Justice from Above or Justice from Below?
Popular Strategies for Resolving Grievances in Rural China *
Ethan Michelson
Indiana University-Bloomington
Department of Sociology
Department of East Asian Languages and Cultures
[email protected]
March 6, 2006
word count (including notes and references): 13,309
DRAFT IN PROGRESS: DO NOT CITE OR CIRCULATE
WITHOUT AUTHOR’S PERMISSION
*
The survey data on which this paper is based were collected with the generous financial support of the Ford
Foundation (Beijing); for this I owe a special thanks to Phyllis Chang and Titi Liu. I would like to thank Feng
Shizheng, Guo Xinghua, Han Heng, Liu Jingming, Lu Yilong, Shen Weiwei, Wang Ping, and Wang Xiaobei for
administering the survey. Jing Tong at Indiana University assisted the data coding process. I would like to extend
my gratitude to Gardner Bovingdon, Lijun Chen, Neil Diamant, Sara Friedman, David James, James Lee, Sida Liu,
Kevin O’Brien, Phillip Parnell, Brian Powell, Benjamin Read, Frank Upham, Jianxun Wang, and Dali Yang for
comments on earlier drafts. Of course I remain solely responsible for all remaining defects and omissions.
Copyright © 2006
Ethan Michelson
ALL RIGHTS RESERVED
Justice from Above or Justice from Below?
Popular Strategies for Resolving Grievances in Rural China
ABSTRACT
It has become common to portray Chinese rural leaders as sources of trouble and as obstacles to
justice, and to portray Chinese villagers as having more trust in and receiving more satisfactory
redress from higher-level solutions including the legal system than from local solutions. In
contrast to this theory of “justice from above,” evidence presented in this paper from almost
3,000 households surveyed in 37 villages in six provinces in the year 2002 supports an
alternative theory of “justice from below.” According to the theory of justice from below, the
social costs associated with appealing to higher authorities, including the legal system, for help
with local disputes tend both to discourage the escalation of disputes and to produce relatively
disappointing experiences and outcomes when such routes are taken. The survey evidence
reveals that local solutions, often with the involvement of village leaders, are far more desirable
and effective than higher-level solutions.
China’s new political leadership has distinguished itself by focusing efforts—rhetorically,
symbolically, and substantively—on the plight of the increasingly contentious peasant. This is
the distinctive tone President Hu Jintao and Premier Wen Jiabao have set since taking over the
top two leadership positions during the First Session of the Tenth National People’s Congress in
March 2003. The widening gap between rural and urban China has become increasingly salient,
galvanizing public debate and popular rural protest. Resolving the “three rural issues” (san
nong)—referring to farming, peasants, and the countryside—is high on the new administration’s
list of priorities, as evidenced by the recent abolishment of all agricultural taxes nationwide as
part of a larger effort to stem further conflict and protest (Huang 2005; Xinhua News Agency
2005; Kahn 2005; RMRB 2005). The larger official effort to contain rising levels of social and
political tension in the Chinese countryside also includes strengthening the legal system, the
official complaints system, as well as other dispute processing institutions outside the formal
legal system (Luehrmann 2003; Minzner 2006; Cai 2004; China Daily 2004a, 2004b; RFA 2005).
How do competing dispute resolution strategies compare in terms of levels of utilization and the
likelihood of delivering satisfaction? Do villagers evaluate their disputing experiences and
outcomes more positively when less formal local solutions are pursued or when more formal
higher-level solutions are pursued? By answering these questions we can assess the magnitude of
the official challenge to resolve the popular complaints of China’s most aggrieved citizens and
the official effort to build a “harmonious society” (hexie shehui)—a leading slogan of the new
central administration. 1
In this paper “justice” refers to the lay notion of a desirable experience and/or a desirable
outcome—in short, of obtaining satisfactory help with a grievance. The advantage of such an
1
On the new official discourse of social harmony, see Marquand (2004), Kwan (2005), and Huang (2005).
1
admittedly fuzzy definition is that it compels the investigator to look in all the places to which—
and at all the methods by which—her research subjects report pursuing remedies to their
grievances, rather than merely in the places and at the methods the investigator deems important.
As we will see, a limitation with the existing research on disputing in rural China is its tendency
to treat “justice” as a euphemism for higher authorities including the courts. Once a more
representative range of third party choices is considered, thus removing a significant source of
selection bias, the theory of “justice from above” is turned on its head.
Justice from Above
In the prevailing image of social conflict in rural China, villagers are depicted seeking the help of
higher authorities, including the courts, to fight injustices suffered at the hands of tyrannical and
predatory local leaders. According to this “justice from above” account, an aggrieved peasant’s
best hope for receiving a fair shake is to go outside her village and appeal to higher authorities, a
now-routine course of action (O’Brien 1996:40-1; O’Brien and Li 2004:88-9; Johnson 2004; Cai
2004; RFA 2005). In this picture of rural China, village leaders try to keep a lid on conflict,
especially when they are the subject of the complaint, by obstructing justice through intimidation
tactics, forging alliances with higher-level leaders, and exacting revenge against successful
peasant complainants.
Thus, the argument runs, from the peasants’ perspective, resolution of a grievance
typically requires the support of outside authorities far removed from the influence of local cadre
networks. Aggrieved villagers reportedly believe their chances of getting justice improve the
further away from local leaders and the closer to the Center they seek help. The Center is
portrayed as relatively sympathetic and responsive (and downright terrified at the prospects of
peasant insurrection) while local officials are portrayed as relatively despotic and/or corrupt.
2
Peasants have therefore adopted strategies of “rightful resistance”: prominently displaying loyal
adherence to official state policy and law as the basis of both the substance of their complaints
and of their methods of appealing to higher authorities (O’Brien 1996; O’Brien and Li 2006).
(Also see Shue [2004] and Straughn [2005] on the strategic use of the language of the state as a
part of repertoires of contention in China and the German Democratic Republic respectively.)
This position is also what I term the “squeaky wheel” theory of dispute escalation—the
notion that the squeaky wheel gets the grease. This school of thought not only contends that
“rightful resistance is invariably noisy, public, and open,” but also asserts that the mobilization
of officially sanctioned tactics of contention spawned by China’s “growing rights consciousness”
and “rights talk” is both a viable means of redress and portends institutional change (O’Brien
1996:34,52-5). “Noisy, public defiance is one of the more perilous but effective ways the
Chinese leadership learns of official wrongdoing, and it can be particularly valuable when more
institutionalized channels of participation are lacking or clogged” (O’Brien and Li 2005:241).
The squeaky wheel theory, however, is not merely a scholarly proposition; it is also
represented in popular discourse by the oft-quoted saying, “a major ruckus leads to a major
resolution, a small ruckus leads to a small resolution, and no ruckus leads to no resolution” (da
nao da jiejue, xiao nao xiao jiejue, bu nao bu jiejue, Zhao 2003:23-4; also see O’Brien and Li
[2005:242] and Cai [2004:442]). The squeaky wheels go by various names, including “shrewd
and unyielding people” (diaomin) (O’Brien and Li 1995; Li and O’Brien 1996), the “professional
complainant” (shangfang zhuanyehu), and the “old-hand complainant” (shangfang laohu)
3
(O’Brien and Li 2005:246). 2 The theory of justice from above is distilled in the following
passage:
Contrary to understandings of Chinese politics that assume the central state weighs heavier on villagers
than its local representatives do, many villagers believe that the higher they go the more likely they are to
receive a satisfactory response. In fact, this may be one reason why a popular maxim in the countryside
holds: “the Centre is our benefactor (enren), the province is our relative, the county is a good person, the
township is an evil person, and the village is our enemy.” If villagers are able to reach higher levels, either
directly or through visiting journalists, relatives outside the village, passing inspection teams, or even
through visits by provincial or national leaders, their chance of gaining a hearing and redress may indeed
improve. (O’Brien and Li 1995:778; also see Li and O’Brien 1996:43)
Research on political trust in rural China, on the lower courts serving China’s rural
population, and on administrative litigation represent three additional pillars buttressing the
justice from above position. With respect to political trust, Li (2004) argues with the support of
survey data that villagers’ trust of central government authorities is far greater than their trust of
local government authorities. With respect to China’s lower courts, Zhu Suli, one of China’s
leading legal scholars, portrays a responsive, accommodating court system that operates flexibly,
bending the rules in order to provide satisfactory solutions to all involved parties (Su Li 2000;
see Upham [2005], Liu [2006], and Chen [2002]). 3 Finally, research on China’s 1989
2
These labels, however, should be treated as socially and culturally constructed categories at least as much as they
should be treated as objective reflections of reality. Indeed, these labels are reminiscent of American stereotypes
about the “frivolous plaintiff” and the “out-of-control litigant” (see Lofquist 2002 and Halton and McCann 2004).
3
It is worth pointing out the multiple personalities of Zhu Suli: In his earlier work (1996), which celebrates China’s
“indigenous legal resources,” he rejected formal state law.
4
Administrative Litigation Law supports the justice from above position by arguing that
administrative litigation—known popularly as min gao guan (“citizens suing officials” or “suing
the state”), a significant portion of which is initiated by rural residents (Tang 2005)—is likely to
produce favorable results for complainants, especially at higher levels of the court system (Pei
1997; Peerenboom 2001).
I will conclude this brief review of the justice from above position with three
qualifications. First, it is important to make clear that the primary focus of most of the research
reviewed on disputing in rural China is on strategies of collective action, with the question of
results—justice from above versus justice from below—being of only secondary interest. Second,
O’Brien and Li (2005:254) have recently tempered their claim of justice from above by stating,
on the basis of three interviews, that “trust in the Center is indeed waning. When rural activists
turn to higher levels, what awaits them more often than not is disappointment.” Third, and by
way of anticipating a central argument I make, O’Brien and Li (2005:253) acknowledge the
problem of selection bias introduced by their research design and therefore recognize the
severely circumscribed ability to generalize from their data.
Data and Methods
In late January and early February of 2002, members of the Department of Sociology at Renmin
University of China administered a survey of rural households in six provinces. They trained
local schoolteachers living and working in the survey sites to conduct the survey interviews. The
2,902 households included in the analyses performed for this paper are distributed across 37
villages: 10 villages in Shandong, 6 villages each in Henan and Hunan, and 5 villages each in
Shaanxi, Jiangsu, and Chongqing. All villages within a province are clustered within a single
township, meaning all 37 villages are clustered within six townships (and therefore within six
5
counties). The survey sites were not selected randomly, but purposively. Because the six survey
sites were selected with the goal of maximizing regional and economic variation, the households
interviewed are not intended to be representative of rural China as a whole, but only of the six
counties from which they were sampled. All indications, however, suggest this is a representative
sample. Age, educational, income, and occupational distributions in the sample closely match
official statistics and published findings from nationally representative samples.
