Rupture and Representation: Migrant Workers, Unions and ... · 2/12/2011  · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (2024)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (1)

Rupture and Representation:Migrant Workers, Unions and the State in China


Eli David Friedman

A dissertation submitted in partial satisfaction of the

requirements of the degree of

Doctor of Philosophy



in the

Graduate Division

of the

University of California, Berkeley

Committee in Charge:

Professor Peter Evans, ChairProfessor Kim Voss

Professor Ching Kwan LeeProfessor Kevin O’Brien

Fall 2011

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (2)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (3)



Rupture and Representation: Migrant Workers, Unions and the State in China


Eli David Friedman

Doctor of Philosophy in Sociology

University of California, Berkeley

Professor Peter Evans, Chair

This project begins with a simple observation: during the first decade of the 21st

century, worker resistance in China continued to increase rapidly despite the fact thatcertain segments of the state began moving in a pro-labor direction. This poses a problemfor the Polanyian theory of the countermovement, which conflates social resistance to themarket with actual decommodification and incorporation of labor. I then pose thequestion, why is labor strong enough to win major legislative and policy concessionsfrom the state, but not strong enough to significantly benefit from these policies? The“partial” nature of the countermovement can be explained with reference to the dynamicsof labor politics in China, and specifically the relationship between migrant workers,unions, and the state, or what I refer to as “appropriated representation.” Because unionsin China are an invention of the state, they have good access to policy makers but arehighly illegitimate amongst their own membership, i.e. strong at the top, weak at thebottom. Labor’s impotence within enterprises means that pro-labor laws and collectiveagreements frequently go un-enforced. As a result, workers are forced to take radicalautonomous action in order to have their grievances addressed. Expanding workerinsurgency strengthens the hand of unions at the national (and potentially provincial andeven municipal) level, but fails to produce a durable re-alignment of power at the point ofproduction that is capable of enforcing laws. Thus, the central state’s initiatives toadvance the interests of migrant workers is simultaneously undermined by theircategorical ban on independent worker organization, a dilemma I refer to as insurgency

trap. Through an analysis of several most-likely cases, I empirically demonstrate theproblems generated by appropriated representation and attempt to discern under whatconditions insurgency trap can be undone. By reconfiguring the theory of thecountermovement, I gain conceptual clarity on the relationship between spontaneousresistance to the market and institutionalization of class compromise.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (4)



Throughout this project and indeed my entire graduate student career I have hadthe exceptional privilege of having Peter Evans serve as my advisor. At every step of theway, Peter has been an exemplary mentor, alternatively advising, instructing, cajoling,and simply leading by example. I have learned a tremendous amount from him, not justabout sociology and doing good scholarship, but also about how to act in the world. Itgoes without saying that this dissertation would not have been possible without hissupport and guidance.

Ching Kwan Lee has selflessly advised me for a number of years, despite the factthat she has in this time never held a position at Berkeley. My intellectual debt to her isapparent enough in the following pages, and if I can count any scholarly or professionalachievements they owe in large part to her ongoing support. Kim Voss has helped methink deeply about union organization, and always provides timely reminders thatChinese labor politics may not be as unique as I think. Kevin O’Brien has consistentlyprovided me with some of the best feedback and suggestions on my work and how tomake it comprehensible to others, as well as invaluable knowledge on conductingresearch in China. His course on contentious politics in China changed the way I thinkabout social movements and continues to be one of the best experiences I had in gradschool. Although Dylan Riley was not on my dissertation committee I would like toacknowledge his contribution to my intellectual growth. He provided me with a rigoroustraining in social theory that helped me to see the world in a radically new light.

The intellectual environment at Berkeley is unparalleled, and I have learned froman incredible group of fellow students. Many of the following people have read andcommented on parts of this dissertation; others have contributed to my development in aless direct but equally important manner: Abigail Andrews, Dan Buch, Ryan Calder, JiaChing Chen, Zongshi Chen, Julia Chuang, Barry Eidlin, Jonathan Hassid, Paul Hathazy,Daniel Immerwahr, Zach Levenson, Mike Levien, Kate Maich, Simon Morfit, MarcelParet, Tianna Paschel, Alina Polyakova, Jeff Sallaz, Aaron Shaw, and Rachel Stern.

My fieldwork would not have been possible without the help of numerous friendsand colleagues in China. Given the political sensitivity of the research I unfortunatelywill not acknowledge them by name. However a significant number of people around thecountry provided contacts, intellectual stimulation, as well as friendship while I was inthe field. I would also like to thank Anthony Spires and the faculty of the sociologydepartment at Chinese University of Hong Kong for hosting me while I conductedarchival research at the Universities Service Centre.

While conducting fieldwork I received generous support from the University ofCalifornia’s Pacific Rim Research Program as well as the Labor and EmploymentResearch Fund. My time at Chinese University of Hong Kong was supported by theGlobal Scholarship Programme for Research Excellence – CNOOC Grant.

My wife Julia Chang not only gave me unconditional love and support throughoutthe challenge of having to spend more than a year apart, but quite literally came up withthe title of this project. The acknowledgement section of a dissertation is an awkwardplace to formally recognize the profundity of a spousal relationship. Nonetheless, I loveher deeply and have counted on her for emotional and intellectual inspiration time andtime again.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (5)


Finally, there are my parents Ellen and Stuart, who have probably set some sort ofrecord for direct contribution to a child’s dissertation. For the first several months of myfieldwork (undoubtedly the most personally challenging part of the project) I lived in anapartment with them on the campus of Sun Yat-sen University in Guangzhou. They ofcourse fed me, listened to me gripe about the challenges of studying labor politics inChina, but also engaged in substantive daily conversations about the research whichprofoundly shaped the development of the project. Additionally, they have read nearlyevery word of this dissertation, sometimes reading it aloud to each other on long car trips.While it is perhaps customary to dedicate a first major project to one’s parents, mine haveactually earned it in spades. I love them very much and dedicate this work to them.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (6)


Table of Contents

Chapter 1 – Introduction p.1

Chapter 2 – Theory p.19

Chapter 3 – The History and Structure of the ACFTU p. 40

Chapter 4 – The Most Progressive Union in China p.64

Chapter 5 – Oligarchic Decommodification? Sectoral Unions and Crises ofRepresentation p.91

Chapter 6 – Worker Insurgency and the Evolving Political Economy ofThe Pearl River Delta p.124

Chapter 7 – Conclusion p.151

Bibliography p.164

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (7)


List of Figures and Images

Figure 1 p.3

Image 1 p.82

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (8)


Chapter 1

Introduction – Labor Politics and Capitalist Industrialization

Shortly after assuming leadership of the Chinese Communist Party in late 2002,Hu Jintao embarked on a significant rhetorical departure from his predecessors. DengXiaoping had famously argued that it was acceptable for “some to get rich first,” and that“it doesn’t matter if the cat is black or white as long as it catches mice,” both indicationsthat economic growth was the unquestioned and ultimate end of action. Jiang Zemin’s“Three Represents” affirmed the Party’s support for private capital and reinforced theprimacy of developing the productive forces. But Hu quickly – if subtly – moved to re-orient the state away from such a single-minded pursuit of growth. Over the course of hisfirst year and a half in office, he unveiled the key slogans that would be associated withhis tenure: “scientific development view,” “putting people first,” (yirenweiben) and mostfamously, “harmonious society.” Though imbued with slightly different shades ofmeaning, in sum these slogans were meant to indicate that the state would no longer beexclusively concerned with GDP growth as an end in itself. Under this new approach todevelopment, the state was to pay greater attention to environmental protection, reducinginequality, expanding the social welfare system, and enhancing rule of law. In short, Huwanted to take steps to soften the edges of the bare-knuckle laissez-faire capitalism that,while leading to many consecutive years of high growth, had resulted in stark classpolarization, ecological destruction, and rapidly expanding social conflict.

And indeed, over the next several years there were strong indications that thecentral government was backing away from full-throttle marketization and re-orientingtheir growth strategy away from one highly dependent on wage repression and export-oriented manufacturing. Although calls for a shift away from exports grew significantlyfollowing the global economic crisis of 2008, the government had been advocating anincrease of domestic consumption since at least 2004.1 In part responding to massiveprotests among laid-off workers in the late 1990s and early 2000s, the high wave ofprivatization of state-owned enterprises (SOE) subsided. It became clear that the publicsector was going to continue to play a large role in the economy, particularly in keyindustries such as energy, arms, transportation, finance, and education. Scholars andmedia commentators began to refer to the phenomenon of “advance of the state, privateretreat” (guo jin min tui) to refer to the process of re-nationalization happening in severalsectors. A series of pro-labor policies and laws were implemented, culminating in thelandmark Labor Contract Law approved in 2007. Additionally, the government took anumber of steps to reform the discriminatory hukou household registration system (Wang2010), increase social insurance coverage of migrant workers, and raised minimumwages. Most significantly for this study, the All China Federation of Trade Unions(ACFTU) appeared to be more aggressive in pushing for collective bargaining andunionizing private employers, as most clearly represented by the high profile Wal-Martcampaign in 2006. Unions around the country began to talk more assertively about

1 See the white paper “China’s Employment Situation and Policies” from the StateCouncil’s Information office: (accessed February 12, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (9)


organizing migrant workers and negotiating better contracts for their members to promote“harmonious labor relations.”

It appeared as if years of high levels of social unrest – chief amongst which waslabor conflict – had taken a toll on the state, and the central government was ready forcompromise. For some scholars, it seemed that the state had embraceddecommodification and a re-embedding of the economy in response to the chaos of themarket, just as theorized by Karl Polanyi in The Great Transformation (1944). Indeed,Wang Shaoguang argued that by the late 1990s,

… the golden tablet (jinzi zhaopan) of market reform toppled, shattering theseeming consensus on the efficacy of market forces… [Those hurt by marketization] feltthat Chinese economic reform had gone astray, and they longed for harmony between theeconomy and society. This initiated the protective countermovement to re-embed theeconomy into the society. (2008:21)

In Wang’s view, by 2008 the central government’s change in direction was alreadysuccessful: “By using state power, the [sic] redistribution breaks the market chain andreconnects everyone. These are the changes China has been experiencing recently.”(2008:22) But is “longing for harmony” enough? Was in fact the market chain broken,with everyone being reconnected?

From the perspective of 2011, we can surely say that Wang’s optimistic prognosiswas pre-mature. Particularly for migrant workers – the focus of this study – needs are stillby and large mediated by the market nexus. Managerial autonomy remains essentiallyuncompromised, and workplaces are subject to endemic legal violations. And workers arenot satisfied. Indeed, from the beginning of the Hu administration in late 2002 up untilthe conclusion of this study in 2010, the volume, and seemingly the intensity, of laborconflict increased without pause. Officially mediated disputes increased rapidly (seefigure 1), while autonomously organized strikes, road blockades, riots, and workersuicides continue to upend social order. In at least two high profile cases, workers thatmurdered their boss were widely hailed as heroes on the internet.2 By 2010, thegovernment was spending RMB 514 billion (US $78 billion) on internal security, nearlymatching its national defense budget of RMB 518.6 billion.3 Clearly, all was not peacefulin the People’s Republic.

2 I’m referring here to the Tonggang steel incident and the Liu Hanhuang case.3 Sun, Wukong. April 28, 2010. “Beijing Hears Dissenting Voice on Unrest.” Asia Times.(USD conversion is based on exchange rates in February, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (10)


Figure 1

Source: zhongguo laodong tongji nianjian 2010. [China Labor Statistical Yearbook 2010]Beijing: zhongguo tongji chubanshe.

This project then seeks to address a problem of the political economy of early 21st

century China: why is it that in the decade since the central government began to shiftaway from full-fledged marketization, worker unrest has continued to grow apace?4

Perhaps one might assume that the answer is simply that unions are weak, and thereforeworker interests continue to be violated. But if this is the case a second questionimmediately arises: why is it that labor is strong enough to win major concessions at thenational and sometimes provincical and municipal level, but not strong enough to allowworkers to signficantly benefit from these victories? In order to answer these questions, Ifocus on the state-controlled unions under the umbrella of the ACFTU, and theirrelationship to workers, capital, and other state agencies. Before outlining the argument,it will first be necessary to establish a basic conceptual framework.

Countermovements, The Institutional Moment, and Appropriated Representation

Just as Polanyi studied the “Great Transformation” of 19th and early 20th centuryBritain (1944), I am concerned here with similar tectonic social and political shifts thatderive from capitalist industrialization in contemporary China. Polanyi’s theory of the“double movement” held that the commodification of land, labor, and money would, if

4 A separate but related question is, why in this period has inequality also continued togrow rapidly? Although I do not aim to explain inequality here, the inability to enforcepro-labor legislation is certainly a primary factor.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (11)


left unchecked, result in the destruction of society and the ecosystem. However, Polanyiargued that commodification generated a countermovement for social protection fromvarious classes in society which would result in decommodification of labor (as well asland and money, but labor is most relevant for this study) and a re-embedding of markets.One of his key examples of a successful countermovement was the American New Deal.But for Polanyi, as well as several contemporary scholars, social resistance tocommodification is conflated with actual institutionalized class compromise, such as thatwhich characterized the post-war political economy of North America, Western Europe,and Japan.5

I make several adjustments to the Polanyian theory of countermovements toaccount for this difficulty. As discussed above, the seeming inclination of the Chinesecentral government towards class compromise has not resulted in a reduction in migrantworker insurgency, as labor remains highly commodified and labor conflicts often cannotbe resolved by unions or through other legal means. Theoretically, this allows us to seethat countermovements against commodification must be broken down into two distinct,if dialectically intertwined, moments: the insurgent moment in which social groupsmarginalized in the process of capitalist development engage in relatively dis-organizedand spontaneous resistance to commodification; and the institutional moment, when classcompromise is established in the political and economic spheres. I takedecommodification of labor as an indicator of the emergence of the institutional momentin the economic sphere; decommodification is defined as social action that lessens the

extent to which workers are immediately compelled to submit the satisfaction of their

needs to the logic of the market. Things such as guaranteed health care, pensions, jobsecurity, increased wages, and having a say in how the labor process is organized allcontribute to decommodification (see chapter 2 for an expanded and more highlyspecified definition). The political aspect of the institutional moment is represented byincorporation of the working class. This means that workers have substantiverepresentation both on the shopfloor (relationship to capital) and in giving the workingclass a voice in government (relationship to the state). If workers are able to resolvecollective problems and contend with capital within rationalized, legal channels(especially collective bargaining) and if they recognize the legitimacy of their legal unionrepresentatives, this serves as evidence of incorporation.

Decommodification and political incorporation are mutually reinforcing trends: tothe extent that workers have greater collective voice in the state and workplace, theireconomic standing is likely to improve; and improvements in economic standing arelikely to increase the legitimacy of union representatives and collective bargainingmechanisms. In China, the state and union were for a long time largely unconcerned withthe economic problem as evidenced by their presiding over a program of radical laborcommodification since the late 1970s. But over the course of the Hu administration,

5 For an analysis of the passage of the National Labor Relations Act in the U.S. that doesnot fit this description, but rather delves into the nuanced relationship between workerinsurgency and decommodification, see: Goldfield, Michael. 1989. "Worker Insurgency,Radical Organization, and New Deal Labor Legislation." The American Political Science

Review 83:1257-1282.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (12)


certain segments of the state (notably the central government and certain provincial andeven municipal governments) became increasingly interested in expanding workers’ability to consume. And the state has long been concerned with the political problem ofincorporation, as they fear instability that may result from expanding labor unrest. Butsince the political and economic are intimately linked, some degree ofdecommodification will likely be necessary in order to attain incorporation. Additionally,strengthened representation in the workplace – especially mobilizational capacity – couldincrease the power of union representatives within the state.

A schematic of my reformulation of countermovements appears below:


In these terms, then, my question is: why is it that the countermovement in Chinahas stalled at the insurgent moment? Why have high levels of resistance among migrantworkers resulted in legislative, regulatory, and symbolic6 victories but have not beentranslated into incorporation and decommodification? My claim is that the transition frominsurgency to the institutionalization of the countermovement (i.e. class compromise) thatwe would expect based on Polanyian theory has been short-circuited because the newclass of migrant workers in China have emerged under conditions of appropriated

representation. “Appropriated representation” is a term originally used by Weber(1978:292) which he juxtaposed to the radically democratic “instructed representation,”(ibid:293) but which he did not develop at any length. I have adopted and reconfiguredthe term to refer to a situation in which the state unilaterally grants exclusive rights ofpolitical representation of an entire class to a particular organization in the absence ofsubstantive or formalistic delegation from membership. Historically, unions in manyother countries undergoing capitalist industrialization have played a crucial role inchanneling insurgent worker energy into the construction of collective power capable of

6 i.e. increased rhetorical support from the state for worker grievances.


Insurgent Moment Institutional Moment



Shopfloor State

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (13)


winning compromise from the state and capital. But, in the case at hand, the ACFTU didnot mobilize or actively involve itself in the lives of migrant workers as they emerged asa new class,7 as it maintains the state’s obsession with reducing labor unrest. Underconditions of appropriated representation, dispersed worker insurgency strengthens thehand of union representatives at the national level (since the state fears instability andmay be willing to promote legislative reform), but simultaneously results in weak,illegitimate unions on the shopfloor which are generally incapable of enforcing laws andcollective agreements.

Such a scenario recalls the problem of union oligarchy, a line of inquiry firstestablished by Robert Michels (1962). I do not conceive of oligarchy8 primarily as aunion that fails to pursue the interests of membership (since the content of “interests” arealways the object of symbolic struggle), but rather in process-based terms. Unions aredemocratic rather than oligarchic to the extent that membership is actively engaged andmobilized in the determination of organizational ends of action and the pursuit of thoseends. In other words, do workers have a say in what the union will do? And, onceorganizational goals have been established, are workers involved in pursuing these goals?ACFTU unions are highly oligarchic, as they are subject to heteronymous control of theParty from the national to the district level9 (formalized in the organization’sconstitution), while remaining highly subordinate to capital at the firm level. And yet, therelationship between the ACFTU and migrant workers is quite distinct from earlierinstances of oligarchic unions. In the West (and in many other countries) the generaltendency was for unions to begin quite democratic and to ossify over time into increasingoligarchy. However, the ACFTU has been re-created in toto by the state once in 1948 andagin in 1978. It is this specific historical trajectory in the relationship between union,state, and working class which I refer to as appropriated representation, and which putsthe ACFTU in a different category than previous cases. In my conceptualization, unionscan be expected to behave “oligarchicly” under conditions of appropriated representation(i.e. they will respond to the wishes of state and capital and members will be excludedfrom practical activities). Whereas appropriated representation refers to the generalpolitical context of the state-union-worker relationship, oligarchy (as juxtaposed todemocracy) refers specifically to union organization and can be empircally observed inspecific cases.

