most court cases. The revived judicial mediation, however, has not been as effective as intended. Although the mediation rate has become a crucial criterion for assessing the performance of the judges and the courts, many judges, especially those with formal legal training, resist this requirement. Although official statistics claimed that the percentage of civil cases resolved by mediation has increased (China Law Yearbooks n.d.), Li et al. (2018: 3) finds that “mediation rate increases at first were largely the result of low-level court involvement in mediation practices outside of the courtroom,” and “the mediation rate was produced strategically by seeking the highest increase possible under the discretion of the definition of mediation in the incentive structure with the lowest cost and risk to the court.” A recent study (Ng and He 2014) documented that of 25 cases they observed during their 2 weeks of fieldwork, only one was successfully mediated, and that was accidental. According to it (Ng and He 2014), one reason for this ineffective policy is that “When the rather rigid format of adjudication is carried over to in-trial mediation, it curtails the flexible, nonlegalistic approach that mediation is meant to promote.”

This contrasts with the effectiveness of judicial mediation in Mao's China. In that era, the judges had not been bound by the format of procedural rules, and they investigated neighborhoods to understand the causes of disputes. As illustrated by Huang (2010), the hearings were held in the neighborhood or government buildings, with the participation of government officials. The judge would also provide government-funded remedies, in addition to the legal remedies provided by the litigation parties themselves.

The crucial role of flexibility and authority has also been shown in the mediation of various businesses in contemporary China. Although scant systematic research is available, preliminary studies suggest the importance of flexibility and authority also here. For example, Huang (2015: 187) suggests that securities dispute mediation in China has been welcomed by the market as a useful mechanism for resolving disputes. Huang (2015: 187) argues that this is because the scope of acceptance is wide, the service is free and “mediation can be conducted in flexible ways, including but not limited to, adopting writing, network, telephone and other off-site means; separately or concurrently meeting the parties or their attorneys.” If the above focuses on flexibility, the last point made by Huang is about authority: securities dispute mediation establishes a “collaborative relationship with courts, arbitral institutions, notarial institutions and other institutions,” so as to “improve the utility and enforceability of mediation agreements.” Huang (2015: 179) continues, “mediation by the industry association is generally regarded by financial consumers as more authoritative than general mediation.” This is echoed by the successful case of medical mediation, in which patients and hospitals are encouraged to apply for judicial confirmation after they reach a settlement in medical mediation (Ding 2015: 168). On the contrary, in explaining the failure of the mediation of environmental disputes, Zhao 2015a, 2015b: 242) argues that a lack of legal force is the primary culprit.

Flexibility and authority are also the key to the effectiveness for informal justice forums in societies with advanced legal systems. For example, in the United States, consumer protection bureaus are associated with the state attorney general's office; New York initiated this trend in 1957, and 44 states had established such bureaus by 1973 (Johnson et al. 1977). The law enforcement activities of the bureaus are related only tenuously to the dispute-settling process (Steele 1975: 1119). They handle complaints through a combination of persuasion, education, mediation, and litigation. Of these, studies have found mediation to be the most effective (cf. Johnson et al. 1977, note 87).

The data demonstrate that the government bureau has a high level of effectiveness in diverting claims from the courts although producing significant resolutions. In Steele's study, 35 percent of complaints were settled, 20 percent were found to be outside the bureau's jurisdiction, 21 percent were invalid claims, 10 percent went unresolved because of inability to contact the seller, 10 percent were dropped because of no further response from the complainant, and only 4.5 percent went on to litigation (Steele 1975). This finding has been supported by several other studies (Johnson et al. 1977, note 98).

What is the secret to this effectiveness, when other informal forums had produced mixed results, and many had run into frustrating obstacles? The existing literature on ADR, particularly that written as advocacy for nonstate solutions and informality, emphasizes the magic of mediation—its nonadversarial features aiming at pragmatic solutions with flexible procedures. Indeed, in dealing with consumer complaints, the bureaus' priority is obtaining compensation for consumers, not reinforcing general rules. The aim “is to get money back as promptly as possible for the citizen who has been cheated or defrauded” (Johnson et al. 1977: 67). In his study, Steele (1975) also resorts to flexibility in explaining mediation's success.

Do most ADRs not share these same features—nonadversarial character, modest goals, and flexible procedures? Why are the U.S. consumer protection bureaus and Chinese LIBs more effective than most others? The explanation lies in the agencies' authority over the disputing parties—in this case, over the companies and retailers. As Johnson et al. put it, consumer bureaus are able to back up their “suggestions” for resolutions with potent sanctions such as injunctions and other orders (1977: 67). This point echoes the “informal power” possessed by the Chinese LIB over businesses under its jurisdiction: as described, sanctions range from holding up clearance for sanitary inspections to confiscating operating licenses. The difference between the two agencies is that the Chinese LIB does not have the power to adjudicate cases directly. This power is vested in its sister branch—the arbitration committee.