Each respondent was presented with a list of 16 problem types plus an open-ended “other
dispute” category, totaling 17 problems or potential disputes, allowing a respondent to report a
maximum of 17 and a minimum of zero grievances. The 16 fixed problem items are: (1) housing
land, (2) water use, (3) debt collection, (4) family planning, (5) consumer, (6) divorce, (7)
neighbor, (8) labor, (9) “responsibility land” (i.e., farmland contracted from the village) or
township and village enterprise contracting, (10) agricultural taxes (nongye fudan), (11)
household (e.g., elderly care or property division), (12) dealings with a government agency, (13)
personal injury (plaintiff), (14) property damage/loss, (15) personal injury or property damage
(defendant), and (16) children’s education. Question skip patterns were implemented to identify
respondents to whom a particular problem or difficulty was not applicable: Only respondents
who reported lending and subsequently attempting to recover money were asked about debt
collection problems; only respondents who reported being engaged to marry or ever married
were asked about divorce problems; and only respondents who reported working for others were
asked about labor problems. With the exception of residential housing land grievances, which
were not time-bound (owing to a design oversight), each grievance by definition occurred with
the past five years. With the exception of divorce, labor problems, dealings with government
6
agencies, personal injury, and property theft/damage, the questions about the remaining ten
grievances were worded in terms of the entire household, not just the respondent.
If the respondent indicated seeking the help of a third party, she was asked, through an
open-ended question, to describe the third party (“To whom or to which agency or unit did you
or a family member seek help?”). The interviewers were instructed to record the respondent’s
description of the third party (or third parties) on the questionnaire form verbatim. Coding the
unique responses—578 unique third party responses reported for all 956 instances of
approaching third parties reported by the 2,902 respondents—into sensible and manageable
categories was the primary challenge posed by this research strategy. But this challenge was
more than offset by the benefits of extremely rich data carrying the original voices of the
involved parties. Rather than relying on fixed, closed-ended third-party categories, the openended questions capture the real choices made by real actors rather than the a priori assumptions
of the investigators. Furthermore, thanks to this research strategy, we are not bound to a single
classification system, but instead can modify our typology of responses. For each of the 4,757
grievances reported by the survey respondents, one of the following courses of action (or
nonaction) was taken: (1) lumped it (did nothing), (2) negotiated bilaterally, (3) mobilized an
informal relation, (4) mobilized a village leader, (5) approached a higher-level administrative or
government office, (6) approached the police, 4 or (7) approached the legal system.
The respondent was also asked to provide a reason for approaching each third party he or
she reported. Like the question on the third party approached for help, the reason for approaching
4
In the remote chance an ambiguously identified local police substation (paichusuo) belongs to a village (rather than
to a township or higher level of government), I replicated all analyses after recoding “police” as “village leaders.”
Coding changes do not significantly alter the empirical results; the substantive conclusions I draw from the results
are identical.
7
this third party was also asked in an open-ended manner and recorded verbatim. I use this
information to identify third parties initially identified as official higher-level offices or
officeholders that were actually informal relations, such as a close relative who happens to work
in the state bureaucracy. When respondents indicated seeking the help of a higher-level third
party because of an informal connection, I recoded the response third party as an “informal
relation.” However, as with all other measures, my results are entirely robust to a variety of
operational definitions. 5
The most important empirical test of the justice from above theory uses information on
evaluations of the performance of different types of third parties elicited from questions asking
whether the outcome and the process of the third-party intervention exceeded, matched, or fell
short of expectations. Separating procedural justice (the way a third party handled the process of
disputing) and distributive justice (the outcome obtained by the third party) permits a more
nuanced and textured picture than would be possible relying on only a single dimension of
satisfaction (see Tyler 1990).
5
More specifically, I recoded as “informal connections” 12% of all 685 instances of seeking third parties originally
coded as “village leaders,” “higher-level administrative or government office,” or “legal system” because the
reported reasons for approaching these higher-level third parties were informal connections (e.g., a prior relationship
with someone working in the office approached, or what O’Brien and Li [2005, 2006] call “mediated contention”). I
took this precautionary measure in order to ensure that any results showing more positive evaluations of the
performance of village leaders and more negative evaluations of the performance of higher-level leaders are not
artifacts of a concentration of informal connections at lower levels and a relative lack of informal connections at
higher levels. However, my findings are highly robust to a variety of coding methods. Coding changes do not
significantly alter the empirical results; the substantive conclusions I draw from the results are identical.
8
Although I analyze a wide range of responses to grievances, from lumping it to bilateral
negotiation to a variety of third parties (on this and similar typologies of responses, see Koch
[1978:4-6]; Felstiner et al. [1980/81:635-6]; Mather and Yngvesson [1980/81:776]), I only
analyze satisfaction with trilateral action—action involving a third party. In other words, I am
only comparing satisfaction between different types of third parties; I cannot analyze satisfaction
with bilateral solutions and with lumping it because the questionnaire elicited evaluations of
trilateral action only.
The Volume and Character of Grievances
Just over half of the respondents (55%) reported at least one grievance from the list of 17
potential grievances on the questionnaire. Almost one-fifth (19%) reported four or more
grievances. However, the geographical distribution of grievances is highly skewed. In the poorer
inland Henan and Hunan samples, 66% of the households surveyed reported at least one
grievance, while only 23% of households surveyed in the more prosperous coastal Jiangsu and
Shandong samples reported at least one grievance. In the Henan and Hunan samples, 26%
reported four or more grievances, while in the Jiangsu and Shandong samples, only 1% reported
this many grievances. Although the poorer inland sites account for 34% of the total sample, they
account for 64% of all reported grievances. At the same time, although the richer coastal sites
account for 31% of the total sample, they account for only 7% of all reported grievances.
Because together they account for 76% of all instances of third-party mobilization in pursuit of
justice, the Henan and Hunan samples dominate the analyses in this paper. At the other end of
the spectrum, the Jiangsu and Shandong samples account for merely 6% of all instances of
seeking the help of third parties in response to grievances.
9
The most commonly reported type of grievance is some sort of conflict with a neighbor,
followed, in descending order, by grievances over water use, agricultural taxes, and some sort of
intra-family issue. These four grievance types account for 47% of all grievances, and 47% of all
households in the sample report at least one of these four grievance types. The probability a
grievance escalates into a dispute involving the intervention of a third party varies enormously
by grievance type. Overall 20% of all grievances escalate to trilateral action, meaning 80% of all
grievances are either lumped or resolved through bilateral negotiation. At the same time, 7% of
all grievances escalate above and beyond the village. Although the housing land grievance
category ranks sixth in terms of prevalence, it is by far the most likely grievance to escalate
above and beyond the village, and thus appears to be the most contentiously disputed problem
type. Conversely, although the agricultural taxes category ranks third in terms of grievances, it
ranks sixth in terms of escalating to higher authorities.
This pattern is consistent with evidence reported elsewhere. According to published
estimates, one-third to 40% of all petitioners (urban and rural combined) complain about land
(CCCCP 2005:68; Zhong 2001:17) and over 60% of rural petitioners complain about land (Wang
2005:68). In a mere ten years over 70 million peasants—a number that continues to rise—have
reportedly been dispossessed of land (Kahn 2004; Yardley 2004; also see Guo 2001). 6 Recent
village riots over unfair land compensation and outright land seizures have been well publicized
(French 2006; Kuhn 2006). In first half of 2004, there were almost 50,000 recorded instances of
efforts to protect rural land rights. Of more than 130 mass conflicts in 2004 studied by a research
6
According to Zhao (2004), between 1990 and 2002, an estimated 66.3 million peasants lost their land. Another
estimate puts the number of peasant who lost their land between 1996 and 2003 at over 34 million (CCCCP
2005:68).
10
team, 87 were about land. Land expropriation and land compensation problems were raised in
55% of 837 letters randomly sampled from a collection of complaint letters sent to central
authorities (Zhao 2004). In a 2004 sample of 632 petitioners in Beijing, 73% reported a
grievance over land requisition (Zhang L. 2005). 7
Petitioning the State in Pursuit of Justice
It is clear from the data that to focus on appeals to higher authorities including the legal system is
to miss the overwhelming bulk of the action (including nonaction). Out of all 4,757 reported
grievances, in only 87 (less than 2%), was either a lawyer, court, or some other office of the
judicial system approached. Nor should it be surprising that bilateral negotiation is more
prevalent than lumping it (46.8% and 33.1% respectively of all grievances). As Felstiner (1974,
1975) has argued, in rural settings the social costs of lumping disputes with other villagers are
exceedingly high, a point I will elaborate later. Popular reliance on village leaders also emerges
in high relief. Of all 956 grievances going to third parties, 31% are brought to village leaders, 9%
are brought to the legal system, and 23% are brought to intervening government or
administrative offices. This pattern is similar to Zweig’s (2002:45) finding from a 1999 survey in
Anhui Province that, among all actions taken in response to problems reported by villagers, half
involved seeking the help of local village leaders and 27% involved seeking the help of higher-
7
In the survey data, “problems with responsibility land or contracting with a township or village enterprise” is a
single item on the survey questionnaire containing two discrete types of grievances and therefore eludes obvious
classification. Unlike housing land grievances, this grievance type, like agricultural tax grievances, is empirically
associated with agricultural production, tends to be contained within villages, and is rarely escalated to higher levels
outside the village.
11
level government authorities. It is to these intervening layers of bureaucracy that I now direct our
attention.
The perils of overlooking the significance of bureaucratic solutions in the disputing
process emerge in high relief in the foregoing findings. Administrative hierarchies are such a
central dimension of social and political organization in China (Barnett 1967; Lieberthal and
Oksenberg 1988:143-7; Chao 1991; Lieberthal and Lampton 1992) that “going up a level” (zhao
shangji or zhao lingdao) in pursuit of a remedy is the most natural—indeed the default—course
of action when bilateral negotiation fails (Shi 1997:51-4; Xin 1999:448). Underscoring the
salience of higher-level administrative solutions in China is one of the great contributions of the
justice from above school (e.g., O’Brien and Li 1995). Indeed, Emerson’s (1992) critique of the
tendency of students of disputing to overlook or downplay the significance of bureaucratic
solutions is arguably nowhere more applicable than in China.
As in many parts of the world, in China there is an implicit, and often explicit, ordering
of dispute forums. That is, there is a popular, shared understanding of the third party actors and
forums that legitimately constitute first resorts, middle resorts, and last resorts in the disputing
process. While courts tend to be last resorts everywhere, what may distinguish China from other
countries is the extent to which this tiered character of disputing is formally institutionalized.