The tension between the insurgent and institutional moments of thecountermovement can be best analyzed at the point where the state is now attempting toincorporate rebellious workers: the trade union structure. Rupture and Representation, thetitle of this project, refers to the relationship of appropriated representation betweenmigrant workers and their union representatives in the process of capitalist

7 I do not mean to suggest that migrant workers have become constituted as a class in thefullest sense of the term. If class formation is a multi-layered process, migrant workerscertainly do not engage in collective action as a class (Katznelson 1986), and thus cannotreally be considered a “class in reality.” (Bourdieu 1985:725)8 It is important to note that “oligarchy” is used throughout as an analytical and not anormative term. For a precise definition of oligarchy, see chapter 2.9 Or to the township level in non-urban areas.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (14)


industrialization. Here, “rupture” refers in fact to a double rupture: first, the rupturebetween union and working class, which although subject to evolving dynamics wasinstitutionalized shortly after the founding of the People’s Republic; and second, therupture between migrant workers and capital which has been intensifying since the early1990s in the process of capitalist industrialization. At the most abstract level then, I askthe question, what are the dynamics of representation within the context of the doublerupture? To put things a bit more concretely, I am interested in how the state, through theauspices of highly oligarchic unions, deals with the problem of labor conflict in the earlystages of capitalist industrialization and attempts to integrate workers into legalizedchannels of contention. This analysis requires an investigation not just into therelationship between workers and union (although this is the primary focus), but alsothese two groups’ respective relationship to state and capital. Although the China case isin many ways unique, it is of incredible importance for understanding the futuredynamics of global capitalism.

Why China is Different and Why it Matters

The study of labor politics during the process of capitalist industrialization haslong interested scholars in various disciplines. Early radical theorists such as Marx andLenin believed that the state could not resolve the contradictions inherent in capitalistdevelopment. As a result, their question was one of revolutionary strategy rather thaninstitutionalization of a countermovement (since they saw the capitalist state as unable toeffectively compromise). But for a group of mid-century neo-Durkheimian scholars, theexamples of the United States and Western European in the 1940s and 1950s caused themto argue that advanced capitalist states were fully able to reduce labor conflict andaccommodate the material interests of the working class.10 Key to this process was theconstruction of civil society populated by representative organizations that could expressthe interests of their membership – chief amongst which were trade unions. Unions in theWest forsook a revolutionary agenda in exchange for access to the state. But armed withofficial recognition as well as mobilizational and political power, labor movements invarious countries played a central role in advancing the decommodifying policiesembodied in the construction of the welfare state.

Without discounting great variation in political and economic outcomes of laborstruggles within the West (Rueschemeyer, Stephens, and Stephens 1992), labor politics inthe initial stage of capitalist development looked quite different in the late developers ofLatin America and the “late-late” developers of East Asia. At the risk of dramaticoversimplification, Latin American governments in the early 20th century used acombination of coercion and concessions to tame labor insurgency in attempting toestablish institutionalized, regularized, and co-optable official labor movements. Successin advancing the political and economic interests of the working class was uneven,

10 e.g. Kornhauser, William. 1959. "The Structure of Mass Society." in The Politics of

Mass Society. New York: Free Press.; Parsons, Talcott. 1964. "Evolutionary Universalsin Society." American Sociological Review 29:339-357.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (15)


though proletarian revolution was averted. The “Asian Tigers” of Taiwan, Korea,Singapore, and Hong Kong, on the other hand, maintained much greater politicalexclusion of labor (Deyo 1987; 1989). With the possible exception of Korea (Koo 2001),workers were much less militant than their counter-parts in the West or Latin Americahad been at a similar stage of industrialization.

Although labor politics during capitalist industrialization in China has manypoints of similarity with earlier cases, there are some very important differences as well.The two features of the institutional moment that I am interested in – decommodificationand incorporation – are based on existing concepts, both of which require modificationfor the Chinese context. For Polanyi, “decommodification” remains vague and hardly ananalytical term at all. But theorists of the welfare state, most notably Esping-Andersen(1990) have refined the concept in focusing on national policies which remove theprovision of basic human needs (e.g. health care, pensions, education, housing, etc.) fromthe market. The key difference between my conception of decommodification and that ofEsping-Andersen is that he analyzes national-level policies while I insist on determiningif potentially decommodifying policies/collective bargaining agreements areimplemented on the ground. This is because one of the primary obstacles todecommodification in China is the strong alliance between the lowest levels of the state(the district level in particular) and capital. Potentially pro-labor legislation and collectivecontracts often go un-enforced with the implicit or explicit approval of precisely thoseofficials that are supposed to enforce them. Thus it is necessary to enter the workplace tostudy decommodification in China.

As for incorporation, China once again calls for a re-configuration of the conceptas it has traditionally been used. Collier and Collier’s (1991) influential work conceivesof incorporation as follows: “State control of the working class ceased to be principallythe responsibility of the police or the army but rather was achieved at least in part throughthe legalization and institutionalization of a labor movement sanctioned and regulated bythe state.” (1991:3) In both this work and other studies of Europe and Latin America, thequestion is how states deal with incorporating unions which developed independently(and which often had highly developed political agendas). Chief amongst these unions’political demands were official recognition and collective bargaining rights.

But Chinese migrant workers are not demanding legal reforms. Indeed, on astrictly formal level, Chinese workers are already guaranteed freedom of association,collective bargaining rights, freedom of speech, etc., as well as relatively strong jobprotections. And they are not fighting for union recognition: the existing union structureis so deeply integrated into state as to render it thoroughly illegitimate amongst itssupposed constituency, and attempts to build organizations outside of the ACFTU areimmediately crushed. Part of the difficulty the state in China faces is that it iscategorically opposed to workers engaging in collective activity such as formulatingdemands, and yet it must respond to the eminently collective problem of generalizedresistance. Without legitimate representation – a means for co-optation – such aprocedure is incredibly difficult. Thus, while in earlier cases, states had to decidewhether/how to integrate worker representatives into the structures of the state, in Chinaincorporation is the state’s struggle to integrate atomized workers into the union, therebyrendering their struggles “intelligible” and potentially co-optable. If in an earlier era, thekey site of analysis was between unions (as relatively unproblematized representatives of

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (16)


workers) and the state, in China the problem of oligarchy means that the focus must shiftto the relationship between dispersed insurgent workers and unions. Certainly a politicalcrisis or democratization could very quickly change these dynamics (as it did in Koreaand Taiwan), but such a consideration cannot be further explored until something of aviable opposition exists within China.

If other East Asian countries maintained similar levels of repression againstindependent union organizing during capitalist industrialization, why then is Chinadifferent? First of all Chinese workers are much more rebellious than their counter-partsin Taiwan, Hong Kong, Singapore or post-war Japan (again, Korea is somewhatdifferent). While there are many factors behind this, greater resistance may be due in partto the wildly unequal distribution of wealth in the reform era, something whichdistinguishes it sharply from other countries in East Asia. Additionally, migrant workershave not benefited significantly from recent increases in social spending, most of whichhas been directed toward urban or rural residents while bypassing the in-betweens. In anyevent, corralling the huge number of strikes and other autonomous forms of action willlikely prove much more difficult a task in China than it was in other East Asian countries.

An even more fundamental difference owes to the differing effects of the spatialand social dynamics of capitalist expansion in China as compared to other smallercountries. While the Asian Tigers were able to shift relatively quickly away fromeconomic dependence on light manufacturing, no such rapid re-alignments will bepossible at the national level in China given its vast size. If in smaller late-industrializers,the “spatial fix” was able to relocate the contradictions of capitalist development abroad(Silver 2003), in China such a process will look quite different. That is not to say thatcapital cannot use spatial mobility as a method for undercutting labor militancy; indeed,there is currently a massive shift afoot in labor-intensive industries from the coastalregions to the interior. But given China’s enormous population and land area and stillquite low GDP per capita (just over US$4,200 in 2010),11 capital will continue topenetrate new geographic and social spheres of accumulation within the nation’s bordersfor quite some time. Since the nation-state continues to exert incredible influence on thedynamics of labor politics, there are significant implications when the social and spatialrelocations of capital occur within a given state rather than between them. To be morespecific, China will likely not be able to “wait it out” to move up the value chain (ashappened to the most extreme degree in Singapore and Hong Kong), especially given itsalready significantly higher levels of worker resistance. With a doubt, labor politics willlook different in various regions of the country (see chapter five), but any fundamentalchanges in worker-union-state relations (e.g. right to strike, recognition of independentunions, etc.) will be difficult to contain in neatly delineated geographic regions.

Finally, it is worth noting that labor politics in China hold profound consequencesfor the future of global capitalism. China occupies an increasingly central position in theglobal economy (Arrighi 2007; Hung 2009; Li 2009), and the nation’s political leadershave lofty geopolitical ambitions. China’s transition to capitalism has alreadyfundamentally reconfigured the structures of the global economy. It is the world’s largest


Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (17)


exporter,12 one of the top recipients of FDI,13 the second largest national economy, andincreasingly dominates the production of all sorts of goods, from the very low-end andlabor intensive to high-end and capital intensive. Given the high degree of concentrationof the globe’s manufacturing, a shift in the country’s mode of accumulation willreverberate internationally. Additionally, although China is of course “dependent” on themarkets of wealthy nations, it is not politically or militarily subordinate to the UnitedStates in the way that Japan, Korea, Taiwan, and any number of Latin Americancountries were and are. China will increasingly be in a position were it is less bound byexternal constraints than has been the case for many newly industrialized countries. Theconsequence is that, if pushed in a pro-labor direction by worker insurgency, China maybe in a position to lead a decommodfying restructuring of global capitalism (Arrighi2007). As unlikely as such a scenario seems as present, it is important to note thissignificant difference in comparison with other late developers.

Existing Literature on Labor in China

At present, there is significant literature on workers and worker protest in China,but as yet no comprehensive studies on how the state and union are responding. Thedestruction of the danwei system that had previously integrated urban workers into statestructures during state socialism (Walder 1983; 1984) resulted in a loss of direct controlover urban workers (Lau 2001; Solinger 1995) and there were massive revolts in the late1990s and early 2000s (Cai 2002; Hurst 2009). Although many of these laid-off workershave suffered immensely in the reform era and have had little success finding re-employment in the private sector, municipal governments have greatly expanded socialinsurance coverage for them (Frazier 2010). But migrant workers, from the verybeginning existing in a precarious economic position and with ambiguous legal statusonce in the city, have emerged as a new social class without an institutionalized channelfor integration of collective demands into legalized state mechanisms. The first wave ofscholarship on migrants identified their legal and economic precarity (Solinger 1999) andthe frequently brutal employment conditions they have been subjected to (Chan 2001;Choi 2003). Subsequent studies have focused on the volume and character of workerresistance. It is certain that Seidman’s findings on worker movements in Brazil and SouthAfrica are quite different from contemporary China, as she argued, “state-led,authoritarian industrialization strategies in late industrializers may tend to producemilitant working-class movement whose demands go well beyond the factory gates.”(1994:12) But if migrant workers are not explicitly political in their demands, there is animportant debate in the field focused on the question of class formation and subjectivity.Lee has a relatively pessimistic perspective, arguing that legal reforms have given rise toa highly legalistic mode of resistance (Lee 2007), and that the state’s project ofindividualizing labor conflict has been actively supported by unions and NGOs alike(Friedman and Lee 2010). She argues that worker resistance in China is characterized by“cellular activism,” in which insurgents are unable to construct durable organization or

12 18, 2011)13 (accessed April 18, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (18)


articulate political demands. On the other hand, Pun Ngai maintains that the category ofdagongmei/zai (working girl/boy) represents a potentially subversive discursiveformation, one which could serve as the symbolic foundation for more broad-basedmobilization. And indeed, she has found evidence of strikes spreading beyond singlefactories, a phenomenon they take as indication of heightened worker consciousness(Chan and Pun 2009; Leung and Pun 2009). Despite his relative optimism about classformation in China, Chris Chan’s key phrase of “class struggle without classorganization” (2010:16) is an implicit recognition of current limits. Regardless of suchdifferent interpretations of working class subjectivity, there is consensus that capitalistdevelopment generated resistance that has been rapidly expanding in scope since theearly 1990s, and that this represents a major political challenge for the regime.

My primary aim, however, is not to describe the dynamics of worker resistance,but rather to provide an analysis of how the state, through the auspices of the unions, isresponding to this conflict. Ching Kwan Lee captures one aspect of the state’s response,namely the expansion in legal rights for workers (2007).14 As has been noted, thisresponse – which culminated in the 2007 passage of the Labor Contract Law – is anattempt to integrate workers into the structure of the state as atomized individuals. In thissense we can see strong parallels with Koo’s characterization of the relationship betweenstate and worker in other export-oriented economies in East Asia.15 But as argued byFeng Chen (2007), the extension of individual rights in the absence of collective rightshas failed to reduce labor conflict. The focus of this study is then how the union respondsto generalized worker insurgency by attempting – within given political parameters – toincorporate workers and potentially advance decommodification.


Gaining access to Chinese unions is a significant challenge, and studying howunions respond to the crisis of worker resistance only compounds the problem. However,I had the good fortune to serve as an interpreter for prominent American labor leaders onseveral exchanges held with national leaders from the ACFTU as well as the ShanghaiFederation of Trade Unions. Given this friendly introduction, I was able to meet people,conduct interviews, and gain access to information that I would not have otherwise.Additionally, while serving as a lecturer at Sun Yat-sen University in Guangzhou, mymother developed a close working relationship with the chairman of the GuangzhouFederation of Trade Unions, Chen Weiguang. Although certain activities of the unionsremained highly opaque, the access I did secure would not have been possible withoutthese personal connections.

14 Lee is not the only one to have written about “rule of law” and labor politics in China,however she has produced the most important analysis in terms of the consequences forcollective action. For other works on labor law see (Chan 2009; Chen 2004; Chen 2009;2006; Cooney, Biddulph, Li, and Zhu 2007; Keith 1991; 1994; Wang, Appelbaum,Degiuli, and Lichtenstein 2009)15 Caraway argues that individual labor rights are a strong suit of East Asian countries,and that protection of such rights in the region is somewhat stronger than the worldaverage (2009:156).

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (19)


This project depends on ethnographic data collected during one and a half yearsof fieldwork in China. I spent approximately ten months in Guangzhou, with theremaining time divided up between Zhejiang province, Shanghai, Hong Kong, andBeijing (in descending order). My respondents include union and government officials,workers, and enterprise managers. As anyone who has conducted qualitative research onsensitive topics in China will attest, such work requires a high degree of flexibility inapproaches to data collection. Sometimes I was able to conduct formal semi-structuredinterviews; other times, I would not be afforded an interview, but would be allowed to“chat.” I attended numerous formal meetings between foreign and Chinese unionofficials, and an equal number of formal meals. These meetings mostly took place duringfour separate multi-day trips by Chinese and American delegations in 2007 and 2008, twoeach in the U.S. and China. As I befriended some of the younger staff of various unions, Iwould sometimes meet up with them for lunch or tea. With workers and managers, mymethods were similarly diverse. Sometimes I would have several hours to conduct aformal interview, or perhaps I would chat with a worker while playing a game of pool.Additionally, much of the data in chapter six comes from supervised interviewsconducted by research assistants.

In addition to interview data, I rely on media reports and historical documents.Although the media’s ability to report on labor issues in China is constrained, it is notuncommon for strikes or other labor conflicts to be reported. In several instances, I foundout about some incident in the media and then would conduct follow up interviews withworkers to gain a greater depth of understanding about the case. The historical data inChapter three come largely from collections of official documents stored at theUniversities Service Centre at the Chinese University of Hong Kong.

Since the working hypothesis is that conditions of appropriated representationserve as a significant impediment to decommodification and incorporation, the studyfocuses on several “most-likely” cases. By analyzing the most progressive attempts ofChinese unions to advance the interests of their membership, I aim to identify theconditions under which the countermovement can be institutionalized. These cases,exceptional though they may be, reveal underlying dynamics which hold implications forthe entirety of the trade union system in China. I look at how three general factors impactpossibilities for institutionalization of the countermovement: 1) internal organizationalfactors (pro-labor leadership); 2) external economic factors (model of development andshifts in global economic conditions); 3) external political factors (center-local relationsand specific incidents of worker insurgency).

Finally, I would like to note how important data triangulation was in this project. Idecided on studying the ACFTU at a time when many in the West were under theimpression that Chinese unions were in a period of significant reform. Based on what Ihad heard from many Chinese union officials during exchanges in 2007, I began myresearch expecting to find unions that were playing a key role in advancingdecommodification. However, I found time and time again that union officials’ claimsabout what was happening in the workplace were not verified by interviews with workersand managers. Thus, I frequently employ a method of contrasting union officials’representation of a particular reality with the lived experience of workers and managers.This is not intended as vindictive attempt to reveal the deceitfulness and/or simpleignorance of union officials; rather, it allows us to understand what sort of a model of

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (20)


labor relations the unions want, and how structural and organizational impedimentsfrequently prevent them from realizing such ideals.

The Argument

I began this chapter noting that over the first eight years of the Hu-Wenadministration, at precisely the time that the central government appeared to be moving inthe direction of class compromise, worker insurgency continued to rapidly expand. Invery broad terms, my argument is that the 2000s were characterized by the statebecoming hemmed in by an “insurgency trap,” a condition that persists up until thepresent. As marketization generated an insurgent response from the new working class,certain segments of the state became interested in institutionalizing thecountermovement. They sought to incorporate workers as atomized individuals throughlegal reforms, a project that failed to reduce labor conflict. There was a real increase insocial spending, but most programs were directed either at urban or rural residents, whileexcluding migrant workers. Despite conditions of appropriated representation, unionactivities hinted at a collective response, but such efforts have encountered difficultybecause of ongoing oligarchy. This combined with a categorical ban on the developmentof alternative autonomous organizations and collective power for the working classmeans there is little capacity to coerce capital into abiding by legal and contractualobligations. This difficulty is compounded by the strong alliance between capital and thelowest levels of the state, those responsible for legal enforcement. The one method likelyto reduce conflict – a countervailing force at the point of production – remains off thetable as far as the central state is concerned; hence the insurgency trap. In light of this, Iargue that we need to distinguish between the institutional and insurgent moments ofcountermovements against commodification. Merely analyzing potentiallydecommodifying legislation is insufficient in cases were non-enforcement of laws is anintegral and necessary feature of the model of development (as is the case in China).Despite indications of an institutional response to insurgency in China, the state andunion’s attempts to improve conditions for workers without allowing them to gain adegree of organized and autonomous power have by and large failed up to the present. Toput it somewhat provocatively, it is precisely the state’s obsession with stability andharmony that is ensuring continual unrest and discord. Thus, despite the legislative effortsof certain segments of the state and top-down administered collective bargaining, Chinaremains stalled at the insurgent moment of the countermovement.