A moment of reflection on other ADR practices of may reinforce this point. One of the most ingenious ADR experiments in the United States was the Neighborhood Justice Center, which reached its height in the late 1970s and early 1980s (Cook et al. 1980; Harrington 1982, 1985; Hofrichter 1987; Tomasic and Feeley 1982). Its advocates insisted on its community character as a solution outside the state. In reality, most of the centers were the product of state legislature, and funded by the government. Initial assessments were mostly positive, but the “success” of the Neighborhood Justice Center, mainly promoted by the scholars who themselves were its advocates, is disputed by critics for lacking a rigorous empirical base (Harrington 1985). More important, by the end of the 1980s, most Neighborhood Justice Centers had run out of funds and had closed, joining the legion of failures in the history of “informal” or “popular” justice.

The termination of funding may have been the logical extension of failure, rather than the cause. The heart of the problem is a lack of authority (Abel 1982b). Unlike traditional townships and villages, the “neighborhood” in contemporary American cities does not comprise an organic unit. It thus lacks the cultural infrastructure necessary to generate community authority on behalf of the “neighborhood.”

7. Conclusions and Implications

This study focuses on not just arbitration or litigation—the disputes that reach the apex of the dispute pyramid—but also the fruit and flowers if the dispute transformation process is regarded as a tree with many branches (Albiston et al. 2014). It represents an effort to evaluate the alternative trajectories of dispute resolution. We have presented a systematic study in which an executive branch of the local government functions as an informal dispute resolution forum by way of its flexibility and authority over the potential disputants. This analysis shows that the LIB, a state-backed channel, effectively resolves disputes. As our findings suggest, it helps resolve a large portion of disputes which would have otherwise been denied access to justice. Although one may argue that the disputes are straightforward, in terms of size they are significant to the disputants. One may also argue that disputes initially routed directly to the formal legal system may be more complex. Nevertheless, the government branch provides an alternative to the formal legal channel in dispute resolution, and it has been doing this well.

We argue that flexibility and authority are the essence of this kind of administrative mediation. It is flexible because it is outside the formal court system. It is also flexible because the executive powers are used “tactically,” as a force to compel disputants to comply with its decisions. Its authority stems from its regulatory position in the government.

Flexibility allows government-sponsored administrative agencies to attract specific types of cases, which are usually screened out by the procedural rigidity of the formal channel. It also allows the handling of disputes in less confrontational ways and with more room for compromise. The authority of the administrative agencies also demands respect for their suggestions, and disputants are thus less resistant to the final resolutions.

The combination of these two factors provides a unique perspective on the formal legal channel and its relationship with informal justice. The formal legal system in the United States, characterized as adversarial legalism by Kagan (2003), enjoys authority but lacks flexibility. It is effective in many ways. It is also important in resolving specific types of disputes, especially those legalistic, complicated, important, and uncertain. It involves an army of professionals and consumes copious time and money. Nevertheless, its formality and rigidity has prevented many disputes from reaching its dockets. That is why informal justice, and especially the ADR movements, has arisen.

Viewed from this perspective, the formal and informal justice channels do not replace each other: each handles different types of disputes, scopes, and disputants. The LIB is feasible partly because it is a halfway, informal option; more difficult cases may be left for the formal system. They are complementary, like different business models catering to the needs of different clients. Reducing the rigidity of formal justice would certainly increase its accessibility to the general public, but resolving disputes through informal channels may not necessarily be a bad idea. Our study thus suggests an alternative agenda to traditional scholarship on access to justice (Sandefur 2008): although it is important to reduce the rigidity and the cost of the formal channel, it is equally important to set up informal mechanisms with the authority to deliver justice. Indeed, as more cases vanish from the formal channel and are diverted into alternative forums (Galanter 2004), this task becomes more pressing.

However, the effectiveness of informal justice should not be overstated. Without formalities established to ensure procedural neutrality, the informal channels of justice are more susceptible to corruption and arbitrariness (Fiss 1984). The success of the LIB illustrates, in a roundabout way, the weakness of the law in China. The Chinese government's goal in establishing such channels is to better control society, and thus laborers' rights are not the top priority; no institutional arrangement is put in place to prevent similar violations from recurring in the future. The coerciveness of mediation in Communist China is a telling example of how cases can turn wild in informal justice. When an adjudicator is also the regulator, there can be many undesirable consequences, which is why administrative laws are established to prevent the abuse of power. Allowing the further combination of judicial power and administrative power is the opposite of specialization and professionalism in the development of the rule of law; it will impede legal development in the long run. In a way this weakness of the legal development continues into the Xi era: the regime now emphasizes rule of law and legal professionalism, but informal justice and the control of society are not less stressed (Ng and He 2017). The role of administrative power remains prominent and the sacrifice of labor rights persists. Moreover, as the informal channels of justice tend to be more suitable for small claims disputes, the efforts made to establish informal forums are not intended to replace the formal channels that are more suitable for complicated and difficult cases.