Dispute processing institutions are arranged in a formally prescribed order, usually beginning
with bilateral negotiation (xieshang jiejue). Typically speaking, only when this fails do people
work progressively upwards, from workplace or community mediation, for example, to higher
levels of the administrative hierarchy, approaching the courts only under extreme circumstances
as a last resort. Indeed, under many circumstances it is prohibited to file a court petition before
other administrative channels are first exhausted. Li (1977) has contrasted China’s system of
12
hierarchical layers of intervention with other systems of conflict management elsewhere in the
world using the metaphor of a slow slide to the bottom (China) versus a sudden drop off the cliff
(the United States). In other words, relative to other countries, in China there are more layers of
bureaucracy intervening in social conflict and functioning to inhibit entry into the court system
(the metaphorical drop to the bottom).
Direct appeals to administrative and government offices are a mainstay action in the
disputing process. One of the most important targets of direct appeals is the xinfang system
(translated literally as “letters and visits”). Formalized in the early 1950s (Wang and Chen
1987:129; Zhang H. 2005:11), xinfang offices have multiplied and strengthened in recent years
as part of an official effort to contain the growing volume of conflict and to preserve social
stability (Wang and Huang 2003; Luehrmann 2003; Cai 2004; Minzner 2006). The xinfang
system includes a large network of complaints offices vertically spanning many levels of
government and horizontally spanning many administrative jurisdictions. Besides the official
Xinfang Administration, there are xinfang departments in the State Council, in the public
security system, in the courts, in the procuracy, and in the Women’s Federation, just to name a
few examples.
Estimates of the annual volume of petitions made to the xinfang system are in the range
of 10 to 13 million. 8 Meanwhile, nationwide, the courts heard about 4.4 million civil cases and
88,000 administrative law suits in the year 2003 (SSB 2004:Table 23-19). But the importance of
8
In the first three quarters of 2002, 8.6 million complaints were made nationally in xinfang offices at the county-
level of government and higher (Wang and Huang 2003). Extrapolating this to the full year yields an estimate of
8.6×1⅓=11.5 million. The volume of complaints made to the xinfang system in 2003 increased by 14% over the
previous year (Zhang H. 2005:12; Yu 2005:27). According to Zhao and Su (2004), the number of petitions made to
the xinfang system in 2003 was in excess of 10 million.
13
petitioning outside the legal system becomes even more apparent when we consider the many
other targets of direct appeals in China besides the xinfang system. According to one estimate,
only two of every thousand direct appeals to state authorities are made through the official
xinfang system (Zhang H. 2005:12). 9 The term shangfang (literally meaning “to petition” or “to
appeal to a higher authority”) is almost synonymous with xinfang. Indeed, virtually any
government office can serve as a site for citizen complaints, for direct appeals and petitions (see
Cai 2004). Besides these terms, additional vocabulary to describe taking grievances above and
beyond the village include “reporting situations” (fanying qingkuang), “filing a written
complaint” (gaozhuang), “lodging a complaint or appeal” (tousu), and “reporting a complaint”
(jubao).
Zhang Yimou’s classic 1992 film, The Story of Qiuju, poignantly reveals the tiered
character of disputing through the story of a peasant woman’s pursuit of justice after informal
conciliation efforts failed. Working her way up through township and county police channels,
she eventually, after much time and enormous personal expense, found her way into the city
court system with the help of a lawyer. A study of 184 peasant appeals to the The Farmers
Daily 10 affirms the authenticity of Qiuju’s experience: The majority of the complainants in this
study had worked their way up through each successive level of government—the village,
township, county, city, and provincial governments—and were in Beijing appealing to central
government authorities because all earlier attempts had failed. While in Beijing, most
9
Because it implies an obviously implausible 5 billion complaints made annually to government offices, this figure
may have been hyperbolically constructed to make the simple point that the xinfang system processes only a tiny
minority of all petitions.
10
Aggrieved villagers often appeal to newspaper publishing houses (O’Brien and Li 2004:86-8; Zweig 2003:119;
Bernstein and Lü 2003:177-8, 185).
14
complainants approached three or more central government offices and media outlets (Zhao
1999). A 2004 study of 632 rural petitioners in Beijing found that the number of government
agencies to which the average petitioner appealed was more than six, the maximum being 18.
The most common agencies approached were the State Xinfang Administration, the Standing
Committee of the National People’s Congress, the Supreme People’s Court, the Chinese
Communist Party’s Central Committee for Discipline Inspection, the Public Security
Administration, the Supreme People’s Procuratorate, the State Land Resources Administration,
the Ministry of Agriculture, and the Ministry of Civil Affairs (Zhao and Su 2004).
This tiered character of disputing is captured in the survey data. Indeed, one respondent
appears to have emulated Qiuju: a 70 year-old male in the Hunan sample, in response to a
dispute with a neighbor, sought the help of village cadres before taking the matter to the
township government and police substation (paichusuo) and eventually appealing to the city
government. Although, of all 578 unique descriptions of third parties provided by respondents,
not a single one contains the words xinfang or shangfang, which in fact should not be terribly
surprising insofar as these are generic and ambiguous terms for appealing directly to government
offices, the responses nevertheless show a hefty incidence of direct appeals to higher levels of
government administration: 26% of all reported third parties (or 33% of all unique descriptions
of third parties) contain the words “government” (zhengfu) (74 grievances), “agency” (suo) (77
grievances), “department” (bumen) (48 grievances), “leaders” (lingdao) (34 grievances),
“bureau” or “administration” (ju) (21 grievances), “office” (ban) (17 grievances), “organ”
15
(jiguan) (12 grievances), “bring to the attention” (fanying) (6 grievances), or “higher level”
(shangji) (4 grievances). 11
Higher-level government and administrative agencies are particularly prominent in
housing land disputes. Whereas 4% of grievances over agricultural taxes escalated to a
government office above the village, 14% of housing land grievances escalated to these
intervening layers of government bureaucracy. Indeed, in the survey data the Land
Administration is reported 33 times as a third party approached for help (27 times in response to
housing land grievances) and the Real Estate Administration is reported 11 times (all in response
to housing land property grievances). Unsurprisingly, personal injuries are unique in their high
likelihood of mobilizing formal law: A full 13% go to the formal legal system and 9% go to the
police, compared to overall averages of 2% and 1% respectively.
Since the respondents were not explicitly instructed to supply information about more
than one third party, the survey data undoubtedly underreport the incidence of layer-by-layer,
hierarchical movement through bureaucratic channels. Nevertheless, the limited information
available about movement between forum types clearly reveals that people are relatively
reluctant to appeal to higher authorities directly and that—entirely consistent with the
hierarchical character of disputing—they are more likely to start locally and to move
11
These words are not mutually exclusive; some responses contain more than one. What is somewhat surprising is
the absence of the media from any responses. In Zweig’s (2002:45) survey of villagers in Anhui, “report to mass
media” accounted for 5.5% of all actions taken in response to the problems they reported. Also see Li and O’Brien
(1996:48).
16
incrementally upwards. While the questionnaire only asks for information about the third party,
9% of third party responses contain multiple categories of third parties. 12
Movement between village and township levels appears in 24 responses (representing 25
disputes); movement between township and county levels appears in 8 responses (8 disputes). In
6 disputes the respondent reported three separate administrative agencies approached for help; in
4 disputes the respondent reported four or more separate administrative agencies. Two examples
are illustrative. First, a 72 year old male in Hunan reported seeking help with a housing land
dispute from the Land Administration, the Water Resources Administration, and Transportation
Administration. Second, an 82 year old female in Hunan reported seeking help with the same
type of dispute from the Land Administration, the Water Resources Administration, the Roads
Administration, and the Forestry Administration.
Figure 1 shows that the probability of reporting multiple types of third parties increases
with distance and formality; people resort to higher-level third parties only after first
approaching lower-level third parties. Grievances tend to reach higher levels not immediately,
but rather through escalation. Disputes appearing at higher levels more likely than not emerged
12
Out of all 956 disputes, in 870 disputes only a single third party type was reported. In all 86 instances of multiple
third party types, I coded upwards: I privilege higher-level third parties over lower-level third parties on the
assumption that a formal, more official third party further away from the village is more likely than an informal,
local third party to be the final, decisive, or most recent third party approached for help. In other words, I assume
that, more often than not, the forum indicated is the last one, that higher-level third parties are more likely to be the
end of line than the beginning of the line. For example, the assumption is that respondents who approached a village
leader before going to court are more likely to report the court than the village leader. To ensure the results are
robust to a variety of coding methods, I replicated all results after recoding the third parties in a manner that
privileges lower-level third parties rather than higher-level third parties. Coding changes do not significantly alter
the empirical results; the substantive conclusions I draw from the results are identical.
17
from lower levels. This pattern reflects a strong association between the stickiness or
intractability of the problem and the hierarchical location of the third party approached for help.
When a third party at a lower level—such as the village head, the township government, or
another local state agency—is unable to resolve a given dispute, the complainant often keeps
complaining to higher levels in search of redress. Thus, not only are the squeaky wheels the ones
that eluded grease at prior stages, but because they are the most difficult ones to grease they are
likely to continue to elude grease at subsequent stages at well. 13
[ FIGURE 1 ABOUT HERE ]
Testing the Theory of Justice from Above
By establishing that the layers of bureaucracy intervening between the village and the legal
system are of considerable importance in the disputing process, the findings presented thus far
offer some support to the theory of justice from above. However, we have also seen that people
tend to turn to higher levels not right away, but generally as a last resort. By testing whether
13
It is worth noting that the vast majority of these disputes seem, at least in the eyes of the respondents, to have
reached a conclusion. Although respondents had the opportunity to indicate whether a dispute remained unresolved
(i.e., still in process), in only 36 (or 4%) of all reported disputes did they indicate so. Although small numbers
prohibit definitive conclusions, the patterns are consistent with the story that squeaky wheels that approach local
third parties are more likely to get greased, while those that approach higher levels represent the most difficult
problems that eluded grease at lower levels and are therefore likely to continue to elude grease at higher levels:
Disputes brought to higher-level third parties above and beyond the village are almost four times more likely than
disputes brought to village leaders to elude resolution (6.2% versus 1.7%; a χ2 equality test shows the difference is
significant at p<.004).
18
higher-level solutions are more or less effective than lower-level solutions, we now turn to a
more direct test. Controlling for geographical location and basic individual and household
background characteristics, what kinds of third parties are most likely to disappoint complainants?