I argue that institutionalization of the countermovement requires both economicand political change. In the economic sphere, I look for decommodification of labor,while in the political it is incorporation of the working class as a collectivity. In a strikingconfirmation of Polanyian theory, we see that even the ACFTU, despite being anexemplar of rigid oligarchy, is attempting to promote decommodification and gainrecognition from its constituency in response to worker insurgency. I analyze how theunion negotiates the tension between the impetus to respond to intensifying workerresistance, on the one hand, and profound structural oligarchy and heteronomy on theother. How then is the union attempting to ameliorate labor conflict given existinginstitutional parameters? Under what conditions is the union relatively successful? Whatsorts of internal organizational changes are taking place? Given the failure of legal

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (21)


reforms to effectively incorporate workers into the state as individuals, can the unioneffectively guide rebellious workers into rationalized legal channels? And can suchlegalized mechanisms resolve conflict? Although in general Chinese unions have notfigured out a way to decommodify and incorporate labor, there are some cases wherethey have been relatively successful – particularly in establishing the formal parametersfor collective bargaining which hold potential for advancing these goals. Even if suchcases remain exceptional, the processes that produced a degree of institutionalization areworthy of investigation.

Before moving to an analysis of contemporary labor politics, it is important tounderstand the historical evolution of the ACFTU in order to grasp how appropriatedrepresentation was established in China. The basic argument (spelled out in chapter three)is that the ACFTU is not simply a union that is deeply integrated into state structures andsevered from its membership (like many other unions around the world). In fact, giventhat the union was founded in the context of semi-colonialism, it has always maintainedthe inter-related goals of realizing ethno-national autonomy and increasing productivecapacity. What’s more, the union completely collapsed and then was resurrected by theParty (on the Party’s terms) on two separate occasions, once in 1948 and once in 1978.These two incidents in which the Party unilaterally arrogated exclusive rights ofrepresentation to an organization of its own creation are the key events in establishingconditions of appropriated representation. Although there have been a number of attemptsfrom ACFTU leadership for slightly greater operational autonomy – if not outrightindependence – each of these efforts has failed. When the state was committed to aprogram of labor decommodification during the era of state socialism, active workerrepresentation such as exists in liberal capitalist democracies was arguably not as crucial.And indeed, conflict between workers and managers during the Mao era was relativelylow.16 However, with marketization and the transition to capitalism acceleratingthroughout the 1980s and 1990s, interests between state and working class havecontinued to diverge (albeit not in a neat or linear manner). Thus, the union’s profounddis-embedding from its constituency has become a pressing political problem for thestate.

When we move to the contemporary era, I have analyzed the manner in whichthree different factors influence the capacity for the union to play a role ininstitutionalizing the countermovement: 1) pro-labor leadership at the municipal level; 2)regional model of development; and 3) center-local relations and the evolving dynamicsof worker resistance. Respectively these represent internal organizational, externaleconomic, and external political factors. There are reasons to believe that under particularconditions, even highly oligarchic ACFTU-subordinate unions could play a role indecommodification which would likely lead to increased recognition from membership.Determining the conditions under which such an event is possible will be crucial inanalyzing the emergence of the institutional moment of the countermovement.

Union Leadership

16 There was of course significant worker mobilization during the Cultural Revolution.However, given the unique character of this protest, I consider it qualitatively differentfrom more typical conflicts between workers and managers.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (22)


The question with regards to union leadership is this: can progressive leadershipat the municipal level advance decommodification and incorporation of labor? Thedebate over the efficacy of union leadership parallels the interminable structure-agencydebate, i.e. can the action of one actor, or one small group of actors, overcome structuralconstraints in realizing desired outcomes? It is frequently suggested by unionpractitioners (and some scholars) in China that leadership can make a very largedifference in outcomes, since laws and regulations are often fuzzy and hence subject tointerpretation in implementation. Since the working hypothesis has been that ACFTU-subordinate unions have not played a decisive role in actively bringing aboutdecommodification and incorporation, I began the study by investigating the places most

likely to provide some counter-evidence. Thus, we begin with the Guangzhou Federationof Trade Unions (GZFTU), chaired by Chen Weiguang – the man widely considered byscholars and practitioners alike to be the most progressive and consistently pro-laborunion leader in China.

The findings on the effects of leadership are inconclusive, though there issignificant evidence that progressive leadership is in general not able to overcomestructural oligarchy. In the “model trade union” enterprises, regular workers do in factenjoy relatively good pay and benefits, and have comparatively good job security. Buteven such enterprises are characterized by a dual labor market, with a significantpercentage of the workforce composed of flexible “intern” labor, something the union hasdone nothing to address. What’s more, it appears as if the relatively good conditions forworkers in model enterprises come about as a result of state power, not from the activityof the union qua worker representative. The result is that the union has done nothing toincrease its legitimacy among rank and file, one consequence of which is ongoing wildcatstrikes.

Based on research in Guangzhou, I additionally argue that a “passive repressive”position is a key strategy of Chinese unions – even those with seemingly sympatheticleadership such as the GZFTU – in quashing bottom-up activism. By passive repression Imean to indicate instances in which union leadership refuses to intervene on behalf ofa*ggrieved workers; simply by “doing nothing” (acting passively) at decisive moments,the inherent repressive capacity of capital to hire and fire is unleashed. This strategy freesunion higher ups from having to actively repress workers, thereby allowing them to stayrelatively above the fray (although we will see that union officialdom sometimes doesend up getting their hands dirty). In the case of the Guangdong Union Hotel we seestrong structural constraints on leadership supporting bottom-up initiatives, and thesurprising (particularly in this case) retreat to passive repression.

In sum, while union leadership at the municipal level is able to pass pro-workerlegislation and to establish a relatively favorable discursive environment, structuraloligarchy has prevented them from constructing durable worker power on the shopfloor.The consequence is that decommodifying activities continue to be highly circ*mscribedand tenuous, and unions remain illegitimate in the eyes of membership.

Regional Model of Development

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (23)


In the course of my research, I continually heard from union leadership inGuangdong as well as in Beijing that some of the most successful unions in the countrywere located in Zhejiang province, just to the south of Shanghai. They tended to attributethis success to the leadership qualities of local union officialdom, as well as support fromthe government. But based on a comparison of sectoral unions in southeast Zhejiang andGuangzhou, I argue that regional models of economic development are crucial indetermining varying capacity for organizational innovation within the structure of theACFTU. Additionally, I present evidence that the union federation’s attempt toadministrate improved conditions while keeping workers atomized will encounter severeobstacles.

The ACFTU has in recent years been heavily promoting the development ofsectoral unions – organizations which attempt to establish common standards for allemployers in a particular industry within a given municipality – as a primary strategy forreducing unrest. National leadership is particularly keen to establish such organizationsand engage in collective bargaining at the sectoral level because they believe that it is away to improve things for workers without serious reform of union structures or givingthe rank and file any power, what I term “oligarchic decommodification.” Municipalitiesin southeast Zhejiang have been relatively successful in establishing such unions, whichis in large part due to the fact that there is a high degree of concentration of a particularindustry in single townships, employers are overwhelmingly local, firms size is small,and capital is organized into legitimate employers associations. But such conditions arequite unique in China, as is evident when we look at another center of capitalistdevelopment, Guangzhou. In Guangzhou, GZFTU leadership has also been intent onestablishing sectoral unions, but thus far has had very little successes. In strong contrastto Zhejiang, Guangzhou’s model of development is characterized by high levels ofmobile foreign investment, highly diverse manufacturing and services, large firms, and anabsence of legitimate employer associations. The consequence has been that inGuangzhou there has been no attempt to establish sectoral unions in manufacturing whileexperiments in the construction industry and the service sector have proven largelyineffective.

Based on studies of the wool knitwear and eyeglass industry, we see that theparticular model of development in Zhejiang has allowed unions there to engage insectoral-level collective bargaining, something that has not been possible in most placesin China. This is an excellent example of highly oligarchic unions attempting toinstitutionalize decommodification within given parameters – most significantly, byadhering to the categorical ban on developing grassroots collective power. And yet, adeeper investigation reveals that the breakthrough of sectoral level collective bargainingis only a partial success. Because of the crisis of legitimacy within the union, workerswere completely unaware of the contract supposedly regulating their employmentrelationship, the consequence of which is that the contract is not being implemented, andunrest continues unabated. Thus it is possible to see that possible decommodification isundermined by lack of incorporation within the Chinese political context.

Political Economy and the Dynamics of Worker Resistance

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (24)


Resistance among Chinese migrant workers has by and large not becomeexplicitly political, with demands generally related only to immediate economic issues.And yet, the accumulation of seemingly particular and unrelated incidents of insurgencyare, at the aggregate level, deeply political. While in general, this study treats laborconflict as an ever-present but relatively abstract threat to capital accumulation, here Ianalyze two specific strikes in similar industries and in the same region, but separated bytwo and a half years. These strikes serve as prisms through which to analyze the rapidlychanging political and economic conditions in the process of capitalist industrialization,touching on the multi-layered and dynamic relationships between workers, capital,unions, and the state from the enterprise level all the way up to Beijing.

Between the first strike at Otis Elevator which occurred in late 2007, and thestrike wave of spring-summer 2010 sparked by workers at Nanhai Honda, major politicaland economic events had taken place. In particular, high-level officials in Beijing and inthe Guangdong provincial government began to talk more forcefully about moving up thevalue chain and expanding domestic consumption (in no small part due to continualeconomic stagnation in the West). Of course this shift in strategy was in part due to thehigh levels of labor conflict the country had been experiencing for years, and which hadintensified remarkably in 2008-9. Regardless, when the strike initially broke out atNanhai Honda, it had implicit support from the central government, as evidenced by theongoing and widespread media coverage. This opened up the space to allow workers toincrease their organizational strength, formulate a range of demands (including somerelated to union organization), and construct internal unity. As a result, the strike gainedin strength and began to exact a severe economic toll on the company, something whichwas not possible to the same extent at Otis. The greater organization and militancy of theHonda strikers allowed them to make gains that never materialized in the earlier strike.One can make a superficial assessment of this comparison and simply argue that strikesthat are more militant and disruptive are more likely to succeed.17 I do not doubt that thisis the case. However, in order to understand why workers are sometimes in a position toexpand organizational capacity, it is important to grasp changes in underlying politicaland economic conditions, particularly as relates to tensions between the various layers ofthe state (and union).

The fallout from the 2010 strike wave caused serious reflection among unionofficials in China. Most significantly, some leaders began to make public calls forlegalization of strikes, particularly in Guangdong, but in other regions of China as well.18

Continual disruption had pushed the state further in the direction of accepting some legaland organized power for workers, even if in the specific case of Nanhai Hondamanagement and higher levels of the union continued to collude to rig elections.Certainly, the movement towards legalizing certain types of strikes is also an attempt toformally delegitimize other strikes, and in this sense the laws under consideration are not

17 Cai Yongshun has argued that incidents of collective action which the governmentcannot pretend to be ignorant of hold a higher likelihood of success (2010:15)18 For instance, in Dalian where a massive strike wave involving 70,000 workers sweptthrough a development zone: September 20, 2010. “dalian tinggong chao 7 wan ren

canyu boji 73 jia qiye, yi gongzi zhang 34.5% gaozhong” [Dalian strike wave of 70kworkers affects 73 enterprises, results in 34.5% wage increases] Caixin.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (25)


necessarily a victory for workers. But it is also a clear example of the insurgent-institutional dialectic at work. Even if unions continued to fail at incorporating workers inthe aftermath of the strike wave, we can see the pressure generated by worker resistanceis forcing continual adjustments, fixes, and occasional compromises. Such a dynamic willcontinue to be an ongoing feature of China’s process of capitalist industrialization.

Finally, given that the withering of civil society is a global phenomenon (Hardt1995), the study of how insurgent workers relate to oligarchic unions is of broaderrelevance. In countries around the world, from authoritarian states like China andVietnam, to liberal democracies such as the U.S. and U.K., union federations that wereintegrated into the state during the 20th century have been unable to effectively respond tothe challenge of neoliberal capitalism. Although the specific dynamics are of course quitedifferent in each national setting, there are some common general characteristics. Unionshave by and large exchanged their capacity to mobilize and disrupt capital accumulationfor access to state power. However, when states around the globe abandoned nationally-oriented growth patterns based on a broad-based increase in wages and social welfare,unions were largely unable to resist. At the same time, neoliberal globalization hasundercut their base of support, either through privatization of state firms or outsourcingof unionized private-sector industries (or both). Capital has successfully introduced muchhigh levels of precarity and flexibility into labor markets, with migrant/immigrant andfemale labor increasingly replacing the regularized employment in the male-dominatedunionized sector. Given high levels of institutional inertia and oligarchy within traditionalunions, it is likely that struggles for decommodification will continue to be organizedautonomously around the globe, and frequently in opposition to established unionbureaucracies. Nowhere was such a dynamic clearer than in the autonomously organizedstrikes during the Egyptian revolution, where a new independent labor federation wasestablished as the Mubarak regime crumbled. Shortly thereafter, demonstrations at thestate-run Egyptian Trade Union Federation headquarters demanded the dissolution of theold union and that its leaders be put on trial, as their obsequiousness to the regime hadprevented them from attempting to protect the working class from privatization andstagnant wages.19 It is not difficult to imagine such a sequence of events playing outelsewhere, particularly in non-democratic regimes.

19 (accessed February 22, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (26)


Chapter 2


In this chapter I develop in greater detail the conceptual tools that will bedeployed in the subsequent empirical sections. I aim to establish a framework forunderstanding the dynamics of labor politics in an authoritarian country undergoingcapitalist industrialization. This framework is then be deployed in demonstrating how theinstitutionalization of a Polanyian countermovement has encountered severe obstaclesdue to conditions of appropriated representation for the new migrant working class inChina. In particular, this requires specifications of the countermovement, oligarchy,insurgency, and the structure of the state. Through the development of the concept of an“insurgency trap,” I will show that the state’s faltering attempts at reaching classcompromise have engendered an increasingly “unstable equilibrium.”

Reformulating the Countermovement

In The Great Transformation (1944), Polanyi provides a sweeping historicalnarrative of the construction of market society in England in the 18th and 19th centuries.With the rise of exchange during the industrial revolution, the form of commercialexchange and its attendant institution of the market began displacing the sociallyembedded economy. Polanyi sees a tension between the socially embedded economyover which human agents have voice and control, and the self-regulating market which issomething alien and indifferent to human and ecological needs. The ideological victoryof market liberalism resulted in a disembedding of the economy, a process which leftsociety at the whim of the market, that unpredictable beast that answers only to the callprofit.

Key to this process was the emergence of the three fictitious commodities, land,labor, and money. What about these things is fictitious? Fundamental to the functioningof a market economy is the production of things as commodities, that is to say, for sale onthe market. Despite the self-regulating market’s insistence on treating land, labor, andmoney as commodities, they are not produced as such, so they are therefore fictitious

commodities. However, these three elements are all essential to the functioning ofindustry, and therefore had to be for sale. The commodification of labor and land wasrealized in England through the enclosures which privatized land and displaced peasants,and the subsequent abolishment of the Poor Laws, which forced this same class to securewages or starve. The implementation of the gold standard ensured that governmentswould not interfere with the money supply, thereby allowing for commodification ofmoney.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (27)


As is vividly portrayed in his depiction of the industrial revolution in England, thecommodification of these three things caused great damage to society. However, Polanyisees in society an inherent drive for self-protection, which is distilled in thecountermovement against the rise of the self-regulating market. Thus, we see the centralconcept of the Polanyian paradigm: the double movement. On the one hand, liberalsfought for the violent expansion of the self-regulating market throughout Europe andbeyond (movement 1); on the other, a wide array of sectors in society rose up to resist thedestructive impulses of this market (movement 2). To skip ahead slightly, the outcome ofthis painful process of industrialization and the rise of capitalism has been the discoveryof “society.” Knowledge of “society” follows the other two constitutive features of theWestern consciousness, namely knowledge of death, and knowledge of freedom. Oncethe knowledge of society has been attained, the choice is between fascism and socialism:“While the fascist resigns himself to relinquishing freedom and glorifies power which isthe reality of society, the socialist resigns himself to that reality and upholds the claim tofreedom, in spite of it.” (1944:268)

Unfortunately, Polanyi’s delineation of the double movement suffers fromvagueness and a mechanistic logic. In Polanyi’s conceptualization of the problem,society’s revolt against the self-regulating market is born of necessity, since otherwise theworld would be destroyed. Aside from its thinly veiled functionalism, this theory fails toexplain in sufficient detail the process by which a revolt of society leads to a “re-embedding” of the market and subsequent decommodification of the three fictitiouscommodities. Additionally, he fails to adequately specify a theory of the capitalist state or“society.” To put it another way, Polanyi has a weak theory of politics.

In order to get past these difficulties in Polanyian theory, a few conceptualadaptations are in order. First, the concept of “embededness,” so central in The Great

Transformation, needs to be clarified. I support Block’s argument that Polanyi in factdiscovered the concept of the “always-embedded market” without being able to name it(2003). Block emphasizes Polanyi’s argument about how “laissez faire” was planned, andthat politics are always crucial in determining economic arrangements. This always-embedded perspective is crucial because it directs our attention towards struggles overthe manner in which markets are embedded (since they cannot be dis-embedded). This ofcourse causes politics to take center stage. Rather than a Marxist-influenced inevitablependular swing from dis-embedding back towards embedding (Silver and Arrighi 2003),we have a much messier, much more contingent view of the countermovement.

The insurgent and institutional moments of the countermovement

The reading of Polanyi that I have argued for views the countermovement not as amovement for “re-embedding,” since the market cannot ever be dis-embedded, but ratheras a movement for political incorporation and resisting the commodification of land,labor, and money. Of course, such a movement implies that there will be a shift in thetypes and degrees of social embedding of the market, but the distinctive feature of suchmovements is not that they seek to embed the market, but rather that they opposecommodification. The point that I would really like to emphasize in this section is thatresistance to the incursion of the market does not, in any straightforward or necessaryway, result in decommodified labor. In doing so I will specify distinct moments of the

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (28)


double movement: movement 1 (commodification), and the “insurgent” and“institutional” moments of movement 2 (the countermovement).

Movement 1 is rather straightforward and does not in my view requirereformulation. Polanyi predicted that increased commodification of land, labor, andmoney would lead to social and ecological chaos. The extreme commodification of laborin post-Mao China (less true for land and money) and the severe social disruptions it hasincurred are testaments to the veracity of this first step in the Polanyian theory. However,things become a bit murkier when we turn to movement 2. I would like to argue that theinsurgent moment is the relatively disorganized, cellular movement of individuals andcollectivities in reaction to the social dislocation brought about by market incursions. Butthis sort of activity is, in and of itself, far from a guarantee of controlling the market. It isnot until the institutional moment that society subverts the naturalist logic of the freemarket to conscious human control that labor can come to be decommodified. Therelationship between the insurgent and institutional moments then is the relationshipbetween the moments of disruption and the rationalized and organized responses of thestate.20 The emergence (or not) of the institutional moment is a political question, ascertain groups will seek to maintain existing levels of commodification.

It should be emphasized that there is not a linear or teleological relationshipbetween the insurgent and institutional moments. While it is almost certain thatinstitutionalized decommodification will not precede the initial outbreak of insurgency, Ido not see a neat causal link between the two. This results in part from the fact thatinsurgency tends to be characterized by immediate economic demands and a politics ofnegation (i.e. protestors frequently do not have a fully-articulated positive vision of thesociety they want) which more conservative elements such as unions and states may thenattempt to co-opt through ameliorative measures. But since insurgency is not organized ina way such that coherent demands are made in the language and logic of the state, therecannot be – almost by definition – a mechanistic relationship between the two. Rather itis likely to appear as an iterative process between insurgents exacting certain symbolicand material costs and the imposition of institutionalized decommodification.

The focus of this study is the institutional moment of the countermovement inChina. The successful institutionalization of the countermovement appears in both thepolitical and the economic fields, in the political as incorporation of labor, and in theeconomic as decommodification of labor. As with the insurgent and institutionalmoments, incorporation and decommodification are dialectically related.

The commodity character of labor

In many ways, Polanyi’s work draws on Marx’s theory of primitive accumulation(1977:873-940). Polanyi’s description of the formation of the labor market in nineteenthcentury England follows Marx’s theory that the process of primitive accumulation wasdirected toward the production of “doubly free” labor: free from the means of production,

20 Of course, the possibility of revolutionary outcomes should not be fully discounted.

Here I generally assume that social revolution is not on the immediate horizon in China

and that an expansion of the social welfare state is a more likely outcome. But social

revolution could obviously also imply a more thorough decommodification of labor.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (29)


and juridically free to be bought and sold on the market. The old system of agriculturalproduction which was constrained by social norms had been destroyed; in its place waserected a new system in which workers were compelled to sell their labor power on themarket in order to meet their needs. Polanyi believed that labor is not actually acommodity, since it is not produced as such. The attempt to treat this “fictitiouscommodity” as if it were a commodity leads to social dislocation and a reactive countermovement from society.

I would like to approach the question of the commodity character of labor in asomewhat more systematic manner than Polanyi manages. Within the broad context of acapitalist society, labor cannot be seen as commodified or not (it always is to somedegree), but there are rather degrees of commodification along a spectrum. The simplestway to define my conceptualization of the commodity character of labor is the degree to

which a worker is immediately compelled to subject the satisfaction of their needs to the

logic of the market. Special attention should be paid to the phrase “satisfaction of theirneeds.” I specifically chose this formulation over “in order to survive,” in order toemphasize the social rather than biological determination of needs. Finally, degree ofcommodification can be measured on three axes: 1) Social protection; 2) Workplacesecurity; and 3) Participation in production.

By “social protection” I mean to indicate provision of things such as health care,pensions, education, and housing, which do not relate in an immediate way to theexchange of labor power for wages. By taking concerns for health, housing, retirement, inshort, general welfare, “out of the market,” workers’ basic livelihood is better ensured.The social protection measurement is particularly important, because it allows for thecommodity character of labor to serve as a sort of index for overall commodification ofhuman needs. Social protection is enhanced as the spatial and social scope of public goodprovision expands (i.e. national provision of public goods is superior to sectoral-basedprovision, which in turn is superior to workplace based provision, etc.). “Workplacesecurity” refers to things that impinge directly upon the content of the employmentrelationship. Things such as base wages, wage increases and seniority, hourly as opposedto piece rates, tenure, etc., can enhance the employment and overall financial security ofworkers, and therefore reduce the ability of market forces to utterly subjugate wageearners. Finally, there is “participation in production,”21 perhaps the most significant (andunlikely to occur) method for advancing decommodification of labor. Participation inproduction indicates the ability of workers to be active participants in the determinationof the structure of the labor process. This means that workers would have greaterinvolvement in decision making processes related to the organization of production,sales, work schedules, workplace rules, managerial compensation, etc. Depending on thedegree to which participation in production is realized, this can result in the mostfundamental decommodification of labor because it holds the potential of eliminating thewage relationship altogether.

21 Joel Andreas has argued that increases in what I would call “workplace security” ends

up reinforcing the capacity of workers to participate in production, since the fear of

losing one’s subsistence is lessened. See:


Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (30)


With this conceptualization we can jettison the vague and mechanistic vision ofsociety moving to re-embed the market, and instead empirically study specific strugglesover the commodity character of labor. While I do not attempt to quantify levels ofcommodification,22 this framework allows for us to see which direction things aremoving in (i.e. in particular enterprises, sectors, regions, etc., is labor becoming morecommodified, less commodified, or staying the same?). We now turn to a discussion ofthe institutional moment of the countermovement in the political field.


The political problem of incorporating labor has been faced by every stateundergoing the process of capitalist industrialization, although it has been dealt with inremarkably different ways. At the most general level, incorporation implies that laborconflicts are dealt with primarily through rationalized and legal means, which are carriedout by officially sanctioned representatives (unions). This of course implies a degree ofcompromise on the part of capital, but one which may become necessary to avoidongoing strikes and other forms of social upheaval. Incorporation is manifested at twolevels: 1) On the shopfloor, through the construction of (relatively) legitimaterepresentative organizations that can exercise collective power on the part of workers; 2)Within the state, where worker representative organizations can push for pro-laborlegislation. Indicators of incorporation include: worker recognition of the legitimacy ofunions; the ability to resolve grievances through collective bargaining, legal strikes, andother rationalized modes of contention rather than through spontaneous disruption suchas wildcat strikes and riots; unions actively contending with capital within the state forpro-labor legislation and social programs. Shopfloor representation and representationwithin the state are inter-related phenomena. To the extent that unions are able to winrecognition from workers, their political power within the state may advance. And to theextent that unions can win pro-labor legislation (and reasonably claim responsibility forthe victory), it may enhance their legitimacy among workers.

As has been mentioned, the problem of incorporation in China is quite differentthan in earlier industrializing countries in the West and Latin America. Whereas in earliercases, the problem faced by industrializing states was whether/how to integrate workerorganizations that had developed autonomously (and which were relatively legitimateamongst workers), in China unions are entirely a state creation, and are in this sensealready formally incorporated. The problem in China is that these unions did not resist,and indeed frequently encouraged, a radical program of commodification for most of thereform era. As a result, they are thoroughly illegitimate amongst workers. Thus, in Chinaincorporation does not mean bringing independent unions into the auspices of the state,but corralling cellularized but increasingly rebellious workers into the union. Bothincorporation and decommodification have faced acute challenges because of the

22 Attempting to establish a relatively objective quantitative measure for levels of

commodification of labor may be a worthwhile endeavor. However, it is not necessary

for my research since I am simply concerned with whether labor is becoming less

commodified or not.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (31)


problem of conditions of appropriated representation and the resultant high levels ofunion oligarchy.

Labor, Appropriated Representation and Union Oligarchy

The term "appropriated representation" was first used by Max Weber, although ina wholly unsatisfactory manner, as he defined it thus: "In [the case of appropriaterepresentation] the chief or a member of the administrative staff hold appropriated rightsof representation." (Weber, 1978:291) Since the phrase "appropriation" capturessomething basic about the representative relationship between migrant workers, unions,and the state in China, I have adopted and reconfigured it. In this work it refers to asituation in which the state unilaterally grants exclusive rights of political representationof an entire class to a particular organization in the absence of substantive or formalisticdelegation from membership. Such a form of representation has a number of distinctivecharacteristics: 1) If representative rights to an entire class are going to be arrogated to anorganizational body, this body must have already been constituted prior to the emergenceof the class; 2) The claim to representation is not dependent on formal membership butrather is encompassing of the class (even if the category of "member" exists); 3) Since therepresentative claim is encompassing of the class, the repressive capacity of the statemust be sufficiently developed to ensure that competitors do not emerge.

Appropriated representation may recall earlier definitions of corporatism, or morespecifically "state corporatism." (Schmitter, 1974:103) Indeed, the corporatist frameworkhas been popular in describing labor politics in China (Unger and Chan 1995; Chan 1993)with some even arguing that China has shifted towards "social representation" (Zhang1997) – an argument that has been marginalized by the empirical fact of continualpolitical exclusion of workers. But why is "appropriated representation" a moreappropriate concept for China's contemporary labor politics than corporatism (or one ofits many variants)? The first reason is quite simply that corporatism implies that thegroup in question has been more or less incorporated into state structures. But as hasbeen mentioned, the state’s attempts to incorporate workers have been failing, hence theexpanding insurgency. A primary purpose of this study is to determine if and under whatparticular conditions it may be possible to incorporate workers, but this has certainly notyet happened at the class level. Second, and relatedly, corporatism (for labor) refers to anhistorically specific arrangement in which the working class was forced to abandonpolitical goals in exchange for economic benefits. Whether in the fascist, state socialist,or Fordist welfare state variant, corporatism implied that the state would preside over aregime of relatively decommodified labor (as well as frequently providing representativeswith material and symbolic benefits) in exchange for acquiescence to national interests.23

The Chinese Communist Party has not been nearly so benevolent towards migrantworkers, which means that the new working class is not dependent on the state in the waythat workers in many Latin American countries were (Cohen 1982). Finally, corporatism

23 States demanded the submission of the immediate needs of the working class to

national interests particularly in state socialist (see chapter three) and fascist regimes (see

Sarti 1971; 1974).

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (32)


frequently implied that independent, mobilized, and perhaps militant trade unions whereto be tamed, co-opted, and integrated into the state, a situation clearly at odds withcontemporary China. Even in previous instances in which the state created new labororganizations to serve corporatist goals, few if any states have been able to police theirlabor monopoly as tightly as in China. In other words, much as corporatism is ahistorically specific political arrangement, appropriated representation can only emergewith a relatively highly developed and diversified state apparatus.

If appropriated representation refers to the general conditions of labor politics, theoligarchy-democracy axis is useful in describing the organizational practices of the union.Robert Michels’ classic Political Parties (1962) is a study of socialist parties in early 20th

century Germany. Methodologically, Michels chooses for his study the organizationwhere one would be least likely to witness the development of organizational oligarchy,namely socialist parties which proclaimed fidelity to the practice of radical democracy.The unfortunate conclusion of the research is that even parties which claim to adhere toradical democracy are – as a result of the technical requirements of instrumentallyrational decision-making – inevitably doomed to abandon their radical origins and ossifyinto oligarchy. This is what is meant by the term the “iron rule of oligarchy.” From thisperspective, processes of delegation and representation are inimical to the full floweringof democracy, since professional officers become oriented to goals and principles ofaction that may differ markedly from that of their constituents.

The study identifies a key paradox of modern society: on the one hand,democracy requires organization, while on the other, organization necessarily underminesdemocracy. But why is this? Michels argues that in a (formally) democratic polity, it isnecessary for different social groups to be able to formulate demands, a fact owing to thestructure of the state. This requirement to formulate collective demands is particularlypressing for dominated classes, as their only advantage is one of numbers. Elites can actmore efficaciously as individuals, and so the problem of organization is less crucial forthem. The working class, however, cannot hope to exercise any power if it is divided, andso parties and unions become a key tool for them in advancing their interests.

This line of argumentation has been developed by a wide array of theoristsworking not on political parties but rather on trade unions. Lipset et al. (1956) set the tonefor much of the literature on unions in the U.S. as they expressed skepticism about thecapacity for unions to function democratically over time. Their argument is thatsomething approaching substantive democracy is only possible in small unions thatrepresent relatively affluent and well-educated workers such as the InternationalTypographers Union (ITU). Subsequent scholarship has focused on the de-radicalizingeffect of union centralization (Roomkin 1976), the frequent suppression of internaldissent (Jacobs 1963), and the “futility” of attempting to attain union democracy(Magrath 1958), among other topics. Kay Stratton (1989) conducted a study of the ITU,the one organization which Lipset had identified as an exception to the iron rule ofoligarchy thirty years earlier. She found that even this glowing example of uniondemocracy eventually resorted to greater authoritarianism in an attempt to deal with theexternal shocks it encountered in the 1980s.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (33)


While the general consensus is certainly that large unions exhibit tendenciestowards oligarchy, there has been quite a bit of pushback against the notion that there isno variation. This has resulted in a number of studies that demonstrate that not all unionsare anti-democratic. Some scholars have pointed to relatively open union structures(Edelstein 1967; Edelstein and Warner 1976), assuming that they are indicators of stronginternal democracy. Additional factors which may increase the democratic functioning ofunions include strong internal opposition groups (Stepan-Norris 1997) and grassrootsworker insurgency (Stepan-Norris and Zeitlin 1996), while more recently some scholarshave argued that internet-based communication technologies may enhance the voice ofrank and file (Greene, Hogan, and Grieco 2003). Though more concerned with unionrevitalization rather than expressly with democracy, Voss and Sherman (2000) show thateven seemingly deeply oligarchic unions can be energized if there is support from theinternational leadership, conflicts among local leadership, and experienced activists in thelocal to help propel organizational change. Similarly, Stepan-Norris and Zeitlin (1991)suggest that radical ideology among union leadership can have a significant effect for theoutcome of contract negotiations. Thus, we can see that there is a rough consensus amongU.S.-focused labor scholars that while oligarchy may be a prevalent trend among unions,it is not as inevitable as Michels would have us believe.

Despite the relatively large literature on union oligarchy, some problems stillexist. The existing literature has failed to critically engage questions of how oligarchyrelates to interests, representation, and outcomes of union activity. Additionally, thesestudies have tended to focus on internal organizational factors in assessing the dynamicsof oligarchy, but have not studied the impacts of worker insurgency and broaderconditions of the political economy. These issues require additional attention before itwill be possible to effectively deploy the concept of “oligarchy” in my own research.

Problematizing Oligarchy and Interests

Michels poses an important problem which he himself fails to satisfactorilyanswer, namely, to what extent do oligarchic organizations produce oligarchic policy(1962:334)? To put it another way, do oligarchic means necessarily produce oligarchicends (of action)? While Michels is deeply pessimistic on the possibilities for humanfreedom in modern society,24 he wavers on whether or not socialist parties have actuallybenefited their constituents, the working class:

The importance attributed to the masses increases, even when the leaders aredemagogues… This may give rise, in practice, to great inconveniences, such as werecognize in the recent history of all the states under a parliamentary regime; in theory,however, this new order of things signifies an incalculable progress in respect of publicrights, which thus come to conform better with the principles of social justice. (1962:333)

This debate about whether or not unions have to be thoroughly democratic in order topursue their members’ interests has certainly been an ongoing one for practitioners. Thedebate among scholars has similarly failed to come to consensus on the issue, with some

24 e.g. “Thus the majority of human beings, in a condition of eternal tutelage, are

predestined by tragic necessity to submit to the domination of a small minority, and must

be content to constitute the pedestal of an oligarchy.” (1962: 354)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (34)


believing that democracy does not necessarily produce better “results” (Jacobs1963:146), with others disagreeing sharply (Levi, Olson, Agnone, and Kelly 2009;Stepan-Norris and Zeitlin 1995). This debate mirrors that in political and socialmovement organizations more broadly about the relative merits/shortcomings ofcentralization vs. decentralization.

This then brings up some theoretical issues related to interests. In the literatureI’ve discussed, working class “interests” tend to be unilaterally assigned by the analyst.This is perhaps a methodologically necessary move, because levels of oligarchy can thenbe determined through an analysis of the divergence between assigned interests and theinterests the representatives pursue. For instance, Michels assumes that socialistrevolution is in the interests of the working class, and so failure on the part of laborparties to pursue a revolutionary agenda is clearly an indication of oligarchy. Without“fixing” interests, it would be impossible to come to such conclusions (otherwise theanalyst would become mired in an eternal debate about the “true” interests of the group inquestion). To abstract slightly, what most scholars have done (and what I do in thisresearch) is to make an assumption about what the ends of action ought to be, therebyeliminating discussions of substantive rationality. Those who have studied unions havemostly assumed that democracy is a means to an already-determined end, i.e. “workingclass interests.”

While I have already admitted that I follow just such an approach, it is importantto point out that, practically speaking, the determination of working class interests cannotbe a unilateral process, accomplished either by a representative or by a disinterestedanalyst. The collective interests of any social group do not enjoy ontological primacy, butrather can only be “discovered” through a process of dialogue. I do not mean to valorize“empirical interests” (i.e. whatever someone says is in their interest is in their interest)over “objective interests.” Rather, the point is that, particularly for dominated classes, theformulation of interests is a social process that demands a dialectic between empiricalinterests and analytically-determined interests (Offe and Wiesenthal 1980:89-91). Thisprocedure is what Habermas refers to as “communicative action,” (1962:86; 1984) and isfundamental to democracy. In this sense, the “interests” of the working class are dynamicand can only be grasped through democratic organization and dialogue. One implicationof this is that class formation and interest formation are necessarily intertwined processes.Thus, oligarchy should be seen as problematic not only because it may inhibit therealization of analytically-determined interests (confounding of instrumental rationality),but at a more fundamental level it prevents class interests from existing as such(confounding of substantive rationality). This formulation puts us in the position to beable to answer Michels’ question about whether oligarchic organization producesoligarchic policy: While oligarchic organizations may be able to win victories which areperceived by some as being in the interests of membership, the policy will necessarily beoligarchic if it is not determined in dialogue with membership.