In order to answer this question I calculate ordered logistic regression models for evaluations of
the dispute outcome (distributive justice) and for evaluations of the dispute process (procedural
justice). To ensure that the results of my analysis are not an artifact of the character of disputes
most likely to escalate to higher levels, I have also controlled for dispute type. That is, I discuss
the net effects of third parties—the effects of different third parties on satisfaction with dispute
outcomes and experiences among otherwise seemingly identical households with seemingly
identical grievances. 14 The descriptive characteristics of the variables employed in the analysis
are presented in Table 1. Regression results are presented in Table 2.
[ TABLE 1 ABOUT HERE ]
[ TABLE 2 ABOUT HERE ]
We can see from the regression models that, relative to the performance of village leaders,
and holding constant the nature of the dispute in question, legal institutions and the police are
evaluated significantly more negatively. To aid the interpretation of the regression coefficients
14
Although small numbers of observations prohibit either the introduction of interaction terms in the regression
models or the calculation of separate regression models for each dispute type, an additional measure I took to ensure
that variation in evaluations of the performance of third parties is not explained entirely in terms of the nature of the
dispute at hand was to cross-tabulate satisfaction by third party separately for each dispute type. The bivariate crosstabulations confirm the main findings presented here.
19
(which I transformed into odds ratios), I calculated the predicted probably of each evaluation
response (i.e., “exceeded expectations,” “met expectations,” or “failed to meet expectations”) for
each third party category, holding all other independent variables constant at sample means.
Figure 2 presents predicted probabilities—which can be interpreted directly and which are
meaningful in an absolute sense—calculated from both regression models in Table 2.
[ FIGURE 2 ABOUT HERE ]
In diametrical opposition to the theory of justice from above, positive evaluations are
associated with local informal solutions, while negative evaluations are associated with higherlevel, more formal solutions. Thus, the evidence fails to support the expectation from the theory
of justice from above that higher-level government offices are more likely than village leaders to
deliver desired outcomes. With respect neither to outcome nor to procedure are higher-level
administrative and government agencies more likely than village leaders to be favorably
evaluated. At the same time, the results do not suggest village leaders are more likely than
higher-level government offices to deliver desired outcomes. With respect to procedure, however,
village leaders are about 40% more likely to “exceed expectations” and about 25% less likely to
“fail to meet expectations.” These findings are consistent with other evidence that village leaders
have become allies in the struggles of ordinary villagers (Thornton 2004; Cody 2005).
However, an interesting pattern with respect to housing land disputes lends some measure
of support for the theory of justice from above. Since the regression coefficients for dispute type
presented in Table 2 are all in comparison to housing land disputes (the omitted reference group),
20
the data show that housing land disputes, the largest source of direct appeals to higher levels of
the state bureaucracy, are more likely than any other dispute type to produce satisfaction.
It is also apparent that, although evaluations of outcome and process are highly correlated,
people seem to be somewhat more satisfied with procedures than with outcomes. 15 An analysis
of the few discrepancies between distributive justice and procedural justice further challenge the
theory of justice from above: Overall, 6.6% of all reported third parties for all 956 reported
disputes meet or exceed expectations with respect to process even when they fail to deliver the
desired outcome. Even when they fail to deliver what complainants want, village leaders are
more likely than any other type of third party to be perceived as fair. When distributive
expectations were not met, village leaders were more than three times more likely than
administrative/government offices above the village to meet or to exceed procedural expectations
(31.7% versus 9.4%). 16
15
Evaluations of outcome and process are identical in 84% of all reported disputes. This degree of overlap is not
dramatically lower than the 92% of identical answers in a similar survey carried out in Beijing in the summer of
2001 (Michelson 2002). On the one hand this pattern may alarm some observers by suggesting that, insofar as a
desirable outcome is generally a precondition of a desirable process, people approach the law instrumentally in the
same manner as functionally comparably means of achieving a given end, including the mafia. On the other hand,
however, this pattern is consistent with research findings from elsewhere in the world. For example, Genn (1999:202)
finds that respondents who lost their court cases were over nine times more likely than respondents who won their
court cases to report that the court acted unfairly. As one respondent in Scotland put it when asked to elaborate on
why she expressed satisfaction with a judge’s performance, “Well, he gave me what I wanted...” (Genn 1999:218).
16
More specifically, among the 104 disputes brought village leaders that failed to meet distributive expectations,
almost one in three nonetheless met or exceeded procedural expectations. In contrast, among the 53 disputes brought
to administrative/government offices above the village that failed to meet distributive expectations, less than one in
ten met or exceeded procedural expectations. A χ2 equality test shows the difference is significant at p<.002.
21
These findings are consistent with recent estimates that only two out of every thousand
rural petitioners find any sort of resolution (Zhao and Su 2004). Xinfang personnel lament their
own ineffectiveness in resolving the grievances of those who approach them, and have reportedly
coined the following saying to express their impotence: “We are in charge of everything but can
take charge of nothing; we cannot overlook anything but are able to neglect everything” (shenme
dou guan, shenme dou guanbuliao; shenme dou bu neng bu guan, shenme ye dou keyi bu guan;
Zhao 2003:23). When they do try to resolve grievances, xinfang personnel may forward
supporting documents brought by the complainant to the relevant government agency—which,
more often than not, had already been approached by the complainant to no avail—and to send
the complainant home. Another method is to summon leaders from the complainant’s place of
residence to accompany the complainant home where they are to resolve the matter amongst
themselves. In either case, the actions of the xinfang office serve to bring the grievance full circle
and to bring “the destination back to the point of origin” (zhongdian you hui dao qidian; Zhao
1999). By prolonging the search for justice from above, the futility of petitioning the state only
contributes to the seemingly inexorable expansion of the ranks of rural petitioners. Consequently,
prominent academics, including Yu Jianrong (at the Chinese Academy of Social Sciences) and
He Weifang (at Peking University), have begun calling for either seriously reforming or
altogether eradicating the xinfang system (see Yu 2005 and Li 2005).
But the survey data suggest that, even more than direct appeals to state agencies,
appealing to the judicial system, including the courts and lawyers, is even more likely to produce
disappointment. Compared to village leaders, the legal system is 75% more likely to fail to meet
outcome expectations (and about 70% less likely to exceed expectations and about 40% less
likely to meet expectations) and more than 100% more likely to fail to meet procedural
22
expectations (and about 70% less likely to exceed expectations and about 35% less likely to meet
expectations). These findings support Bernstein and Lü’s (2003:190-6) bleak assessment of the
accessibility and effectiveness of the court system for the problems facing China’s villagers.
Indeed, recent research suggests the growing population of rural petitioners is, to some measure,
a consequence of the failure of the courts. Almost two-thirds of the 632 petitioners surveyed in
Beijing in 2004 had tried local courts first. Among those who had appealed to the courts, 43%
reported that the courts had refused to hear their case (Zhao and Su 2004; Yu 2005:27). Although
national and local policies prohibit skipping or bypassing administrative levels, such official
efforts are routinely ignored or surmounted by petitioners (O’Brien and Li 1995:766,778;
O’Brien and Li 2004:89; Cai 2004:447; Li 2001:585n39; Bernstein and Lü 2003:183). Indeed,
the failure of higher-level of institutions to absorb complaints, which subsequently spill over to
lower-level government authorities, has been labeled an “upside down pyramid” (dao jinzita)
(Zhang L. 2005:11).
Not only do courts reject the cases of China’s aggrieved villages, but lawyers, too, tend to
screen out rural clients. According to a study of screening practices in the Beijing bar, lawyers do
their best to avoid representing shrewd and unyielding “rightful resisters” (diaomin). Petitioners
and migrant workers are perceived by lawyers as low-quality (suzhi di) troublemakers who, by
reneging on a fee agreement or by filing a malpractice suit, for example, may turn their tactics of
resistance against their own lawyers (Michelson 2006:18-21). Although it is ironic that fitting the
profile of a “rightful resister,” of someone who aggressively mobilizes legal resources, is a
liability when it comes to hiring a lawyer, lawyers’ aversion to representing some of the neediest
members of Chinese society helps explain why villagers report so much difficulty obtaining
justice from above. Moreover, the cultural stereotypes about the “low quality” of villagers that
23
inform lawyers’ profiling practices undoubtedly shape screening practices elsewhere, including
the petition filing sections of lower courts, legal aid offices, and so on.
How can we reconcile these findings with the theoretical expectations of justice from
above? In the sections that follow I identify and discuss methodological and theoretical
limitations that help account for the findings of the justice from above school. Overcoming these
methodological and theoretical limitations weakens support for the theory of justice from below
position and strengthens an alternative theory of justice from below.
Methodological Limitations of Justice-from-Above Research
To illustrate the problem of selection bias in the simplest terms, the squeaky wheel theory
proceeds from a research design that systematically excludes from consideration the squeaky
wheels that do get the grease at lower levels, obviating the need to escalate to higher levels. If
one’s research design entails deliberately seeking out only the most troublesome problems
brought to higher authorities, by definition one is limiting the scope of analysis to the problems
that had eluded successful resolution at earlier stages. This type of selection bias is known
alternately as “sample truncation” and “selecting on the dependent variable” (see Winship and
Mare 1992; Geddes 1990; and Collier and Mahoney 1996). Insofar as complainants’ satisfaction
with various third parties depends on how high their grievances escalate, the exclusion of less
extreme grievances resolved at lower levels will necessarily lead to biased estimates of the
performance of competing third parties. In other words, research designed to capture the most
extreme problems that had already escalated to full-blown disputes brought to higher-level third
parties, and that accordingly ignores problems that did not escalate, (1) will obviously exaggerate
assessments of the magnitude and severity of local problems and (2) will unduly bias
24
assessments of the dispute-resolution capacity of local village institutions versus that of higher
authorities.
One of my central arguments in this paper is that appearances at higher levels of the
dispute pyramid implies failure at lower levels, and that failure at lower levels is associated with
more negative assessments of the dispute-resolution capacity of lower-level less formal third
party intermediaries. In short, the findings that undergird the justice from above position, I
submit, are in no small measure artifacts of selection bias. The justice from below position, in
contrast, emerges from a research design that includes grievances that may be abandoned,
grievances that may be resolved locally, and grievances that may be pursued at higher levels.
Only by including all these possibilities can the relative effectiveness of appealing to higher
authorities be fairly assessed.
More specifically, two sources of selection bias impeding the study of disputing in
general and the study of disputing in China in particular are forum bias and grievance bias.
Forum bias refers to the restriction of the scope of analysis to formal legal and administrative
settings and to the exclusion of alternative third-party actors and forums. Grievance bias,
meanwhile, refers to the exclusion of the less thorny problems more amenable to lower-level,
local solutions and to the disproportionate inclusion of the intractable grievances more likely to
elude local resolution and to percolate up to higher-level legal or administrative forums. Of
course in practice these two types of selection bias overlap: Higher-level forums tend to select
for tougher grievances. Therefore these two sources of bias can be thought of as a general
selection bias syndrome.