It is now possible to articulate a framework for understanding the operation ofoligarchy for my own research. I would like to argue that there are two levels on whichwe can analyze oligarchy within unions. The first is to determine to what extentmembership is actively involved in determining the ends of action for the organization asa whole, something I will call ends formulation. Ends formulation is democratic, ratherthan oligarchic, to the extent that rank and file membership is able to freely vote in and

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (35)


stand for elections, recall incompetent or corrupt representatives, participate in andoversee budgeting processes, form internal opposition caucuses, openly criticizedelegates, and participate in union interactions with representatives of state and capital.Conversely, lack of transparency, unaccountability, restrictions on debate and internalassociations, and non-engagement with/disregard for the wishes of membership on thepart of union leadership, elected or otherwise, are indications of the existence ofoligarchy in ends formulation. I should once again emphasize that I do not believe thatworkers are infallible in assessing their own interests (i.e. what the union should do forthem); intellectuals and analysts have the capacity to make suggestions which can expandthe imagination of the possible. It is only through a dialogue between the theoreticallypossible and the assumed horizon of action that truly democratic ends formulation canoccur. Finally, I here make the assumption that the discussion about ends will necessarilyentail a conversation about means. It is hard to imagine democratic ends formulation andoligarchic means formulation existing within a single organization.

In the event that democracy in ends formulation is imperfect (as it is in allactually existing organizations), we can add “pursuit of analytically-determined interests”to our understanding of oligarchy. As has already been indicated, by “analytically-determined interests,” I mean the interests of a particular group as determined unilaterallyby a disinterested analyst. This has sometimes been referred to as “objective interests,” aterm I reject on epistemological grounds,25 but which implies a similar assigning ofinterests. Pursuit of analytically-determined interests provides an indication of oligarchybecause we can measure the gap between the interests of the membership and the goalsthat representatives pursue. This is precisely the operation Michels employs in arguingthat the failure of socialist parties to pursue a revolutionary agenda is an indication ofoligarchy. But why is failure to pursue analytically-determined interests an indication ofoligarchy? As argued extensively by Weber and Michels, processes of bureaucratizationand delegation entail the creation of a class of political elites, the representatives.Bourdieu (1985; 1989; 2003) has done the most work to demonstrate that Gramsci’snotion of “organic intellectual” (1971:6) cannot be realized in practice; oncerepresentatives enter the political field they are subject to different rules, logics, andstakes of struggle than those of their constituents, regardless of their social origins. Thisdoes not mean that the activities of representatives are completely severed from thewishes of their constituents, as their position in the political field is to a certain extentdependent on their ability to convincingly stand in as the embodiment of the group inquestion. But it does imply separation. The question then becomes, to what extent are thegoals pursued by representatives distinct from the analytically-determined interests oftheir membership? To the extent that this gap grows, we can say that oligarchy isincreasing as it indicates a deepening of the separation between the interests of therepresenters and the represented. For the purposes of this project, the analytically-

25 It should be noted that I do not fully reject the possibility of “objective interests,” but

rather that I reject the way in which they have tended to be determined. I believe that

objective interests can only come into existence with the emergence of a subject capableof action. In this sense, working class formation and determination of the "true" interestsof the working class are one and the same process. And this process is predicated ondemocratic organization.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (36)


determined interests of the working class will be assumed to lie in expanded

decommodification. Finally, it is important to note that this formulation focuses on theactive pursuit, but not necessarily realization of analytically-determined interests. This isbecause realization of a set of interests is to a certain extent dependent on structuralconditions which are beyond the immediate control of representatives. Thus, it is thepursuit of interests rather than a necessary realization that is the appropriate site foranalyzing oligarchy.

While we cannot assume that pursuit of a set of interests will necessarily lead totheir realization, it should be mentioned that it is not just the ends an organizationpursues, but also the means it employs which can serve as an indicator of oligarchy.Means that do not involve broad-based, democratic organizing, but rather depend on theadministrative, political, and symbolic capacities of a small group of representatives areoligarchic. The clearest example of such a tactic, and one which is frequently used bynearly all unions, is lobbying of political elites. On the other hand, when membership ismobilized for mass-based activities such as strikes, direct actions, or more mundane taskssuch as receiving training in relevant labor laws, bargaining skills, and handlinggrievance procedures, these constitute democratic means. I do not mean to imply thatunions choose one or the other; indeed, regardless of the ends that they are pursuing,most unions will employ a combination of some means that are relatively democratic andbroad-based with more oligarchic ones. The significant point is that the relativeweighting of such means says a lot about power relations within the organization.

Now that I have provided re-worked definitions of the countermovement andoligarchy, it is time to bring these two theories into confrontation with each other in anattempt to gain greater clarity on the relationship between worker insurgency and unionactivity in China.

The Polanyi-Michels Dilemma

With this conceptual background in mind, we are now in a position to bringPolanyi and Michels into confrontation with each other in the context of contemporaryChina. Over the course of China’s market reform era, severe commodification of labor(among other things) has led to rapidly increasing social unrest, an occurrence whichwould lead Polanyi to predict that society would soon move to protect itself. Here I relyon Burawoy’s (2003) specification of the Polanyian concept of society as a spacebetween the state and the market where resistance to both is formulated. Rapidlyincreasing worker insurgency in China is an indication that just such resistance isexpanding. But when it comes to sustained organizing of workers across time and space,there is no society, only the state26. The experience of other industrializing countriesindicates that labor cannot be decommodified without the working class exercisingrelatively coordinated collective power both at the point of production and in the politicalsphere. Historically, labor unions have been the organizational form through which such

26 All independent worker organizations are either crushed or severely limited in their

action. In 2006 local activists near Guangzhou were removed from industrial zones after

passing out leaflets which contained nothing other than text from the labor law. This is an

indication of the level of repression when it comes to labor in China.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (37)


power has been exercised. However, in China the only unions tolerated by the state –those that are subordinate to the ACFTU – are nearly ideal-typical cases of oligarchicorganization, and support the state’s agenda of maintaining highly commodified labor.Unless this oligarchy can be broken, it is highly unlikely that decommodification cancome to pass. In other words, in China the countermovement has to go through the ironrule of oligarchy. It is this tension between social rejection of commodification andchallenges in realizing decommodification and incorporation at the institutional level thatis at the core of my investigation. In this sense, one of my primary concerns is about howto analyze the consequences of insurgency.

Insurgency and its Consequences

To start with, it is important to distinguish worker insurgency of the type thatexists in China from a social movement as defined by the “political process” theorists.Charles Tilly, one of the key theorists associated with the political process approach,defines a social movement as “a sustained challenge to power holders in the name of a

population living under the jurisdiction of those power holders by means of repeated

public displays of that population’s worthiness, unity, numbers, and commitment.”(1999:257) Classic works from this tradition (Andrews 2004; McAdam 1982; McAdam,McCarthy, and Zald 1996; Tarrow 1998) have not deviated much from the abovedefinition. These studies have focused on movements that tend to display the followingcharacteristics: 1) Relatively coherent political program and well-articulated goals; 2) Apreponderance of formal “social movement organizations” which are necessary inarticulating said goals; 3) Targeting of the state; 4) Exploitation of political space whichis available in liberal democracies (e.g. through public marches, media outreach, politicallobbying, etc.) As described by McAdam, one of the major reasons for the emergence ofsuch a perspective was because of a reaction to previous theories which viewed socialmovements as necessarily “irrational.” (1982:5-20) The consequence was that politicalprocesses theorists fought hard to demonstrate that social movements are in factcompletely rational responses to deeply-held grievances that cannot be resolved through“politics as usual.”

This conception of social movements as relatively coherent entities with definedgoals27 has resulted in a particular approach to studying the outcomes of protest. Mostsignificantly, an analysis of policy outcomes has been a recurring theme in much of theliterature (Giugni 1998; Snow and Cress 2000; Soule, McAdam, McCarthy, and Su 1999;Tilly 1999), likely owing in part to the types of movements that were studied, but also tothe methodological coherence it provides. By the late 1990s, this group of scholars hadidentified the problem of “unintended consequences” as well as more amorphous culturaland political changes that can result from social movement activity. But the followingcomment from Giugni, McAdam, and Tilly is instructive in terms of how politicalprocess theorists think of this question: “While it is certainly true that social movementsare rational efforts aiming at social change, their consequences are often unintended and

27 I do not mean to caricature these studies, as many have noted that social movements

contain a broad array of actors, organizations, and goals, which do not necessarily work

in perfect concert with each other.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (38)


are not always related to their demands.” (1999:xxi) Here we can see that even ifconsequences are unintended, the movements these scholars are concerned with are“rational” and have specific political “demands” against which outcomes can bemeasured.

Unfortunately, the framework developed by political process theorists cannotexplain the dynamics or the consequences of labor protest in China, in large part becauseof the lack of formal organization and the apolitical appearance (if not reality) ofcontention. I follow Ching Kwan Lee’s (2002) assessment of Chinese worker activism asa form of “insurgency” rather than as a social movement. This concept of insurgency wasfirst fully articulated by Ranajit Guha – one of the primary scholars behind thedevelopment of subaltern studies – in Elementary Aspects of Peasant Insurgency in

Colonial India (1983). Guha admits that the movements he studies are “apparentlyunstructured” (1983a:5) but takes to task those who believe that politics is only“conscious” when there is, “a ‘conscious leadership’, secondly, some well-defined aim,and thirdly, a programme specifying the components of the latter as particular objectivesand the means of achieving them.” (1983a:5) Though he is arguing that activism does nothave to display such characteristics in order to be considered conscious, he is alsoacknowledging that insurgency does not display said characteristics. An additionalfeature of insurgency is that it is generally a politics of negation, in the sense that attackson the, “superordinate elite… carried no elaborate blueprint for its replacement”(1983a:9). Clearly, such features put insurgency in a markedly different category thantypical social movements.

And yet, the thrust of Guha’s argument is that insurgency is not an irrational,psychologically-motivated outburst of anger at rulers. Rather his point is that despite theappearance of incoherence and the relative lack of a well-defined plan, insurgency isconscious and is also eminently political: “The risk in ‘turning things upside down’ underthese [highly oppressive] conditions was indeed so great that [peasants] could hardlyafford to engage in such a project in a state of absent-mindedness.” (Guha 1983b:1) Theconscious nature of the politics of negation is discernible in two phenomena that can beobserved in insurgent actions, namely “discrimination” and “atidesa.” Discrimination, or“selective violence” implies that the targets of insurgency are not chosen at random.Rather, certain locations, individuals, buildings, etc., are targeted for their specificeconomic, political, or symbolic value. The capacity to engage in such discrimination is astrong indication of political consciousness, even if it is not a highly articulated one. Thesecond phenomenon is atidesa, or a “logic of extension” by which targets are expandedfrom the level of the particular to the categorical level. That is to say, it is a politics ofanalogy whereby it is not just, for instance, “our landlord” that is a target, but landlords ingeneral. From this we can see that, for Guha, insurgency denotes a form of activism thatdespite initially appearing as apolitical and spontaneous, is actually necessarily politicaland requires the development of a certain level of consciousness.

In general, Guha’s framework is sufficient for conceptualizing worker protest inChina, though I would like to make some further specifications. First, it should be notedthat atidesa-like extension has been relatively rare. Such analogy-based targeting ofprotest indicates a higher level of political capacity than Chinese workers currentlyposses. Related to this point, it is incredibly important to note the cellular nature ofworker insurgency in China (Lee 2007). What this means is that, with a few notable

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (39)


exceptions (Chan and Pun 2009), labor protests in China remain workplace based, andare not organized across sectors or across other enterprises within a given sector. Therehave been many incidents of “copycat” strikes, but the organization of such resistance isstill taking place at the workplace level. Worker insurgency has failed to extend acrossspace or, for the most part, time. This is related to a final point I would like to makewhich is on the relationship between consciousness and politics in insurgency. Guha’swriting is framed as a polemic against those who claim that insurgency is spontaneousand apolitical. I agree with him that insurgency does require planning and that thecapacity to discriminate in target selection is an indicator of some sort of consciousness.But, certainly in the case of insurgent Chinese workers, that consciousness exists only inembryonic form. When workers go on strike, block roads, petition government buildings,etc., their demands are not framed as political demands (e.g. allow for freedom ofassociation, legalize strikes, end the discriminatory household registration system), butrather as immediate economic ones. Thus, the “virtual unity” (Hardt and Negri 2000:262)of worker insurgency in China appears apolitical when viewed as a series of seeminglyunrelated incidents, and workers themselves generally do not claim a specific politics.However, when viewed at the aggregate level, this insurgency has a deeply politicalcharacter since it implies a rejection of commodification. That insurgent workersthemselves do not grasp this as such is an indication of the fact that their politicalconsciousness is in embryonic form; the capacity to comprehend and articulate thepolitical nature of insurgency in particular struggles would be an indication of a muchhigher level of consciousness and the beginning of the passage to political subjectivity.Thus far, workers have not possessed the means for reflecting the aggregate of seeminglyapolitical cellular insurgency back on the cells, thereby revealing the emerging politics ofdecommodification. This of course is only possible through organization, an opportunitynot currently afforded to workers in China.

The question of the relationship between insurgency and organization requires abit more attention, and in doing so it will be helpful to draw on Piven and Cloward’sPoor People’s Movements (1977). One of their main arguments is that it is throughdisruptive, un-organized outbursts that social movements secure most of their victories.Squarely placing themselves in the Weberian-Michelsian tradition, they argue thatorganization is a necessarily conservatizing force, and that movements lose their efficacyonce they are integrated into such formal bodies. My perspective is somewhat different,in part because the structuring of political space is quite different in China than it is in theUnited States (where Piven and Cloward’s study was focused). In China, the insurgentmoment of the countermovement has been confounded by the problem of oligarchy. Toput it another way, worker disruptions have not been translated into the type of victoriesPiven and Cloward would expect precisely because of a lack of organization. While theorganizations they studied may have had a conservatizing effect on their respectivemovements, the deep oligarchy of ACFTU unions, and non-existence of moreautonomous organizations, threatens the possibility of any political victories. Withoutorganizations that maintain any legitimacy among workers, the working class does notexercise coherent collective power either in the political sphere or at the point ofproduction. This second point implies that even pro-labor legislation that is passed inresponse to insurgency is frequently un-enforced on the shopfloor. The authoritariannature of China’s polity makes it much more difficult for social insurgency to be

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (40)


effectively co-opted, one result of which is that the insurgent moment may not translateneatly into an institutional response.

But how specifically should we analyze the responses of state and union? Thefirst and most important question, of course, is whether or not worker disruption leads todecommodification and incorporation. This can be analyzed both at the micro-level (theenterprise) and at the sectoral, regional, national, and even international level. The secondequally important question is to see how the state, through the auspices of the union,responds to the threat (or reality) of continued disruption. Does the union become moreresponsive to worker needs? Are power relations at the point of production reconfiguredsuch that workers (or their legitimate representatives) exercise some authority (i.e. doesworker power expand)? Merely analyzing decommodification on a formal level (i.e. iflabor appears to be decommodified by a particular piece of legislation or a collectivebargaining agreement) is insufficient. Thus, it is not just formal (on paper) butsubstantive decommodification that I am interested in, and the capacity to enforceformally decommodfying administrative agreements is crucial to this.

Verifying that seemingly decommodifying formal agreements are actuallyenforced is important in determining whether the insurgent moment has actuallyprecipitated the institutional moment of the countermovement. This is why the micro-level perspective is so important in my study. Because of the model of capitalaccumulation and the structure of the state in China, many contracts and labor lawsfrequently go unenforced. Given this, we will need to understand a bit more about thenature of the Chinese state in order to have a more nuanced understanding of the politicsof decommodification. We will now turn to an analysis of the complex, oftencontradictory state structures and how the various levels may respond to worker protest.

The State and Labor Politics

To point out that the Chinese state is a complex, contradictory, and fragmentedentity will come as no surprise to any student of politics. The question for anyonestudying the state is not whether it requires disaggregation, but precisely how it ought tobe disaggregated. I would like to argue two things: 1) Decentralization of economicdecision making in the reform era has created a strong alliance between capital and thelocal state in China; 2) The “relative autonomy” (Block 1977) of the central state inpassing pro-labor legislation is undermined by local political dynamics and the lack ofcountervailing forces in society. Additionally, the appearance of autonomy on the part ofthe central state emerges in large part from an ideological operation intended to enhancethe stability of domination. As an integral part of the state structure, these dynamics arealways present in union actions aimed at reducing worker insurgency.

I follow Nicos Poulantzas’ conception of the state as the “factor of cohesion”between the various classes in capitalist society (1973:44). This of course does not meanthat the state is monolithic, with its various levels marching in lockstep: “…we can see anindication of this [cohesive] function of the state in the fact that, although it is a factor ofcohesion of a formation’s unity, it is also the structure in which the contradictions of thevarious levels of a formation are condensed.” This evokes Bourdieu’s distinction betweenthe left and right hands of the state (1998), in the sense that conflicts within the state are –to a certain extent – a mimesis of class conflict in society. But this is not to say that state

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (41)


power is exercised autonomously from existing class relations; in a moment in which theinterests of capital are hegemonic (such as is the case in China) cohesion is necessary tomaintain a particular political economy the benefits of which – both symbolic andmaterial – are necessarily classed in their distribution. Fundamental to the maintenance ofcohesion of totality in capitalism is ensuring that class struggle appears as a purelyeconomic, rather than political, conflict. This is achieved through an effect of isolation, or“the effect of concealing from these agents in a particular way the fact that their relationsare class relations.” (Poulantzas, 1973:130) This effect requires that social conflictsappear as conflicts between individuals rather than classes. The juridico-politicalsuperstructure within a given formation is critical in this process as it produces workersas juridical-subjects, endowing them with a set of individual rights according to whichconflicts are to be adjudicated. Effects of isolation also operate on the ideological level inthe sense that the state attempts to appear as classless, as a, ”popular-national-class state,in the truest sense. This state presents itself as the incarnation of the popular will of thepeople/nation.” (Poulantzas, 1973:133) “Isolation” refers to the effect that these variousprocesses serve to both isolate individuals from dominated classes from other members oftheir class (i.e. prevent organization) and, as a corollary to this, isolate economicstruggles from political ones that inevitably have transformation of the state as a goal.