The upshot is that valid comparisons of the performance of competing dispute forums—
i.e., a proper empirical test of justice from above versus justice from below—demands the
25
inclusion of a greater variety of grievances brought to a greater variety of forums. As Felstiner et
al. (1980/81:636) write, “A theory of disputing that looked only at institutions mobilized by
disputants and the strategies pursued within them would be seriously deficient.” This statement
would be even more fitting had the word “institutions” been replaced with “legal institutions.”
Forum bias stems from the premise inherent in much survey research that all roads lead to court
and other parts of the state bureaucracy. Other research, however, tells us that disputes follow a
less linear, less teleological, and more meandering path, twisting and turning both vertically and
horizontally, often never even entering the court system (see Merry [1990:92, 95] for such a
critique). The disputing process is characterized by much forum shifting; complainants
frequently move through multiple institutions throughout the course of a single dispute. Thus, a
proper evaluation of the impact of lawyers and courts, for example, demands some knowledge
about the problems that do not avail themselves of lawyers and courts but are rather dealt with
through alternative means. If we want to evaluate how effectively labor laws protect aggrieved
workers, for example, we need to know something about the labor problems that do not get
processed according to official labor grievance procedures. Finally, to draw meaningful
conclusions about the effectiveness of administrative litigation presupposes some knowledge
about citizens’ grievances with government agencies that do not end up as legal petitions. (See
Gallagher 2005; Guthrie 2002; Pei 1997; and Peerenboom 2001 for additional examples of
research designs that embody this methodological limitation.) In short, in assessing the question
of access to justice, it is essential to take a bottom-up approach that captures more of the dispute
pyramid than the implicitly and explicitly top-down approaches taken in much of the research on
disputing. In the absence of the denominator, the numerator is of little value.
26
Forum bias is often difficult to disentangle from grievance bias: Grievances typically
percolate to higher-level forums as a last resort only after the failure of lower-level solutions.
Thus, disputes entering the legal system and other formal settings tend to be the most intractable
problems, problems that that eluded remedy at lower levels and consequently likely to continue
to elude resolution at higher levels. This principle can be illustrated by the phenomenon of the
“three 80-percents” that has emerged in popular discourse in China: It is popularly asserted that
over 80% of complaints made in the xinfang system have been determined to be (1) complaints
about basic-level government, (2) complaints made for good reason, or (3) complaints that were
produced by the neglect of local agencies (e.g., Zhao 2003:24; RFA 2005; Zhang L. 2005). 17 As
we have already seen, the squeaky wheel is the one that did not get oiled at earlier levels and that
subsequently kept on squeaking to higher levels where it will remain unlikely to get oiled. In
other words, when a person directly appeals for help to a higher authority, there is a considerable
probability the grievance at hand has a long history of prior attempts at resolution:
Most disputes are resolved by the parties themselves and only those cases involving anger, intractable
problems, or both end up before judges and mediators. Hence, authorities are constantly faced with the
most difficult cases....In such a situation, it is seldom possible to give both parties what they want or feel
they deserve. Dissatisfaction may, therefore, by inevitable. (Tyler 1997: 875)
Insofar as complaints made to higher authorities are more likely to have failed to receive
redress at earlier stages, a research design that privileges complaints that escalate—i.e., that
17
In another published report, the various “80 percents” are actually “over 80%” and refer to collective complaints.
A fourth “80%” sometimes cited is, “over 80% of complaints reflect problems created by the reform and
development process” (Wang and Huang 2003; Zhang H. 2005:12).
27
discounts complaints made to local authorities—is not well suited to an assessment of the
question of justice from above versus justice from below. If we only look at full-blown, escalated
disputes, we will overlook a substantial population of grievances that do get resolved by local
authorities to the satisfaction of complainants. Thus, to assess the question of access to justice,
we need to start looking far upstream at a wider array of problems and to examine how they
develop and escalate. As the findings I presented above demonstrate, when we consider a greater
variety of grievances in a greater variety of forums, higher authorities no longer look so
appealing and lower authorities begin looking a lot better.
Theoretical Limitations of Justice-from-Above Research
A classic in the canon of legal anthropology, The Disputing Process—Law in Ten Societies
(Nader and Todd 1978), deals directly with the issue of forum choice in a variety of rural settings,
including Lebanon, Turkey, and Mexico. One of the central empirical themes of this book is the
importance and desirability of flexible, informal, customary methods of dispute resolution vis-àvis the relatively rigid performance of the courts (also see Abel 1974). Owing to the importance
of preserving local social relationships, local solutions are normally preferable to outside
solutions. This finding has been replicated repeatedly in studies of why people choose and how
they experience state law versus customary, indigenous law in Africa (Gluckman 1955; Gulliver
1969; Comaroff and Roberts 1981), the United States (Llewellyn and Hoebel 1941), and Mexico
(Nader 1965; Collier 1973; Parnell 1978, 1988), for example.
People resort to higher-order solutions when lower-order solutions fail (Emerson 1981).
While formal jurisdiction can partially explain the hierarchical order of remedies, social reasons
are at least as important. An important theoretical proposition in the ethnographic literature on
disputing is that social relationships discourage direct appeals to the state, including formal legal
28
action, that there are important social limits and social disincentives to approaching higher
authorities for help with local problems. Thus, the theory of justice from below posits that people
choose third parties bearing in mind the social costs of escalating disputes. While people have a
significant measure of legal freedom to choose dispute forums, their choices are also socially
constrained.
At the foundation of this theoretical proposition is the distinction between multiplex and
simplex relationships. While multiplex relations are diffuse and affective, spanning multiple
domains (e.g., neighbors who are also colleagues whose children attend the same school),
simplex relationships are limited to a single domain (Gluckman 1955). In general, the closer the
relationship between two parties approaches the multiplex end of the continuum, the less likely
formal legal action (including litigation) will occur in the event of a dispute between these two
parties (Merry 1982; Felstiner 1974). Insofar as “Village dispute processes effectively preserve
ongoing, multiplex, affectional relations” (Starr 1978:138), the likelihood of escalation to higher
authorities increases in proportion to the unimportance of the complainant’s relationship with her
adversary (Felstinter 1974, 1975). 18
18
As Galanter (1974:130) writes, “Which human encounters are we likely to find regulated at the ‘official’ end of
our scale and which at the ‘private’ end? It is submitted that location on our scale varies with factors that we might
sum up by calling them the ‘density’ of the relationship. That is, the more inclusive in life-space and temporal span a
relationship between parties, the less likely it is that those parties will resort to the official system and more likely
that the relationship will be regulated by some independent ‘private’ system.” Similarly, according to Black, “If we
know that the relational distance between the adversaries differs in two otherwise identical cases, we can predict
which is likely to attract more law....We cannot be certain, but our ability to anticipate the result is vastly improved.
In sum: Law varies directly with relational distance.” (Black 1989:12, emphasis in original; also see Black 1976:408; 1973:133-4). Finally, as Engel (1984:577) has observed in rural Illinois, pressures to litigate were seen as
“intrusions upon existing relationships...”
29
More so than their urban counterparts, villagers embroiled in disputes risk damaging
multiplex relationships (Gluckman 1955; Van Velson 1969; Kidder 1973; Felstiner 1974, 1975;
Engel 1978, 1984; Kritzer et al. 1991:502, 532-4; Greenhouse et al. 1994:Chapter 4). In the case
of a dispute between two villagers, the relationship between petitioner and respondent typically
spans multiple roles; social and economic intercourse can rarely be avoided. “Today they are
disputing in court, tomorrow they may be collaborating in the same work party” (Van Velsen
1969:138, cited in Nader and Todd 1978:12-13). It is worth repeating Gluckman’s (1969:22) oftquoted statement that “...the parties (and often the judges too) are normally involved in complex
or multiplex relations outside the court-forum, relations which existed before and continue after
the actual appearance in court,...” (cited in Nader 1984:77). Owing to the importance of
multiplex relationships, villagers tend to prefer informal, negotiated settlements to formal, legal
ones: “Villagers avoid and resent the ‘law powers’ of their fellow villagers. They prefer to
negotiate on an informal basis, since the outcomes are more flexible and negotiation as a process
is part of everyone’s daily, routinized behavior” (Starr 1978:149). Thus, “Courts are resorted to
where an ongoing relationship is ruptured” (Galanter 1974:108). To be sure, the centrality of
multiplex social relationships does not preclude the possibility of litigation in rural settings, but
does result in a greater propensity of local courts to adopt mediation, conciliation, and
negotiation (see Kidder 1973; Zhu Suli 2000; Liu 2006). Although the importance of maintaining
relations is more prominent in rural areas, the same pattern has also been found in urban areas
(Conley and O’Barr 1990; Merry 1990). Even American businessmen have been observed
privileging social relationships over contractual relationships (Macaulay 1963).
Finally, direct appeals to higher authorities carry the palpable risk of revenge, the threat
of local retribution in response to the mobilization of legal solutions. Parnell (1978:317)—
30
echoing Collier’s (1973:228) statement that “...few plaintiffs take their cases to San Cristobal
[the state court] because they prefer restitution and reconciliation to revenge and continuing
hostility”—writes, “...those who choose the state are faced with the prospect of revenge and
continuing hostility.” Revenge not only reflects the failure of state coercion (Rieder 1984), but
also the intended consequence of state power (e.g., Madsen 1990) and the unintended
consequence of the mobilization of state law. Zweig (2003:119) underscores the importance of
revenge and retribution as a disincentive in rural China to appeal to higher authorities: “In fact,
winning a case is not always the best solution as cadre revenge may negate a victory in court:
According to one article, family members said that: ‘No matter whether you win or lose, you lose.
You still have to live here and the township government won’t forgive you.’”
Conclusions and Directions for Future Research
I have endeavored to demonstrate in this paper that, after expanding the scope of empirical
observation to a wider array of third party actors and forums than has been measured in previous
research, satisfactory resolution to disputes appears to be more forthcoming from local, less
formal third parties than from higher-level, more formal third parties. This empirical pattern,
revealed through survey data from rural China on real-life grievances and real-life strategies of
redress, is inconsistent with “justice from above” expectations that complainants in rural China
fare better appealing to higher levels of the state, including the legal system, than to local
authorities. My findings also point to the importance of moving beyond the study of generalized
political trust based on questions posed to villagers in the abstract without reference to actual,
concrete encounters with the institutions they are asked to evaluate. While Li (2001:579n30) has
asserted that, insofar as villagers are reluctant to criticize local leaders, their assessments of
village leaders are biased in a positive direction, there are also good reasons the opposite may be
31
true, why any such bias may be moderated or cancelled out by a countervailing source of bias.