This effort on the part of the state to maintain the unity of a formation should notbe confused with actual unity. As has been mentioned, the state contains a distortedreflection, but a reflection nonetheless, of the various conflicts in society. The divergencebetween the appearance of unity and empirical existence of deep conflicts within the stateis nicely captured by Migdal’s distinction between the image and practices of a state(2001:15). In China in particular, the state attempts to project an image of unity,coherence, and discipline within its various branches and levels. But this image of unity,as the unmediated expression of the will of the nation, of course is disproved in practicewhere numerous conflicts, contradictions, and subversions are enacted between variousparts of the state (see especially Chapter 6). Such messiness is inevitable if the state is toattempt to provide coherence and unity within a necessarily conflict-laden capitalist modeof economic organization. The acuteness of such conflict, and therefore the divergencebetween state image and practice, is of course intensified in the process of capitalistindustrialization.

These conflicts do not, short of social revolution, spell the end of the state.Echoing Gramsci (1971:219), Poulantzas argues that the primary task of a capitalist stateis to maintain an “unstable equilibrium of compromise.” (1973:192) Unstable refers torecurrent class conflict in society; equilibrium denotes a balance of forces in society suchthat basic order is maintained, though it does not imply equivalence between the forces;and compromise refers to the deal which must be struck between warring classes andwhich must be realized and enforced through state power. An unstable equilibrium ofcompromise is the condition under which continual capital accumulation can occur, ascenario impossible without the active intervention of the state. The question nowbecomes, how effective has the Chinese state been in maintaining such an unstableequilibrium of compromise?

Insurgency trap and the structure of the Chinese developmental state

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (42)


Beginning with the establishment of special economic zones (SEZ) in southernChina in the late 1970s, the central government has gone about dismantling the commandeconomy and devolving signficant authority over economic policy to lower levels of thestate. While the autonomy of SEZs was originally quite exceptional, the subsequentdevolution of economic decision making power to lower levels of the state resulted in a“market-preserving federalism,” (Montinola, Qian, and Weingast 1996) meaning thatstrong incentives were put in place to prevent local governments from overturningmarketization. Administrative reform reconfigured the fiscal incentives such that thebenefits of rapid economic growth would accrue increasingly to the local state (Jin, Qian,and Weingast 2005). In addition to the formalized incentives for local states to pursueinvestment and accumulation, lack of transparency and oversight within the bureaucracygives individual officials great leeway in securing personal benefits (i.e. engaging incorruption) from providing capital with a favorable investment environment. Theconsequence is that both local governments and local agents of the state have strongincentives to align themselves with the interests of investors.

This has given rise to a political situation Ching Kwan Lee terms decentralized

legal authoritarianism (2007). Lee argues that marketization and decentralization ofeconomic policy-making has produced massive worker insurgency which is, by andlarge, directed against the local state. The central state, more concerned with the stabilityof the economic system as a whole, has passed a series of laws which theoreticallyincrease individual rights for workers. However, the lack of collective rights, i.e. freedomof association, undermine the possibility of strictly enforcing these individual rights(Chen 2007). Additionally, decentralization provides local states with a strong incentiveto side with capital in adjudicating labor conflicts. The result is that the legal reformsenacted by the central state have had the effect of individualizing much labor conflict,while many of the laws themselves often gone un-enforced. The central state’s project ofrule by law (rather than rule of law) has thus greatly influenced dynamics of labor protest,but has not given rise to a transparent, rule-bound, political process, nor has it resolvedthe underlying source of conflicts.

Despite the strong incentives local governments have to side with capital, oneshould resist the temptation to caricature these levels of the state as fully subordinate tothe demands of investors. Though perhaps less concerned with the long-term capacity foraccumulation than the central state, local officials often times take an interest in theoverall stability of production. The resolution of labor conflict is of course central tomaintaining production, and so various segments of the judiciary, the government, and,most importantly for this project, the trade union, will occasionally side with workersagainst capital. Since it is these agents of the state who preside over the actualimplementation (or not) of potentially decommodfying legislation, their capacity toovercome the strong incentives to side with capital in defense of workers is of utmostconcern.

Let us recall for a moment the earlier formulation of political power for thecapitalist state being based on an unstable equilibrium of compromise (Poulantzas1973:192). In China we can see that the structure of the state and its model of capitalaccumulation have produced a severe disequilibrium between labor and capital,particularly at the point of production. This disequilibrium paired with deepcommodification has produced great conflict, which in turn has engendered an attempt at

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (43)


“compromise” from various levels of the state. In my terms, this means that China hasstalled at the insurgent moment of the countermovement, and it is unclear that acompromise, i.e. institutionalized decommodification and incorporation, can be realizedin a political context in which the state is constitutionally opposed to the working classdeveloping a (semi-)autonomous base of power. To push this a bit further, the propositioncan be made that it is necessary for the capitalist state to allow for a degree of politicalpower for dominated classes (particularly the working class) in order to maintain theequilibrium necessary for continual hegemonic rule of capital and relatively stableaccumulation. Without this power for the working class, compromise becomes incrediblydifficult to realize in practice.

Thus we can see how crucial an understanding of the ACFTU becomes. This isthe arm of the state which is concerned with worker representation, and is thereforenecessary for any sustainable compromise to be made. There are intense pressures on theunions to resolve the conflicts generated by commodification, but it is unclear if thesepressures will be acute enough to overcome profound oligarchy. This presents us with afundamental paradox of China’s current political economy: that which is categoricallybanned by the state, i.e. organized worker power, may be the only means by which theclass compromise necessary to overcome worker disruptions and maintain continualaccumulation can be realized, a predicament I refer to as an “insurgency trap.”28 Laborpolitics in general, and particularly the activities of ACFTU unions, become the ideal sitein which to analyze whether this trap can be undone.

Additionally the point must be made that in a country as large as China, variationin regional political economy creates different dynamics of labor politics (Hurst 2009).Although the conditions I have described thus far hold to some extent in differentregions, the particulars of the models of development will influence the triangularrelationship between workers, state, and capital. In particular, the type of industry,national origin of investment, qualities of social embeddedness, etc, will influence thestate of equilibrium and possibilities for reaching compromise. This intra-nationalvariation in class politics will be made apparent in Chapter five.

Is China Capitalist and Does it Matter?

As a final theoretical point, I would like to address the question of whether Chinais capitalist, and whether it matters. Over the past several years, there has been a largeamount of literature which has attempted to come to terms with the contradiction of aformally socialist state presiding over a markedly capitalist approach to development

28 While there are some areas of overalap with Minxin Pei’s (2006) “trapped transition,” I

should be clear that an insurgency trap is actually quite different. Pei argues that China

has failed to transition to a fully marketized and liberal-democratic society, and that this

“trap” may inhibit future growth potentials. He believes that this stalled transition is due

to the particular, contingent decision making of high-level leaders. An insurgency trap,

by contrast, develops because of the structure of the state, model of accumulation, and in

particular, modes of class domination. Pei argues that the stalled transition will hold

negative effects for Chinese society as a whole, while I draw attention to the classed

nature of domination and unequal gains from growth.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (44)


(Hart-Landsberg and Burkett 2005; Hung 2009; Weil 1996; Zhang and Ong 2008). Somescholars as well as capitalists and agents of the state will claim that we live in a “post-ideological” era, and so labels such as capitalism and socialism are unimportant.29 Thisargument is, however, deeply ideological in the sense that it claims the current socio-economic arrangement as natural and without a specific political content. I argue that oneof the most remarkable developments of the past thirty years is that China has becomecapitalist, and that this fact is of the utmost importance in understanding contemporaryChinese politics.

For the purposes of this research, I conceive of China as capitalist for two primaryreasons. The first is that the state has taken the interests of capital as hegemonic. Here Ido not make a strong distinction between domestic, foreign, and state capital, since state-owned enterprises are to an increasing degree operated accorded to market principles(Gallagher 2005; Zhang 2008) including hiring and firing policies, managerialcompensation, and private appropriation of surplus. The fundamental organizingprinciple of all such enterprises is the realization of capital valorization, though of coursethe methods employed toward this end vary widely. That the trade union – perhaps thearm of the state most likely to side with workers – has become subordinated to thehegemonic interests of capital will become increasingly clear as we proceed. For the statemore broadly, its function as a factor of cohesion is represented on the discursive level bythe concept of “harmonious society,” a framework which is an attempt to winsubordination to reconfigured relations of class domination within a marketized context.Workers, among other dominated classes, are asked to unquestioningly submitthemselves to high rates of exploitation in the service of maintaining the coherence of thecurrent system of capital accumulation. Such subservience is claimed to be necessary inorder to maintain a good investment environment, develop the economy, and mostsignificantly, build the (classless) nation. As for the effect of isolation discussed earlier,the project of rule by law (fazhi) is a clear case of the state attempting to separate theeconomic from political struggles. The individualization of employment rights signalssurrender to a basic tenant of the logic of capital, namely that success or failure in themarket is due to individual capacity rather than class-based power relations. Finally, thestate-led destruction of the social contract including the evisceration of comprehensivehealth, education, and pension systems, has been framed as necessary for developmentand becoming a “modern” and powerful country. The end of generalized social welfare isof course important in the production of a free labor market.

This then leads to the second reason why China ought to be characterized as aproperly capitalist society, namely the high levels of commodification of labor. As hasalready been discussed, I consider labor to be commodified to the extent that workersmust immediately submit the satisfaction of their needs to the logic of the market.Additionally, levels of labor commodification can serve as a sort of index of overall

29 For an analysis of this issue in China, see Guo, Yingjie. 2009. "Farewell to Class,except the Middle Class: The Politics of Class Analysis in Contemporary China." The

Asia-Pacific Journal 26. As well as, Wang, Hui. 2006. "Depoliticized Politics, From Eastto West." New Left Review 41:29-45.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (45)


commodification of human needs. Without state-guaranteed access to things they need,e.g. housing, education, health care, etc., workers’ labor power has become increasinglycommodified over the past thirty years. Some may argue that many migrants’ ability toretain land rights has a decommodifying effect; this is surely the case, and if the statewere to go through with full privatization of land (as it was considering in 2008), thiswould make the situation even more dire. Yet with prices of grain falling after WTOentry, the destruction of the rural social welfare system, continual inflation, and withhuman needs increasing due to cultural work of commodity marketing, migrant workersneeds are highly mediated by the market. The situation is perhaps even more severe forlaid-off workers from the former state-owned sector who do not even have access to land.This commodification of labor has been institutionalized at the national level through theexercise of state power. It is for these reasons that by the beginning of the second decadeof the 21st century, China must be considered capitalist.

I should point out that there are many other ways in which China could becounted as capitalist. These include, but are not limited to: 1) A preponderance of theproduction of goods as commodities; 2) Formal protections for the rights of many formsof property (land being an incredibly important exception); 3) The widespread existenceof the capitalist habitus; 4) Increasing formal representation for capital within the state;30

5) The extension of the logic of capital as a basic organizing principle of social relationsin a variety of fields; and perhaps most significantly, 6) Deep integration into thecapitalist word system. Claims that China is still socialist tend to be based on a few typesof arguments. The first is that the percentage of GDP derived from state-ownedenterprises is still quite large. I have already addressed the problem of state capital, andwhy it behaves increasingly like private capital (i.e. it is primarily oriented towards self-valorization rather than socially defined needs). The second is that the state remainsactively involved in regulating many markets, including most importantly finance,energy, and transportation. But states in the West have long engaged in active regulationof similar markets, even if the ideological claim contradicts this. The difference can onlybe a quantitative rather than a qualitative one. Finally, there is the tautological positionthat China is socialist because the state says it is socialist. I do not think this finalargument is worth seriously engaging.

That China has become capitalist is crucial to understanding the operation of thecountermovement. If labor were not highly commodified, and if the state had notaccepted the interests of capital as hegemonic, the Polanyian framework would be of littlevalue. However, it is precisely because market logic – previously tightly confined ingeographically and politically circ*mscribed SEZs – has seeped into the very DNA of theeconomic and political structures of Chinese society that we are now facing massive

30 In early 2011, it was reported that the richest 70 members of the National People’sCongress had a combined wealth of 493.1 billion yuan (US$75.1 billion), and increasingnumbers of private capitalists were being invited to serve on the Chinese People’sPolitical Consultative Congress. See March 4, 2011. “Wen Sees Billionaires in Congressas Gap in Wealth Widens.” Bloomberg News. And, March 4, 2011. “Business influencegrows in China.” Financial Times.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (46)


worker insurgency. Institutionalizing the countermovement will require a rejection of thislogic in the economic and the political spheres.


In this chapter I have developed the overall theoretical framework for the entireproject. I have attempted to redirect attention in the Polanyian countermovement towardsresistance to commodification and its outcomes. This necessitates an understanding ofhow worker insurgency is or is not translated into institutionalized decommodificationand incorporation which has been accomplished by distinguishing the insurgent momentfrom the institutional moment of countermovements. Given that the emergence of theinstitutional moment in the process of industrialization has, in the experience of othercountries, required the existence of worker-based organizations, conditions ofappropriated representation and the union oligarchy it produces beomce of centralimportance. I have argued that oligarchy within unions can be measured according toboth ends formulation and pursuit of analytically-determined interests. Because of thelack of space for autonomous worker power, I have argued that deep oligarchy in China’sunions may confound the countermovement.

Additionally, I have sought to gain greater clarity with regards to the form andoutcomes of worker protest in China. I have argued that seemingly apolitical cellularinsurgency has developed among China’s working class. This insurgency is organizedaround immediate economic demands, but has failed to articulate a specific politicalagenda, and it has in general not extended across time or space. The inability of theworking class to develop as a coherent political force is in large part due to the statepolicy of “rule by law” which is fundamentally opposed to such an outcome. Althoughthe state as a whole has taken the interests of capital as hegemonic, there are divisionsbetween its various levels and arms. In particular, the central state’s function as a factorof cohesion has encountered a challenge from local governments’ strong alliance withcapital and the lack of a countervailing force at the point of production. Thiscontradiction is expressed most clearly by the inability of individualized legal rights to berealized in practice, a state of affairs which leads to increased worker protest. The centralstate is thus in a bind where the one way to get its laws enforced and to reduce laborconflict is to do the one thing it refuses to do, namely provide the political space for theworking class to amass collective power. This insurgency trap implies that it will bedifficult for the capitalist state to make the compromise necessary to maintain an“unstable equilibrium.”

Before seeing how these processes play out in our empirical cases, it will benecessary to understand a bit more about the history and structure of the ACFTU.Through such an analysis we will see how the Party systematically deprived the workingclass of an autonomous base of power – a process that started long before the economicreforms of the past three decades.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (47)


Chapter 3

The History and Structure of the ACFTU

The headquarters of the Guangdong Federation of Trade Unions (GDFTU) – themost important provincial level union organization in China – is situated directly betweentwo buildings, each respectively imbued with profound and seemingly diametricallyopposed symbolic values. On the left is the East Garden, the building which housed theshenggang (Guangdong-Hong Kong) strike committee during the revolutionary upsurgeof 1925-27. Initiated in June of 1925, the shenggang strike remains the longestcontinuous general strike in history and it played a crucial role in the development of theChinese revolution. Today the site is closed to the public, although there is a smallcommemorative plaque to the side of the guarded entrance. To the right of the unionheadquarters (indeed, located within the same building complex) is the Guangdong UnionHotel. This hotel is owned in whole by the union federation, but operates just as a normalhotel would and is used for generating profit. In early 2009, management fired an activistin the hotel after she pushed for expanded health insurance, housing subsidies, andovertime pay for low-paid migrant worker employees. Following this incident of blatantand illegal retaliation, the union federation which owned the hotel did nothing to defendthis activist, but stated that they would “not take sides.” Given the prominence this storyhad received in the media, this sent a strong message to potential activist unionmembers.31 Thus, we have two materially objectified symbols of radically opposedmodes of political action. On the left we have the ACFTU’s past: organization,mobilization, and confrontation; and on the right we have its present: atomization,acquiescence, and heteronomy.

And yet, as we will see in this chapter, if the methods have changed dramatically,there is a coherent logic that connects the past and present of the ACFTU. Founded inGuangzhou just one month before the initiation of the shenggang strike in 1925, theACFTU actively organized and mobilized the working class to engage in militant actionsagainst capital and foreign imperialism, including strikes, marches, and armed pickets(Kwan1997; Perry 1993:69-87). However, even at this time, mobilization was primarilydirected towards defeating imperialism rather than destroying capital (Smith 2002).Before final military victory in 1949, ACFTU unions began working to maintain

31 This case is discussed in greater detail in Chapter 3.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (48)


workplace harmony and increase efficiency and output, and any attempts at greater unionautonomy were to be crushed. With marketization and the emergence of a new class ofmigrant workers in the 1980s and 1990s, unions in China found themselves utterlyincapable/unwilling to adapt to capitalist labor relations, bound by heteronymous controlof capital at the enterprise level and the state at higher levels. It is within the context ofsuch a profound and systemic severing of ties between representer and represented thatironic cases such as a union activist being fired from a union-owned hotel can occur. Inthis sense, the ACFTU is as close to an ideal typical case of oligarchic organization ascan be found. But it was the conditions under which the ACFTU was first constitutedwhich defined a persistent internal logic that has demanded the union pursue the inter-related goals of ethno-national autonomy and economic growth. With the destruction andre-creation of the union on two different occasions, conditions of appropriatedrepresentation became generalized at the class level.

In this chapter, I aim to accomplish four things. First, an empirical account of thebasic historical activities of the ACFTU, from its founding in 1925 until 1989. Second, Iwill delineate the major changes in trade union structure. Of crucial importance in thisregard is the dynamic relationship between union, Party, and working class. Related tothis is the third point, which is an analysis of the emergence of appropriatedrepresentation that – although continually subject to contention – was solidified in thepost-revolutionary period. Finally, I will argue that, in general, working classorganization in China has since the 1920s been accepted the goals of promoting ethno-national autonomy and increasing productive forces. Although the danwei systemeffectively incorporated and decommodified labor during the era of the commandeconomy (Lau 2001; Walder 1983; 1984), the organizational logic of unions became thatof the state. As the goals of the Party and those of the working class increasinglydiverged throughout the 20th century (though certainly not in a linear manner), theunion’s subordination to the state meant that the representative relationship to theworking class was increasingly characterized by practical divergence.

The Revolutionary Years - Mobilization

The emergence of labor activism among the ACFTU and its subordinate tradeunions cannot be viewed in isolation from the development of the Chinese CommunistParty (CCP) and the struggle against foreign imperialism. Prior to the Guomindang’s(GMD) purge and massacre of Communists in April of 1927, the CCP had beenfollowing Marxist-Leninist orthodoxy by focusing their organizing efforts on the urbanworking class. While much analysis of the Chinese revolution has focused on theCommunists’ mobilization of peasants (Bianco and Bell 1971; Hinton 1966; Skocpol1979), most of their pre-1927 energies were devoted to building industrial unions and anational union federation, the ACFTU. As the revolutionaries became painfully aware oflater, the Chinese urban proletariat in the 1920s was still quite small. Although theextractive and transportation industries were somewhat developed in many places in thecountry, modern industrial manufacturing was concentrated in two places: Shanghai andGuangdong province. In each of these places, the ACFTU and Communist-affiliatedunions actively sought to organize and mobilize workers, sometimes in defense ofimmediate economic rights, and sometimes for broader political struggles.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (49)


One of the ACFTU’s official histories, A Concise History of Chinese Trade

Unions (Wang, He, and Cao 2005), provides an account of the heroic activism ofCommunist-affiliated unions in the 1920s. Even before the official establishment of theCommunist Party, prominent Chinese Communist Chen Duxiu argued in August of 1920that most unions in Shanghai were controlled by bosses and were therefore useless. Chenclaimed that “10,000 of these [yellow] unions could be formed and it wouldn’t matter,”and that workers should, “unite and organize real worker groups!” (ibid:9) Shortly afterthe formation of the CCP and the establishment of the Party’s Labor Secretariat, therewas a strike wave throughout the country between January 1922 and February 1923.During this time, the Secretariat was actively involved in protesting against the officialban on strikes and organizing strikes in which a total of 300,000 workers participated innationwide (ibid:14). These strident pro-labor activities and proclamations surely helpedsolidify the position of Communist unions among the working class. However, inresponse to a strike of workers on the Beijing-Hankou railroad the ruling warlordslaunched a vicious nationwide crackdown on organized labor. In addition to the dozens ofrailroad workers killed and imprisoned, Communist labor organizations around thecountry went underground.

This low tide of revolutionary labor activity did not last long. By 1925 morefavorable political conditions prevailed in certain regions, and Communist labor activistsbegan regrouping in Shanghai and Guangzhou. Though the Communists and Nationalistswere operating under the “united front” policy, right wing Nationalist labor activistsboycotted the second National Labor Conference, which commenced on May 1st, 1925.This boycott allowed the Communists to avoid the confrontations and gridlock which hadplagued the first national conference and to realize their goal of formally establishing anational labor federation under their control: the ACFTU (Lee 1986:8).

Just weeks after the ACFTU was established, the federation and its subordinateindustrial unions were to receive a series of opportunities to mobilize and increase theircredibility among Chinese workers. The Communist-controlled Japanese Cotton MillsUnion began a series of strikes in 1925. On May 15th, union activist Gu Zhenghong waskilled by a Japanese manager, stoking both nationalist and class based outrage (Perry1993:80). Gu instantly became a martyr of the anti-colonial and class struggle, and hismemorial service was attended by thousands of workers. Presided over by the CottonMills Union chair Liu Hua, a reporter at the event said, “I dare say that this sort of grandproletarian gathering is unprecedented in Shanghai. Nine out of ten participants wereworkers.”(ibid:80) On May 30th, there was a huge solidarity protest in which workersfrom a wide variety of industries gathered to express their dismay over the murder oftheir fellow countryman and worker. But this protest was not to end peacefully: TheBritish authorities opened fire on some of the protesters, eventually killing 13 workers.Then, on May 31st the just-established Shanghai Federation of Trade Unions (SHFTU)called for a general strike which was to last several months. This revolutionary event wascrucial in establishing the SHFTU as a worker-representative organization, and in givingcoherence to the Shanghai working class: “…after the [Shanghai] General Union hadgradually been deserted by its allies, it had done a tremendous job of organization duringJuly and August, and had given the working masses of Shanghai a unity and a crusadingspirit they had never known before.” (Chesneaux, 1968:269) Additionally, the unionfederation very quickly gained recognition in the political field, as indicated by its

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (50)


hosting of a delegation of Soviet trade unionists and increased consultation with theBeijing government on pending labor laws. Between the start of the strike in May andSeptember of the same year, SHFTU-subordinate unions increased in number from 20 to127, and its membership grew from 20,000 to 220,000, a phenomenon which was notlimited to Shanghai: “[during the summer of 1925] regional union federations andnational industrial unions were set up under the leadership of the ACFTU one afteranother.” (Wang, et al., 2005:20) The so-called May 30th incident became a rallying cryfor the emerging wuzhou movement, which immediately thereafter burst on the scene notjust in Shanghai, but throughout China.

The labor movement in Guangdong was not to be left out. With some politicalprotection coming from the Communist-Nationalist united front policy (Li 2001), ageneral strike in Guangdong and Hong Kong was initiated in late June of 1925. TheACFTU and Communist activists were able to gain control over the strike committee,though some Nationalist sympathizers were also involved. As was the case with the laborupsurge in Shanghai, the shenggang (Guangdong-Hong Kong) strike was propelled byanti-colonialism. After Communist-influenced unions in Hong Kong initiated widespreadstrikes starting in mid-June of 1925, the ACFTU called for solidarity action amongworkers in Guangzhou. On June 23rd French and British soldiers attacked protestingworkers and students just outside of Shamian, a foreign concession in Guangzhou. The“Shaji Incident” left 52 people dead and 117 wounded, and led to great outrage amongstpeople throughout the province and the country. Approximately 250,000 workers inHong Kong and Guangdong began a general strike which stretched out over manymonths (Wang et al. 2005:21).

The Primacy of Anti-Imperialism

As has already been alluded to, both the ACFTU and the Party were primarilyconcerned with the anti-imperialist struggle during the mid-late 1920s. I follow S.A.Smith’s characterization of labor struggles in this period as “class-inflected anti-imperialist nationalism.” (2002:190) Although anti-imperialism was the dominantdiscursive frame, Communist unions in particular occasionally weaved class-basedanalyses into their rhetoric. In a sense, anti-imperialism frequently coincided with classstruggle given the large number of foreign-owned enterprises in Shanghai andGuangdong. That the institutional foundations of the Communist union federations wereconstituted largely in opposition to imperialism rather than in opposition to capital was tohold important consequences for the development of labor politics in China.

Through the revolutionary upsurge of 1925-27, union leadership frequentlyemployed the language of minzu when speaking of the need for liberation. The termminzu – which can alternately be translated as “nation” or “ethnic group” – was first usedby nationalists in the late 19th and early 20th century in political rhetoric (Wu 1991:161),and was an attempt to sharply differentiate Chinese from foreign. In open letters fromACFTU leadership to workers involved in the shenggang general strike, the need fornational liberation was always paramount. In one such letter from July 1925, the secondmonth of the strike, union leadership said:

We must safeguard the national [minzu] movement and workers’ movement inGuangdong and Hong Kong in order to safeguard China. We should unite with all the

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (51)


people of Shanghai, and workers, and all oppressed people in the nation in order to resistthe imperialist offensive. (zhongguo gonghui lishi wenxian, 1958:93)

In a separate letter from the same week, the ACFTU very clearly lays out its argument forthe importance of the general strike: “Workers! Why do we strike? We strike to win thefreedom of the nation [minzu]. Who stole our nation’s freedom? It was stolen by theimperialist.” (zhongguo gonghui lishi wenxian, 1958:94)

As revealed in these letters, union and Party leadership primarily viewed theshenggang general strike as a means to eliminate foreign aggression in China. When theACFTU announced the formation of the shenggang strike committee in June of 1925, thissentiment is unequivocal: “Compatriots from Guangdong and Hong Kong areimplementing an allied strike because of their hatred of British and Japanese imperialistmassacres in Shanghai, Hankou, Qingdao, and other places.” (Shenggang Da Bagong

Ziliao, 1980:149) Deng Zhongxia, an early ACFTU leader and one of the centralorganizers of the shenggang strike, was quite direct in his assessment of the movement’sgoals: “This national strike of the working class is originally not an economic struggle to‘increase wages,’ or ‘reduce hours,’ but rather is a political struggle to ‘opposeimperialism and liberate the nation.’” (Shenggang Da Bagong Ziliao, 1980:150) Sinceforeigners were highly dependent on the labor of Chinese workers, particularly inShanghai, Hong Kong, and Guangzhou, the strike exacted a high economic toll on theirprofits.

Despite the refrain from union leadership that “the entire country has a commonenemy,”32 more traditional class-based Marxist language occasionally appeared in officialspeeches. In particular, the commemoration of the 100th meeting of the shenggang strikecommittee in March 1926, nearly one year into the strike, was marked by specificallyclass-based rhetoric and mention of the liberation of the working class rather than thenation. Deng Zhongxia (who in 1925 said that the goal of the strike was to opposeimperialism) stated the following in his address to the assembly:

If the working class wants to achieve full liberation, it will only come after theworking class has seized state power. If the working class tries to implement peaceful andreformist policies before the capitalist system has been eliminated, that is to say under therule of the capitalist class’ government, then liberation will never be realized. The workingclass can only rely on its own class organization and exercise continuous and absolute statepower to eliminate all previous forms of the state, before full liberation will be possible.The Paris commune represented the first historical step in this direction, and Russia’sSoviet government is the second step. (Shenggang Da Bagong Ziliao, 1980:187)

And although “the enemy” remains undefined in this official union statement from April1926, the implication seems that it is capital in general, rather than specifically foreigncapital:

In order for the oppressed working class to liberate itself, it must unite across allboundaries and increase its own strength and defeat the enemy. Workers must manageproduction by themselves. Policies of compromise or reconciliation either intentionally orunintentionally help the enemy and bring our working class into a realm where they willnever be able to achieve liberation. (Zhongguo gonghui lishi wenxian, 1958:197)

32 e.g. ACFTU letter to all national unions which begins with the phrase juguo tongchou,

(the entire country has a common enemy). (Shenggang Da Bagong Ziliao, 1980:151)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (52)


Additionally, many union demands during the wuzhou movement and shenggang strikefocused on bread and butter workplace issues, such as limiting the working day,increasing the minimum wage, and the provision of benefits such as housing andeducation for workers.

It is unclear why such class-based analysis occasionally appears in the officialrecord of the strike, even if it is certainly less emphasized than ethno-national unity.There likely was intense debate and negotiations among workers and intellectuals as towhether an anti-capitalist or an anti-imperialist frame was most appropriate for theChinese labor movement. Additionally, changes in rhetoric may be related to ongoingstruggle between the Communists and the GMD for control over the movement. In thestruggle between the right wing of the GMD and the Communist Party, perhaps theACFTU was encouraged to play up class struggle as a means of differentiation.

And yet, such strong class-based language was clearly the exception in thisperiod. Deng Zhongxia, in a separate address at the very same conference where heclaimed that the working class could never attain liberation without destroying thecapitalist state, had this to say while ticking off the victories of the shenggang strike:

Another victory has been economic development in Guangdong… Since the shenggang

strike, the Central Bank has been fortified. Previously, Guangzhou was flooded with Westerncurrency, and even in the villages, Western currency was ‘better than gold.’ But since the strike,there has been a boycott of Western currency while only using Central Bank currency, so nowCentral Bank currency is ‘better than gold.’ The Central Bank has not only been fortified but hasexpanded, establishing branches in various places. The Central Bank now has more credit thanforeign banks. (Shenggang Da Bagong Ziliao, 1980:214)

Not only was there economic nationalism, but union leadership directly and explicitlysought assistance from Chinese capital in furthering the anti-imperialist struggle. Forexample, in March of 1926, the ACFTU wrote a letter to the Shantou Chamber ofCommerce explaining their position on a cross-class alliance: “[unions] should call onnormal civic organizations to establish a common front to attack the enemy, and shouldestablish a national representative assembly to seize political power.” (Zhongguo gonghui

lishi wenxian, 1958:182) “Normal civic organizations” in this case refers to employerassociations.

In Shanghai, the SHFTU also sought cooperation with the local bourgeoisie aswell as with organized crime. Following the May 30th massacre, the local ChineseChamber of Commerce was the largest contributor to the union’s strike fund (Perry1993:82). The petit bourgeoisie was organized to participate in the general strike,although union leadership later expressed disappointment at this group’s lack ofcommitment. And SHFTU chair Li Lisan, as well as other union activists, had extensiveties with Shanghai crime syndicates, something deemed necessary in order to mobilizethe proletariat. In SHFTU documents from the summer of 1925, aggression is clearly(and justifiably) directed towards Japanese cotton mill owners and the foreign police thatcommitted the massacre of protestors.

But perhaps it is in the closing slogans of union communiqués that the primarilyanti-imperialist slant is most evident. Below is a sampling of such slogans from therevolutionary upsurge of 1925-1927:

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (53)


• From an open letter from the ACFTU to shenggang strikers: “Workers! Keepstruggling till the end! Eliminate the unequal treaties! Defeat imperialism! Longlive the victory of the national revolution!” (Zhongguo gonghui lishi wenxian,1958:95)

• From the closing of the Guangzhou Workers’ Assembly, April 1926: “Ourslogans are: Protect the national government! Support the Northern Expedition!33

Protect the shenggang strike! Defeat the imperialists and their running dogs! Longlive the victorious national revolution! Long live the world revolution!”(Zhongguo gonghui lishi wenxian, 1958:200)

• From ACFTU Secretary Liu Shaoqi’s remarks at the opening ceremony of theLabor Institute in June 1926: “Workers, peasants, business, and students unite!Defeat imperialism! Defeat reactionaries! Establish the revolutionaryfoundation!” (Zhongguo gonghui lishi wenxian, 1958:247)

Reaction and Retreat

The period of revolutionary activity starting after the May 30th massacrewitnessed some of the most significant working class mobilization of the 20th century.Massive general strikes in Shanghai and in Guangdong-Hong Kong had been organizedby newly established union organizations, and had given the working class a coherencethat had not existed previously. And yet, the anti-imperialist struggles of the ACFTU andits subordinate unions ended in defeat when the right wing of the GMD undertook aseries of attacks on the Communists, thereby decisively ending the fragile united frontalliance and driving much labor organization underground.

The beginning of reaction was signaled by Chiang Kai-shek’s (Jiang Jieshi) coupin Guangzhou on March 20, 1926. During the initial stage of the shenggang strike, GMDleadership had publicly expressed support for the workers, though they were careful toemphasize the national, rather than class-based, character of the movement. Communistleaders had failed to grasp the growing threat from the right wing of the GMD, and sowere completely unprepared when Chiang seized power in March (Kwan 1997:164).However, immediately after the coup (which involved confrontation with both Chineseand Soviet Party officials), Chiang quickly engaged in damage control. Rhetorically, atleast, he attempted to maintain his favor with the left wing of the GMD, and continued topledge allegiance to the national revolution. His political maneuvering was at leastsomewhat successful as he received a warm reception at the ACFTU’s 3rd NationalCongress in May of that year (Wu 1968:592).

If the Communists and Chiang were able to paper over the growing rift revealedafter the March 20 coup, no such détente would be possible just 13 months later. Afterthe GMD captured Shanghai in March 1927, Chiang quickly moved to establish analliance with the city’s powerful organized crime syndicates and leaders of industry(Smith 2000:190). And although the general strike and armed insurrection led by theSHFTU on March 21st had played a decisive role in the GMD’s capture of Shanghai,Chiang showed little gratitude towards the Communist activists. Early in the morning of

33 The Northern Expedition was a GMD military campaign to defeat warlords that

controlled large swaths of China.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (54)


April 12th, armed groups of thugs began a systematic attack on union offices throughoutthe city, killing scores and arresting many high-ranking labor leaders. The following day,a group of more than 100,000 workers stopped work and went to protest against themilitary commander in the northern part of the city. With the full support of the foreignpowers, the military opened fire on the protestors, killing one hundred and woundinghundreds more (Chesneaux 1968:3370). Similar attacks against Communist labororganizations took place in many of the other cities under control of the newly-emboldened right wing of the GDM. By the end of April, the strength of these unions hadbeen greatly diminished.

Even if Communist labor activism was not completely eliminated in the cities(Stranahan 1998), space for leftist political activity was extremely limited. With the leftwing of the GMD holed up in Wuhan and Chiang’s reactionary forces dominating thecentral and southeastern parts of the country, the Communists had little room formaneuver. Retreating from the city, they reconvened in relatively remote Jiangxiprovince, where they led a brief takeover of the capital Nanchang in August, 1927.Although they did not hold the city for long, many Communists retreated to the ruralareas of the province, and began laying the organizational groundwork for the JiangxiSoviet. The success with rural organizing in Jiangxi was crucial in the emergence of MaoZedong’s rural strategy (Averill 1987:280). With the initiation of the Long March, Mao’sascendancy in the Party, and the eventual establishment of a base in Yan’an, it was clearthat the working class was not going to be the primary revolutionary agent. The ACFTUfell into disuse at this time, and would not be resurrected until military victory neared in1948.

Socialist Trade Unionism – The First Rupture

Post-revolutionary trade unionism was to be very different from the revolutionaryyears of the anti-imperialist struggle. As the Communist military victory neared, theACFTU was rehabilitated, although in remarkably different political conditions thanthose of the mid-1920s. With the defeat of the Japanese and the turning of the tide in thecivil war, large swaths of the country were firmly under control of the Communists. As aresult, the role of the union federation was inevitably going to change dramatically fromwhat it had been in the anti-imperialist mobilizations in Shanghai and Guangzhou.Starting in the late 1940s, the ACFTU and its subordinate unions were primarilyinterested in encouraging workers to increase efficiency and production. Advocating forthe particular interests of the working class came to be denounced as economistic, and theParty successfully undermined attempts at greater operational autonomy for the union.This resulted in the “first rupture”: that between union and working class. Although thegeneral direction of state policy was for greater decommodification of labor, the workingclass was deprived of any independent base of organized collective power.