To the extent that it is politically legitimate to criticize local leaders but not politically legitimate
to criticize the Center (e.g., Weston 2004:75), greater expressed trust in the Center could also be,
to an important measure, an artifact of political desirability bias (see Rosen 1987; Manion 1994),
namely, the imperative to choose the politically “correct” answer.
The empirical patterns I have revealed reaffirm an older “justice from below” research
tradition based on research in a variety of societies around the world consistently identifying
local informal solutions as the preferred and more effective means of balancing the needs of
justice with the imperative to preserve local social relationships. While I hesitate to invoke for a
second time the 1992 Chinese film, The Story of Qiuju, I do so only because it so aptly illustrates
the “justice from below” theoretical proposition about the social limits and social costs of
petitioning higher levels of the state. From a strictly technical standpoint, Qiuju’s efforts to
escalate her dispute (a personal injury her husband sustained by a kick to the groin inflicted by
the village head in response to some measure of verbal provocation in a land dispute) to higher
administrative levels of the state were successful. By ultimately winning her court case in the
second instance on a legal technicality, justice, in the narrowest of senses, was delivered.
However, this was a hollow victory insofar as it undermined Qiuju’s overriding desire to repair
and preserve her family’s relationship with the village head and his family. The ambivalence of
the director, Zhang Yimou, with respect to the ability of higher levels of the state, including the
legal system, to resolve village-level disputes is captured by the expression of horror on Qiuju’s
face as she watches the village head being dragged away by the police the day after he saved her
life and the life of her baby in a complicated childbirth.
32
Indeed, consistent with the theory of justice from below, a popular reliance on village
leaders to solve problems emerges in high relief from the survey data. Not only are village
leaders a more popular source of help than other third parties located above and beyond the
village, but village leaders also appear, relative to higher-level, more formal sources of help, to
be adept at solving problems to the satisfaction of local complainants. In contrast to the popular
image of social conflict in rural China in which village leaders are portrayed as a primary source
of problems that escalate to higher levels, the survey data presented in this paper suggest
precisely the opposite, that village leaders in fact do more to solve and contain problems from
escalating to higher levels.
The lessons from rural China for anyone interested in studying disputing, conflict
resolution, and access to justice anywhere in the world include new methodological possibilities
for using survey techniques to overcome the problem of an overly narrow array of grievances
and third parties approached for help—the problem of selection bias afflicting, to greater and
lesser extents, much of the prior research on disputing. By excluding from consideration disputes
successfully resolved informally by local levels, research designs that privilege difficult cases
brought to higher-level third parties will inevitably produce biased estimates of the desirability
and effectiveness of the competing third-party choices available to aggrieved individuals. In
short, a defensible assessment of the dispute resolution capacity of various third parties cannot
proceed from a consideration only of what happens to the toughest cases at the highest echelons.
Notwithstanding the methodological strategy that led to the empirical and theoretical
contributions in this paper, there are many tasks for future research, many areas in need of
improvement in the future. One of the tasks for future research is to open up the black box of
dispute processing: What is it about local village leaders that causes their performance to be
33
evaluated relatively positively and perceived as relatively fair? To be sure, part of the answer lies
with the relative stickiness and intractability of the problems escalated to higher levels. Even if
state actors closer to the Center are as able and willing as village leaders to resolve the local
disputes of villagers, insofar as disputes brought to higher levels are, on balance, more difficult
to resolve than disputes brought to lower levels (because efforts at resolution are more likely to
have failed at lower levels), state actors closer to the Center are consequently more constrained
in their ability to deliver justice to aggrieved peasants. But a more thorough answer to the
question of the precise methods and mechanisms of dispute processing must await further
research.
One of the paramount goals of China’s legal reforms is to preserve social stability, and
thus regime stability, by resolving the popular grievances of an increasingly contentious citizenry.
In the process of expanding the courts and popularizing legal education, the state has heightened
popular expectations of justice from above and has equipped aggrieved individuals with the
weapon of the law. Therefore, in a context in which the legal system systematically denies
justice to a highly aggrieved and potentially volatile segment of society, an unintended and
paradoxical consequence of the legal reforms may be to jeopardize social stability and to erode
regime legitimacy.
34
References
Abel, Richard L. 1974. “A Comparative Theory of Dispute Institutions in Society.” Law &
Society Review 8(2): 217-347.
Barnett, A. Doak. 1967. Cadres, Bureaucracy, and Political Power in Communist China. New
York: Columbia University Press.
Bernstein, Thomas B., and Xiaobo Lü. 2003. Taxation without Representation in Contemporary
Rural China. Cambridge and New York: Cambridge University Press.
Black, Donald. 1973. “The Mobilization of Law.” Journal of Legal Studies 2: 125-39.
—————. 1976. The Behavior of Law. New York: Academic Press.
—————. 1989. Sociological Justice. New York: Oxford University Press.
Cai, Yongshun. 2004. “Managed Participation in China.” Political Science Quarterly 119(3):
425-51.
CCCCP [Central Committee of the Chinese Communist Party’s Nanjing Municipal Party
Committee Party School Research Team]. 2005. “Shidi Nongmin de Liyi Weihu Yu
Goujian Nongmin Hexie Shehui Yanjiu” (Protecting the Rights and Interests of Peasants
Who Lost Their Land and Research on Constructing a Harmonious Society). Zhonggong
Nanjing Shiwei Dangxiao Nanjing Shi Xingzheng Xueyuan Xuebao 4:68-74.
Chao, Chien-min. 1991. “T’iao-t’iao vs. K’uai-k’uai: A Perennial Dispute between the Central
and Local Governments in Mainland China.” Issues and Studies 27(8): 31-46.
Chen, Alfred H.Y. 2002. “Socio-legal Thought and Legal Modernization in Contemporary China:
A Case Study of the Jurisprudence of Zhu Suli.” Paper presented at the Law and Society
Association Annual Meeting, Vancouver, BC, Canada, May 30-June 1.
China Daily. 2004a. “New Petitioning Rules in Pipeline.” January 15, accessed from
<http://www.china.org.cn/english/government/84796.htm> on June 15, 2004, on file with
author.
—————. 2004b. “Citizen’s Complaints: Give All Justice.” March 13, accessed from
<http://www.chinadaily.com.cn/english/doc/2004-03/13/content_314386.htm> on June
15, 2005, on file with author.
Cody, Edward. 2005. “China’s Rising Tide of Protest Sweeping Up Party Officials: Village
Chiefs Share Anger over Pollution.” The Washington Post, September 12, p.A01.
Collier, David, and James Mahoney. 1996. “Insights and Pitfalls: Selection Bias in Qualitative
Research.” World Politics 49(1): 56-91.
Collier, June F. 1973. Law and Social Change in Zinacantan. Stanford, CA: Stanford University
Press.
Comaroff, John L. and Simon Roberts. 1981. Rules and Processes: The Cultural Logic of
Dispute in an African Context. Chicago: University of Chicago Press.
Conley, John M., and William M. O’Barr. 1990. Rules versus Relationships: The Ethnography of
Legal Discourse. Chicago: The University of Chicago Press.
Emerson, Robert M. 1981. “On Last Resorts.” American Journal of Sociology. 87(1): 1-22.
—————. 1992. “Disputes in Public Bureaucracies.” Pp. 3-29 in Studies in Law, Politics and
Society, Vol. 12, Part A, edited by Susan S. Silbey and Austin Sarat. Greenwich, CT: JAI
Press.
Engel, David. 1978. Code and Custom in a Thai Provincial Court. Tuscon, AZ: The University
of Arizona Press, for the Association of Asian Studies.
35
—————. 1984. “The Oven Bird’s Song: Insiders, Outsiders, and Personal Injuries in an
American Community.” Law & Society Review 18(4): 551-82.
Felstiner, William L.F. 1974. “Influences of Social Organization on Dispute Processing.” Law &
Society Review 9(1): 63-94.
—————. 1975. “Avoidance as Dispute Processing: An Elaboration.” Law & Society Review
9(4): 695-706.
Felstiner, William L.F., Richard L. Abel, and Austin Sarat. 1980/81. “The Emergence and
Transformation of Disputes: Naming, Blaming, Claiming...” Law & Society Review 15(34): 631-54.
French, Howard W. 2006. “Police in China Battle Villagers In Land Protest.” The New York
Times January 16, p.1.
Galanter, Marc. 1974. “Why the ‘Haves’ Come out Ahead: Speculations on the Limits of Legal
Change.” Law & Society Review 9(1): 95-160.
Gallagher, Mary E. 2005. “‘Use the Law as Your Weapon!’ The Rule of Law and Labor Conflict
in the PRC,” Pp. 54-83 in Engaging the Law in China: State, Society and Possibilities for
Justice, edited by Neil J. Diamant, Stanley Lubman, and Kevin J. O’Brien. Stanford, CA:
Stanford University Press.
Geddes, Barbara. 1990. “How the Cases You Choose Affect the Answers You Get: Selection
Bias in Comparative Politics.” Political Analysis 2: 131-50.
Genn, Hazel. 1999. Paths to Justice: What People Do and Think About Going to Law. Oxford
and Portland, OR: Hart Publishing.
Gluckman, Max. 1955. The Judicial Process among the Bartose of Northern Rhodesia.
Manchester: The University Press.
—————. 1969. Ideas and Procedures in African Customary Law. London: Oxford
University Press.
Greenhouse, Carol J., Barbara Yngvesson, and David Engel. 1994. Law and Community in Three
American Towns. Ithaca, NY and London: Cornell University Press.
Gulliver, P.H. 1969. “Dispute Settlement without Courts: The Ndendeuli of Southern Tanzania.”
Pp.24-68 in Law in Culture and Society, edited by Laura Nader. Chicago: Aldine.
Guthrie, Doug. 2002. “The Transformation of Labor Relations in China’s Emerging Market
Economy.” Pp.139-70 Research in Social Stratification and Mobility, Vol. 19 (The
Future of Market Transition), edited by Kevin T. Leicht. Amsterdam: JAI Press.
Guo, Xiaolin. 2001. “Land Expropriation and Rural Conflicts in China.” The China Quarterly
166: 422-39.
Halton, William, and Michael McCann. 2004. Distorting the Law: Politics, Media, and the
Litigation Crisis. Chicago: University of Chicago Press.
Huang, Cary. 2005. “Premier Vows to Spread the Wealth: Agricultural Taxes Will Be Abolished
Next Year, Two Years Ahead of Schedule, Promises Wen Jiabao.” South China Morning
Post, March 6, p.6.
Johnson, Ian. 2004. Wild Grass: Three Stories of Change in Modern China. New York: Pantheon.
Kahn, Joseph. 2004. “China’s Elite Learn to Flaunt It While the New Landless Weep.” The New
York Times, December 25, p.A1.
—————. 2005. “China to Cut Taxes on Farmers and Raise Their Subsidies.” The New York
Times, February 3, p.A3.
Kidder, Robert L. 1973. “Courts and Conflict in an Indian City: A Study in Legal Impact.”
Journal of Commonwealth Political Studies 11: 121-39.
36
Kuhn, Anthony. 2006. “Inside China’s Angry Villages.” Los Angeles Times February 11, p.B17.
Koch, Claus-Friedrich. 1978. “Access to Justice: An Anthropological Perspective,” Pp.1-7 in
Access to Justice, Vol. IV, The Anthropological Perspective, Patterns of Conflict
Management: Essays in the Ethnography of Law, edited by Klaus-Friedrich Koch. Milan:
A. Giuffrè; Alphenaandenrijn: Sijthoff and Noordhoff.
Kritzer, Herbert M., W.A. Bogart, and Neil Vidmar. 1991. “The Aftermath of Injury: Cultural
Factors in Compensation Seeking in Canada and the United States.” Law & Society
Review 25(3): 499-543
Kwan, Daniel. 2005. “Building Social Harmony Top of Party’s Agenda: Hu’s Concept Expected
to be the Hot Topic at Two Major Parliamentary Conferences this Week.” South China
Morning Post, February 27, p.6.
Li Li. 2005. “Judicial Independence Should Come First.” Beijing Review 48(45) November 10.
Li, Lianjiang. 2001. “Support for Anti-Corruption Campaigns in Rural China.” Journal of
Contemporary China, 10(29): 573-86.
—————. 2004. “Political Trust in Rural China.” Modern China 30(2): 228-58.
Li, Lianjiang, and Kevin J. O’Brien. 1996. “Villagers and Popular Resistance in Contemporary
China.” Modern China 22(1): 28-61.
Luehrmann, Laura M. 2003. “Facing Citizen Complaints in China, 1951-1996.” Asian Survey
43(5): 845-66.
Li, Victor H. 1977. Law without Lawyers: A Comparative View of Law in China and the United
States. Stanford, CA: Stanford Alumni Association.
Lieberthal, Kenneth G. and David M. Lampton, eds. 1992. Bureaucracy, Politics, and Decision
Making in Post-Mao China. Berkeley, CA: University of California Press.
Lieberthal, Kenneth, and Michel Oksenberg. 1988. Policy Making in China: Leaders, Structures,
and Processes. Princeton, NJ: Princeton University Press.
Liu, Sida. 2006. “Beyond Global Convergence: Conflicts of Legitimacy in a Chinese Lower
Court.” Law & Social Inquiry 31:75-106.
Llewellyn, Karl N., and E. Adamson Hoebel. 1941. The Cheyenne Way: Conflict and Case Law
in Primitive Jurisprudence. Norman, OK: University of Oklahoma Press.
Lofquist, William S. 2002 “Closing the Courthouse Door: Constructing Undeservingness in the
Tort and Habeas Corpus Reform Movements,” Sociological Spectrum 22: 191-223.
Macaulay, Stewart. 1963. “Non-Contractual Relations in Business: A Preliminary Study,”
American Sociological Review 28(1): 55-67.
Madsen, Richard. 1990. “The Politics of Revenge in Rural China during the Cultural
Revolution,” Pp. 175-201 in Violence in China: Essays in Culture and Counterculture,
edited by Jonathan N. Lipman and Stevan Harrel. Albany, NY: State University of New
York Press.
Manion, Melanie. 1994. “Survey Research in the Study of Contemporary China: Learning from
Local Samples.” China Quarterly 139: 741-65.
Marquand, Robert. 2004. “In China, Stresses Spill Over into Riots.” The Christian Science
Monitor, November 22, p. 1.
Mather, Lynn, and Barbara Yngvesson. 1980/81. “Language, Audience, and the Transformation
of Disputes.” Law & Society Review 15(3/4): 775-822.
Merry, Sally Engle. 1982. “The Social Organization of Mediation in Nonindustrial Societies:
Implications for Informal Community Justice in America,” Pp.17-45 in The Politics of
37
Informal Justice, Vol. 2: Comparative Studies, edited by Richard L. Abel. New York:
Academic Press.
—————. 1990. Getting Justice and Getting Even: Legal Consciousness Among WorkingClass Americans. Chicago: The University of Chicago Press.
Michelson, Ethan. 2002. “How Much Does Law Matter in Beijing?” paper prepared for Law and
Society Association Annual Meeting, Vancouver, BC, Canada, May 30-June 1.
—————. 2006. “The Practice of Law as an Obstacle to Justice: Chinese Lawyers at Work.”
Law & Society Review 40(1):1-38.
Minzner, Carl. 2006. “Xinfang: An Alternative to the Formal Chinese Legal System.” Stanford
Journal of International Law 42.
Nader, Laura and Harry F. Todd Jr., eds. 1978. The Disputing Process—Law in Ten Societies.
New York: Columbia University Press.
Nader, Laura. 1965. “Choices in Legal Procedure: Shia Moslem and Mexican Zapotec.”
American Anthropologist 67(2): 394-99.
—————. 1984. “From Disputing to Complaining,” Pp. 71-94 in Toward a General Theory of
Social Control, Volume 1: Fundamentals, edited by Donald Black. Orlando, FL:
Academic Press.
O’Brien, Kevin J. 1996. “Rightful Resistance.” World Politics 49(1): 31-55.
O’Brien, Kevin J., and Lianjiang Li. 1995. “The Politics of Lodging Complaints in Rural China.”
China Quarterly 143: 756-83.
—————. 2004. “Suing the Local State: Administrative Litigation in Rural China.” The
China Journal 51:75-96.
—————. 2005. “Popular Contention and its Impact in Rural China.” Comparative Political
Studies 38(3): 235-59.
—————. 2006. Rightful Resistance: Contentious Politics in Rural China. New York and
Cambridge: Cambridge University Press.
Parnell, Philip. 1978. “Village or State? Competitive Legal Systems in a Mexican Judicial
District,” Pp. 315-50 in The Disputing Process—Law in Ten Societies, edited by Laura
Nader and Harry F. Todd Jr. New York: Columbia University Press.
—————. 1988. Escalating Disputes: Social Participation and Change in the Oaxacan
Highlands. Tuscon, AZ: The University of Arizona Press.
Peerenboom, Randall. 2001. “Globalization, Path Dependency and the Limits of Law:
Administrative Law Reform and Rule of Law in the People’s Republic of China.”
Berkeley Journal of International Law 19(2): 161-264.
Pei, Minxin. 1997. “Citizens v. Mandarins: Administrative Litigation in China.” China Quarterly
152: 832-62.
RFA (Radio Free Asia). 2005. “New Rules Do Little to Help China’s Petitioners.” May 16,
accessed from
<http://www.rfa.org/english/news/in_depth/2005/05/16/china_petitioners/> on June 15,
2005, on file with author.
Rieder, Jonathan. 1984. “The Social Organization of Vengeance,” Pp. 131-62 in Toward a
General Theory of Social Control, Volume 1: Fundamentals, edited by Donald Black.
Orlando, FL: Academic Press.
RMRB [Renmin Ribao]. 2005. “Ba Nongcun Shuifei Gaige Quanmian Yinxiang Shenru” (An
All-Round Deepening of the Rural Tax Reforms), June 8, p.1.
38
Rosen, Stanley. 1987. “Survey Research in the People’s Republic of China: Some
Methodological Problems.” Canadian and International Education 16(1): 190-7.
Shi, Tianjian. 1997. Political Participation in Beijing. Cambridge, MA: Harvard University
Press.
Shue, Vivienne. 2004. “Legitimacy Crisis in China?” Pp. 24-49 in State and Society in 21stcentury China: Crisis, Contention, and Legitimation, edited by Peter Hays Gries and
Stanley Rosen. New York and London: RoutledgeCurzon.
SSB [State Statistical Bureau]. 2003. China Statistical Yearbook. Beijing: State Statistical
Publishing House.
Starr, June. 1978. “Turkish Village Disputing Behavior,” Pp. 122-51 in The Disputing Process—
Law in Ten Societies, edited by Laura Nader and Harry F. Todd Jr. New York: Columbia
University Press.
Straughn, Jeremy Brooke. 2005. “‘Taking the State at Its Word’: The Arts of Consentful
Contention in the German Democratic Republic.” American Journal of Sociology 110(6):
1598-650.
Tang, Yuen Yuen. 2005. “When Peasants Sue En Masse: Large-scale Collective ALL Suits in
Rural China.” China: An International Journal 3(1): 24-49.
Thornton, Patricia M. 2004. “Comrades and Collectives in Arms: Tax Resistance, Evasion, and
Avoidance Strategies in Post-Mao China.” Pp. 87-104 in State and Society in 21stcentury China: Crisis, Contention, and Legitimation, edited by Peter Hays Gries and
Stanley Rosen, New York and London: RoutledgeCurzon.
Tyler, Tom R. 1990. Why People Obey the Law. New Haven, CT: Yale University Press.
—————. 1997. “Citizen Discontent with Legal Procedures: A Social Science Perspective on
Civil Procedure Reform.” American Journal of Comparative Law 45: 871-904.
Upham, Frank. 2005. “Who Will Find the Defendant If He Stays with His Sheep? Justice in
Rural China,” Yale Law Journal 114:1675-1718.
Van Velsen, Jan. 1969. “Procedural Informality, Reconciliation, and False Comparisons.”
Pp.137-52 in Ideas and Procedures in African Customary Law, edited by Max Gluckman.
London: Oxford University Press.
Wang Gan. 2005. “Goujian Hexie Shehui Ying Qieshi Weihuhao Shidi Nongmin de Hefa
Quanyi” (In Constructing a Harmonious Society We Should Earnestly Protect the Lawful
Rights and Interests of Peasants Who Lost Their Land). Yangzhou Daxue Shuiwu
Xueyuan Xuebao 10(3):68-70.
Wang Xiantang, and Chen Hongbin. 1987. Xinfang Xue Gailun. (An Overview of the Study of
Letters and Visits) Shenyang: Liaoning Daxue Chubanshe.
Wang Yongqian, and Huang Haiyan. 2003. “Guojia Xinfang Ju Juzhang: 80% Shangfang You
Daoli.” (The Director of the State Xinfang Administration: 80% of Complaints are
Reasonable) Ban Yue Tan, November 20.
Weston, Timothy B. 2004. “Dilemmas of Party Adaptation: The CCP’s Strategies for Survival.”
Pp. 67-86 in State and Society in 21st-century China: Crisis, Contention, and
Legitimation, edited by Peter Hays Gries and Stanley Rosen. New York and London:
RoutledgeCurzon.
Winship, Christopher, and Robert D. Mare. 1992. “Models for Sample Selection Bias.” Annual
Review of Sociology 18:327-50.
Xin Chunying. 1999. Zhongguo de Falü Zhidu Ji Qi Gaige (Chinese Legal System & Current
Legal Reform). Beijing: Falü Chubanshe.
39
Xinhua News Agency. 2005. “Scrapping of Agricultural Tax Relieves Farmers of Big Burden.”
January 19, accessed from <www.chinaview.cn> on January 20, 2005, on file with author.
Yardley, Jim. 2004. “Farmers Being Moved Aside by China’s Real Estate Boom.” The New York
Times, December 8, p.A1.
Yu Jianrong. 2005. “Zhongguo Xinfang Zhidu Pipan” (A Critique of China’s Complaint System).
Zhongguo Gaige 2:26-8.
Zhang Haihong. 2005. “Xin Zhongguo Xinfang Zhidu de Youlai Yu Xianzhuang” (The Origins
and Current Condition of New China’s Xinfang System). Tuanjie 11:11-5.
Zhang Liping. 2005. “Ba Wenti Jiejue Zai Jiceng” (Resolve Problems at the Lowest Levels).
Gongchan Dangyuan 5:10-11.
Zhao Donghui. 2003. “Xinfang de Tizhi Pingjing Jidai Tupo: Rang Minyi Shun Chang Shang
Song” (A Bottleneck in the Xinfang System in Urgent Need of a Breakthrough: Let the
Sentiments of the People Flow Upwards Unobstructed) Liao Wang Xinwen ZhouKan 40
(October 6): 23-5.
Zhao Ling. 2004. “Diaocha Xianshi Nongmin Weiquan Zhongxin Chuxian Zhongda Bianhua”
(Survey Reveals a Major Shift in the Focus of Peasants’ Rights Activism) Southern
Weekend, September 2.
Zhao Ling and Su Yongtong. 2004. “Guonei Shou Fen Xinfang Baogao Huo Gaoceng Zhongshi”
(China’s First Report on Petitioning System Received High-Level Attention). Nanfang
Zhoumo, November 4.
Zhao Shukai. 1999. “Shangfang Shijian He Xinfang Tixi—Guanyu Nongmin Jin Jing Shangfang
Wenti de Diaocha Fenxi” (Complaint Incidents and the Xinfang System: An Analysis of
a Survey on the Problem of Peasants Coming to Beijing to Make Complaints).
Unpublished report dated December 1999 (the report’s author is a researcher in the State
Council’s Development Research Center), accessed June 15, 2005 at
<http://www.usc.cuhk.edu.hk/wk_wzdetails.asp?id=2355>, on file with author.
Zhong Qing. 2001. “Farmers Speak Out.” Beijing Review. 32(August 9): 17-8.
[Zhu] Su Li. 1996. Fazhi Jiqi Bentu Ziyuan. (Rule of Law and Its Indigenous Resources) Beijing:
Zhongguo Zhengfa Daxue Chubanshe.
—————. 2000. Song Fa Xia Xiang: Zhongguo Jiceng Fazhi Zhidu Yanjiu. (Bringing Law
Down to the Countryside: Research on China’s Basic-Level Legal System) Beijing:
Zhongguo Zhengfa Daxue Chubanshe.
Zweig, David. 2002. Democratic Values, Political Structures, and Alternative Politics in Greater
China. Peaceworks No. 44. Washington, DC: United States Institute of Peace.
—————. 2003. “To the Courts or to the Barricades: Can New Political Institutions Manage
Rural Conflict?” Pp. 113-35 in Chinese Society: Change, Conflict and Resistance, 2nd
Edition, edited by Elizabeth J. Perry and Mark Selden. London and New York:
RoutledgeCurzon.
40
Figures and Tables
Figure 1. Movement between Third-Party Categories, Rural China, 2002
lawyer, court, or judicial office (n=90)
70%
police (n=60)
30%
73%
administrative/government office above village (n=214)
27%
79%
village leader (n=391)
84%
informal relation (n=295)
90%
all third-party types
83%
0%
20%
40%
60%
80%
Proportion of Disputes
21%
16%
10%
17%
100%
Legend:
% disputes for which this is the only reported third party type
% disputes for which additional third party types are reported
NOTE: 956 disputes reported by 585 households. χ2=27.1, p<.001. In 870 disputes, one third party was
reported; in 78 disputes two third-parties were reported; and in 8 disputes three third parties were reported. (1
× 870) + (2 × 78) + (3 × 8) = 1,050 = (90 + 60 + 214 + 391 + 295).
41
Figure 2. Predicted Probabilities of Various Evaluations of Disputing Outcome and Process,
Rural China, 2002
A. Evaluations of Dispute Outcome
lawyer, court, or judicial office
2%
police
35%
4%
63%
44%
52%
administrative/government office above village
7%
57%
37%
village leader
7%
57%
36%
informal relation
10%
overall predicted probabilities
63%
7%
0%
57%
20%
27%
36%
40%
60%
80%
Proportion of Disputes
100%
B. Evaluations of Dispute Process
lawyer, court, or judicial office
3%
police
40%
5%
administrative/government office above village
56%
48%
8%
46%
57%
36%
village leader
11%
62%
27%
informal relation
11%
62%
27%
overall predicted probabilities
9%
0%
20%
59%
40%
60%
80%
Proportion of Disputes
32%
100%
Legend:
exceeded expectations
met expectations
failed to meet expectations
NOTE: The above estimated probabilities were calculated from the regression models presented in Table 2. The
“overall predicted probabilities” were calculated by letting all independent variables equal their means (as presented
in Table 1). The predicted probabilities for the various third parties were calculated by holding all variables constant
at their means (as presented in Table 1) except the third-party variables. The response categories comprising each
third party type do not always sum to 100% owing to rounding error.
42
Table 1. Descriptive Statistics of Variables
Mean
DEPENDENT VARIABLES
evaluations of dispute outcome
1. exceeded expectations
2. met expectations
3. failed to meet expectations
evaluations of dispute process
1. exceeded expectations
2. met expectations
3. failed to meet expectations
St. Dev.
Min.
Max.
2.283
.088
.542
.371
2.226
.106
.562
.332
.615
.283
.499
.483
.622
.308
.496
.471
1
0
0
0
1
0
0
0
3
1
1
1
3
1
1
1
.369
.319
.172
.050
.091
.483
.466
.377
.218
.288
0
0
0
0
0
1
1
1
1
1
3.752
31.698
6.651
3.061
1.133
10.472
2.052
.871
1
11.5
1
1
7
75
14
5
SAMPLE
Jimo (Shandong)
Ru’nan (Henan)
Taicang (Jiangsu)
Zhong (Chongqing)
Yuanjiang (Hunan)
Hengshan (Shaanxi)
.026
.571
.034
.074
.188
.107
.160
.495
.182
.262
.391
.309
0
0
0
0
0
0
1
1
1
1
1
1
DISPUTE TYPE
water use
debt collection
family planning
consumer
divorce
neighbor
labor
responsibility land or enterprise contracting
agricultural taxes
household/family
dealings with gov’t office
personal injury
property damage/loss
children’s education
accused of personal injury of theft
housing land
other
.035
.032
.091
.046
.020
.183
.027
.086
.066
.072
.059
.053
.050
.028
.010
.026
.114
.185
.176
.288
.209
.142
.387
.163
.281
.248
.258
.236
.225
.218
.166
.101
.160
.318
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
THIRD PARTY APPROACHED
informal relation
village leader
administrative/government office above village
police
lawyer, court, or judicial office
HOUSEHOLD CHARACTERISTICS
# members
mean age
mean years of education (if age>14)
subjective relative economic status
NOTE: N=879 disputes among 536 families in 32 villages. Subjective relative
economic status is measured as the response to the following question: “Compared
to other families in your village, do you feel your family’s economic situation is
better or not?” Response categories form a five-point scale: “(1) My family’s
economic situation is much worse than other families; (2) somewhat worse than the
typical family; (3) about the same as others; (4) somewhat better than other families;
and (5) much better than that of other families.”
43
Table 2. Correlates of Evaluations of Dispute Outcomes and Experiences,
Odds Ratios from Ordered Logistic Regression Analysis
Disappointment
with Outcome
THIRD-PARTY HELP SOUGHT
informal relation
administrative/government office above village
police
lawyer, court, or judicial office
village leader (comparison group)
Disappointment
with Process
.677*
1.043
1.985#
3.031***
1.016
1.487#
2.333*
3.496***
1.000
.993
1.014
.885
1.047
.999
1.010
.885
DISPUTE TYPE
water use
debt collection
family planning
consumer
divorce
neighbor
labor
responsibility land or enterprise contracting
agricultural taxes
household/family
dealings with gov’t office
personal injury
property damage/loss
accused of personal injury or theft
children’s education
other
housing land (comparison group)
.921
1.845
1.011
.786
2.678
1.460
1.731
1.751#
2.246*
1.466
2.609*
1.174
3.592**
1.744
1.527
4.605**
1.639
2.767*
1.576
.813
3.050#
1.572#
1.687
1.506
1.840
1.516
2.534**
1.239
3.215**
2.001
.484
1.747
SAMPLE
Jimo (Shandong)
Ru’nan (Henan)
Taicang (Jiangsu)
Zhong (Chongqing)
Yuanjiang (Hunan)
Hengshan (Shaanxi) (comparison group)
3.455*
1.994*
1.022
2.592**
3.005***
3.038*
1.913*
.821
2.846**
2.479**
HOUSEHOLD CHARACTERISTICS
# members
mean age
mean years of education
subjective relative economic status
2
Pseudo R
Wald χ2
.066
91.3***
.046
65.4***
NOTE: # p<.10 * p<.05 ** p<.01 *** p<.001, two-tailed. N=879 disputes among 536
families in 32 villages. The dependent variables have three values: (1) exceeded expectations,
(2) met expectations, and (3) failed to meet expectations. An odds ratio of 1.000 means one’s
chances of evaluating a third part one point more negatively (on the three-point scale) neither
increase nor decrease. An odds ratio of 2.000 means one’s chances double. An odds ratio
of .500 means one’s chances halve. Whether or not multiple third parties were reported (a
possible proxy for the stickiness or intractability of the dispute and for how long it eluded
resolution) is also not statistically significant. Other household characteristics (such as the
presence of political leaders) are not statistically significant. The models were calculated with
sampling weights and robust standard errors.
44