The key figure in the post-war development of the ACFTU is Li Lisan. Li, whohad been one of the leaders of the Shanghai labor movement in the 1920s, remained apowerful figure in the Party until 1930. After the split between the GMD and the CCP,the “Li Lisan Line” which called for armed insurrection in the cities gained prominencein the Party. However, these uprisings were ultimately not successful, most notably in

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (55)


Nanchang. Li’s “extremist” politics fell into disfavor and he was banished to Moscow forfifteen years.

But by 1948, when labor organization was back on top of the Party’s agenda, Lihad already admitted his past mistakes and had returned to a position of power. In Marchof 1949, the CCP Central Labor Movement Committee announced a conference to beheld in Ha’erbin in June, the goal of which was to establish a “Liberated Areas WorkerAssociation.” However, as planning was underway for this conference the GMD held aconference in April that they claimed was organized by the ACFTU (despite the fact thatthe ACFTU had always been a Communist organization). In order to combat what wasseen as an attempt to divide the labor movement, the Communists cancelled the Juneconference and on May 1st announced that they would be convening the 6th NationalLabor Congress at which time they would re-assert their rights to the name “ACFTU.”The congress attracted representatives from unions with 2.83 million members, and thedelegates ratified a new constitution for the union federation and passed resolutions tofight the GMD and American imperialism (Li and Liu, 2005:294). Although Chen Yunwas formally the chairman of the revived union federation, Li was elected first vice-chairof the union and assumed effective leadership over day-to-day operations.

Even before the formal re-establishment of the ACFTU, the question emerged asto what precisely unions in areas under Communist control would do. Clearly, theiractivities would be quite different than they had been during the massive strikemobilization of the 1920s. The general direction was definitively established in a 1948report by Chen Yun in which he proclaimed a policy of, “developing production, makingthe economy prosper, caring for both public and private, benefiting labor and capital” fortrade unions in liberated areas. Of central importance to the Party was the rehabilitationof production in areas under their control in order to have sufficient supplies for theongoing civil war against the GMD. In order to do this, union leadership exhorted theworking class to distinguish between “short term” and “long term” interests, i.e. tosacrifice immediate economic and political advancement for the good of the nation.

That increasing production was the top priority for the union was madeabundantly clear, and cooperation with private capital was to be encouraged. In a tradeunion handbook from the immediate post-revolutionary period, it was explicitly statedthat, “the major criterion of judging a trade union was the production record of itsenterprise.” (Lee 1984:19) Unions were also exhorted to increase their leadership inpromoting “labor competitions,” which were seen as a means to increase theproductiveness and efficiency of workers (Jianguo Yilai Zhonggong Zhongyang Guanyu

Gongren Yundong Wenjian Xuanbian, 1988:379; Lee 1984). Li Lisan wrote a veryimportant article published in People’s Daily on May Day, 1949, in which the union’sposition on the issue was expounded at length. He started by pointing out that ChairmanMao supported the principle of, “developing production, making the economy prosper,caring for both public and private, benefiting labor and capital,” and that there wereconsequences which followed for the appropriate resolution of labor conflicts in privateenterprises: “Hence, ‘developing production and benefits for labor and capital’ should be

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (56)


the starting place for the resolution of all problems within private enterprises.”34 Theadditional argument was made by union leadership that because China was no longer acapitalist country, the character of labor conflicts had necessarily changed (even ifownership was still private) (ibid:407).

When it came to the possibility of worker-controlled production, Li wasunequivocally opposed. He argued that while rural land could be expropriated andoperated collectively, the same was not true for industrial production:

The property of the capitalist is a factory, and a factory cannot be divided. If it isdivided, there can be no production. For instance, if a factory was broken up and given toworkers, you would get one wheel, he would get one driving belt, and all of the machineryand the entire factory would be destroyed. Then will there still be production?… So,workers absolutely must not divide capitalists’ factories, but must do precisely theopposite: work hard to protect the factory and increase production, and this is necessary forworkers to receive a benefit.35

Li went on to address concerns that workers might have, particularly as relates toimmediate workplace issues. However he claimed that, “in this historical period,[exploitation] is impossible to eliminate, it will require patience… The working classneeds to have the spirit of eating bitterness first and enjoying life later, to labor hard andset an example for the entire nation.”36 He even went so far as to claim that the interestsof labor and capital were, as far as an expansion in production was concerned, identical:

If we took all the money that private enterprises earn and used it to improveworkers’ lives, may we ask how we would have capital accumulation to expand productionand develop industry?… an appropriate amount will serve as profit for the capitalist, tomake the capitalist interested in expanding production. So on this point, the interests ofcapital and labor can be identical.37

The inter-relatedness of the Party’s twin goals of increasing production and ethno-national autonomy were neatly summarized by the following line: “…without greatdevelopment of industry, not only will socialism be out of the question, the Chinesenation’s economic independence from imperialism will be out of the question.”38

The contradictory imperatives to increase production and represent workerstended to be resolved in favor of the former. This was the case not only in state-ownedenterprises, but also in private ones as well. As described in the Workers’ Daily in early1951:

In private enterprises, there are a few trade union cadres who not only are not good atgiving consideration to the interests of the working masses, but who even take the place ofthe capitalists in carrying out ‘firing workers and lowering wages’, and who openly speakon the capitalists’ behalf. Because of this, in some trade unions the phenomenon of the

34 Li, Lisan. May 1, 1949. “guanyu fazhan shengchan laozi liangli zhengce de ji dian

shuoming.” [A few explanations on the policy of developing production and benefits for

both labor and capital] Renmin ribao.35 ibid.36 ibid.37 ibid.38 ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (57)


masses’ [sic] not trusting the trade union and the trade union being divorced from themasses has appeared.39

But if union leadership and cadres frequently proclaimed support for the Party-dictated policy of non-confrontation with capital, workers were less accommodating.Bearing an uncanny resemblance to the early 21st century, workers in the immediate post-’49 era often criticized union officials for their, “bureaucratism, isolation from rank andfile, a preference for coercive or commandist methods, failure to trust in the workers,arrogance and high-handedness, formalism, and a lack of democracy within enterprisesand in the trade unions.” (Sheehan, 1988:14) As argued by Paul Harper (1969), workers’expectations were raised by the success of the revolution, and were not necessarilywilling to accept the wage reductions and increased workload that union leadershipfrequently encouraged them to accept.

Owing to this increasing pressure from below, national leadership began to shiftaway from unconditional support for management towards support for worker demands.Starting in late 1950 there was a union rectification campaign that aimed to make unioncadres more accountable to their membership. Even in state owned factories, many cadresargued that the interests of the working class and Party were not identical. And Li Lisanattempted to increase centralization of power within the ACFTU, such that lower leveltrade unions were more vertically integrated and less subject to the power of horizontallyconnected Party units.

It should be noted that Li Lisan’s position on the relationship between the union,Party, and enterprise management is remarkably inconsistent in existing documents, afact which perhaps is indicative of the rapidly and constantly changing politicalenvironment of the time. For instance, in July 1949, at the very same time he was sayingthat the interests of labor and capital were frequently identical, he also argued for greaterunion autonomy: “The management committee is an administrative organization, thetrade union is a mass organization, the Communist Party is a Party organization. Thesethree organizations are independent and none of them can command any of the others.”(Li and Liu 2005:318) This autonomy was necessary precisely because, as argued in June1949, “This contradiction between public and individual interests must be reflected in therelationship between the administration [of a factory] and the union. The location andenvironment of the administration will inevitably lead to it representing public interestsmore, and it is very difficult for them to be concerned with every person’s day-to-dayinterests.” (ibid:317) And yet, in March 1951 he would directly contradict this earlierstatement: “the administration and union are part of an organic whole, and we should beclear that the union is for helping the administration complete tasks… the interests of theadministration and the union are basically identical, and there are not any contradictionsat all. If there are some contradictions, it is just the result of bad temperaments.”(ibid:318) The relevant point is that there was a constant back and forth in the early yearsof the PRC as the ACFTU was frequently tugged in opposite directions by membershipand the Party.

Official pronouncements aside, by the end of 1951 the Party had decided that Lihad pushed too far with constructing greater union autonomy. In December of that year,the ACFTU’s Party organization held a meeting at which they passed a resolution 39 January 1, 1951. Workers’ Daily. (quoted in Sheehan, 1998:29)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (58)


condemning him for “errors of principle,” including economism, formalism, andsyndicalism. Li was forced to step down from his position in early 1952, and a broadcritique of the former chair spread throughout the union structure.

The fear of being singled out for excessive focus on the particular interests ofworkers once again put union cadres in a position where they were squeezed betweenworkers with increased expectations and the Party which was focused on increasingproduction. This led to more distrust between workers and their representatives:

From this [critique of Li] union officers were afraid of committing economism orsyndicalism and so they dared not speak out on behalf of the interests of workers. Thiscaused the union to become lifeless, and to occupy a non-essential position, leading to theproblems of the Party committee handling everything and separation from the masses.(Tang, 1989:191)

This environment led to a situation referred to in official jargon as “separationfrom the masses,” (tuoli qunzhong), which became increasingly apparent with an uptickin strike activity among workers in the first half of 1953. In July of that year, the ACFTUsubmitted a report to the Party’s Central Committee in which they detailed at length theseproblems. In particular, the report was concerned that, “These [strikes] reflect that theunion cannot connect with the masses, and that union cadres do not have the trust of themasses.” (Jianguo Yilai Zhonggong Zhongyang Guanyu Gongren Yundong Wenjian

Xuanbian, 1988:174) Just as this worker activism was expanding, the ACFTU’s SeventhCongress in May 1953 was used to firmly re-establish Party control over the unionorganization (Sheehan, 1989:13).

Despite unequivocal subordination of the union to the Party, from 1953 to 1956the state seesawed between a focus on trying to increase workers’ participation infactories on the one hand and unremitting pursuit of production on the other. The “threeanti and five anti” movements attempted to address problems among the state/unionbureaucracies and the capitalist class, respectively. While autonomy from the Party wasstill forbidden, union cadres were encouraged to mobilize the masses for greater say inthe functioning of the enterprises. This was thought to be necessary in private enterprisesin order to eventually make the transition to full nationalization. However, by 1955, thestate had experienced enough of mass mobilization and once again returned to anexclusive focus on increasing production. This included a strong reassertion ofmanagerial prerogative, particularly in SOEs.

Chinese workers, now deemed the “masters of the enterprise” in the officialrhetoric, were not pleased with the new Soviet-inspired centralization of authority withinthe workplace. This increased focus on worker discipline and efficiency was a majorfactor in the emergence of a strike wave in the country starting in the latter half of 1956.This series of strikes, though not coordinated by any national-level organization, affectedmajor swaths of Chinese industry. The unrest was particularly acute in Shanghai in early1957, at which time there was more widespread labor unrest than even in therevolutionary years of the Republican period (Perry 1994). Having coincided with anoutbreak of student strikes, the Central Government became quite concerned that anuprising such as had taken place in Hungary that same year could develop in China.

In March 1957, the Party’s Central Committee issued a “Directive on Handling ofStudent and Worker Strikes.” The document placed blame for the social disturbancessquarely on the shoulders of state officials: “These [worker and student strikes] have

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (59)


occurred primarily because of we have not done our work well, and especially because ofbureaucratism by officials.” (Jianguo Yilai Zhonggong Zhongyang Guanyu Gongren

Yundong Wenjian Xuanbian, 1988:508). If bureaucratism was the cause of the strikes,then more democracy in the workplace (and schools) was seen by the Party as theantidote: “In order to prevent the appearance of student and worker strikes… the mostimportant [method] is to overcome bureaucratism and expand democracy.” (ibid:509)While there were certainly no calls for union independence from the Party, once again,the pendulum appeared to be swinging back towards greater accountability for officials.

Not coincidentally, the Party launched the Hundred Flowers Movement in early1957, at which time various sectors of society were invited to openly criticize the Partyand other officials. Certain elements within the ACFTU took quite liberal positions,likely in response to the upsurge in independent worker organizing which hadmaterialized with the strike wave. Chairman Lai Ruoyu, originally brought in during1952 because of his lack of experience within the trade union (and therefore supposedallegiance to the Party), began arguing that the union should side with workers under allcirc*mstances. This of course was a marked change from the previous rhetoric of unionsworking hand in glove with factory administration. Gao Yuan, another high level officialin the ACFTU went as far as to say that workers should be allowed to set up independentorganizations. This was a truly exceptional moment in ACFTU history.

But just as before, this loosening at the top was quickly met with a harsh responsefrom the Party, this time in the form of the Anti-Rightist Movement, launched in thesummer of 1957. While much of the literature on the Hundred Flowers Movement and itsaftermath has focused on intellectuals (Goldman 1962; MacFarquhar 1960) workers andsupportive union officials were also active participants in the outpouring of criticism ofthe Party that emerged in this period. Chairman Lai Ruoyu was eventually sacked for hissyndicalist tendencies, and other union leaders throughout the organization were purged.Once again, insurgent and autonomous worker activity had pushed union leadershiptowards greater liberalization, only to be ultimately foiled by a nervous Party leadership.

This awkward balancing act between trying to maintain legitimacy with insurgentworkers while not disrupting the relationship with the Party has been a definingcharacteristic of the ACFTU in the People’s Republic. During the course of the CulturalRevolution, the dynamics changed somewhat, as popular mobilization became officiallysanctioned from the top, and all sorts of autonomous worker organization becameprevalent. Indeed, the ACFTU ceased to function for all intensive purposes during thistime period, failing to hold any national congresses. Given the peculiarity of labororganization in the Cultural Revolution, I will not provide details on this time period.However, it will be worth mentioning one last major struggle for greater unionindependence that was once again crushed: that which took place in the spring of 1989.

The ACFTU in 1989

At the ACFTU Congress in 1978, the first one held since the 1960s, DengXiaoping blamed Lin Biao and the Gang of Four for the “paralysis” suffered by the unionfederation during the Cultural Revolution (Wang, 1990:3)(Wang 1990). Furthermore,class struggle was off the table, and the union’s role was to encourage market-orientedreforms:

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (60)


The Center believes that various union federations will deeply enter the masses todo propaganda work so that enterprises can smoothly enact these reforms. This will be forthe interest of socialism and the interest of the Four Modernizations, and the working classwill play a vanguard function of considering the public and denying the individual. (ibid:5)

The union federation was being resurrected completely on the Party’s terms, quiteexplicitly to serve goals which were understood as not being in the “short-term” interestsof the working class.

As before, the loosening of the political environment in the late 1980s causedsome in the ACFTU to begin to take limited, yet significant, steps toward greateraccountability to their membership. The clearest example of this tendency was in the“Basic Plan for Union Reform,” a resolution passed by the ACFTU executive committeein October 1988. This document said that previously, “leftist” thinking had incorrectlyled to union leaders being accused of economism and syndicalism. The assessment in1988 was both that the working class had distinct interests and that this would requiregreater independence for the unions:

For a long time, there was an excessive emphasis on the unity of social interestsand high levels of centralization of leadership, with no distinction between the functions ofParty, state, and social organizations. This caused the union to be unable to embody its roleas a mass organization and did not function according to the role it is supposed to have. Inreality, the union became a work department of the Party committee or a subordinate organof the administration. (Li and Chen, 1989:325)

As far as the relationship to the Party was concerned, the resolution said that horizontallylocated Party committees should, “respect the organizational independence of the unionand the democratic system of the union, and support the local union in independently,autonomously, and creatively carrying out its work.” (ibid:328) Additionally, it went onto argue that the existing system of appointing, rather than electing, cadres should bereformed.

While it is not difficult to find official boilerplate about promoting democracythroughout the history of the ACFTU, the rhetoric in late 1988 and early 1989 wasnoticeably more strident. In this sense it echoed previous calls for reform from the 1950s.When students and workers began taking to the streets en masse in Beijing and othercities in April and May of 1989, it is likely that an internal debate within the unionappeared, much as was taking place in the Party. But it is unlikely that union officialswere prepared for the very direct threat they faced as represented by the formation ofindependent worker organizations.

Although the student activists were the main focus of the media and subsequentscholarly work on the 1989 protests (Gold 1990; Zhao 1998; 2001) workers played a keyrole, particularly in Beijing. As the movement in that city grew in late April, dozens ofworker activists began meeting up in Tiananmen Square to discuss politics and potentialforms of organization. Understanding that their own participation was riskier than it wasfor the students, the existence of the newly formed Beijing Workers’ AutonomousFederation was not made public until student hunger strikers took over the square on May13th (Walder and Gong 1993:6) Staking out a much more radical position than most ofthe students, the workers’ organization openly rejected the leadership of the Party andofficial unions, and called for direct worker supervision of production. Unsurprisingly,

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (61)


workers eventually suffered more severe state violence and repression than was the casefor the students.

How did the ACFTU respond to this political crisis? To some extent, the workerprotest enhanced the position of union officials as they officially represent the interests ofthe working class within the state bureaucracy. Anita Chan relates a story from thisperiod in which the mayor and deputy mayor of Shanghai went to the union offices (anunusual reversal of protocol) and that union officials successfully put forth a number ofproposals for improving workers’ conditions. Chan concludes that, “[Trade unions] wereexpanding their corporatist power by laying claims to representation of a restive socialforce.” (Chan 1993:56) Back in Beijing, ACFTU officials had occasionally expressedsupport for the students, even going so far as to make a large donation to support hungerstrikers in the square (Walder 1991:485). Of course there were limits to what waspossible: when members of the Beijing Workers Autonomous Federation asked theACFTU for help in formally registering, they were rejected.

Any sympathy union leaders may have felt for the protestors was quickly brushedback under the rug with the incursion of the military and subsequent massacre on June 3-4. With the hardliners acting decisively to end the standoff, there was no more questionbut that political liberalization was off the table. Reformists within the ACFTU got themessage, and talk of independent worker organization or even a more assertive role forthe official unions ended abruptly.

If the state’s actions in the street were not clear enough, in December of 1989 theParty released a wide ranging documented entitled, “Notice on Strengthening andImproving Party Leadership of the Trade Union, Youth League, and Women’sFederation.” The Party’s deep concern about the emergence of independent tradeunionism is quite apparent:

Currently, the tasks of rectifying governance and deepening reform are incrediblyformidable. Foreign and domestic enemy forces have used our temporary difficulties to tryto change our forward direction… Party organizations at various levels must enforceunified leadership over the trade union, youth league, and women’s federation according tothe Party line, such that these organizations will maintain the correct political direction.(Wang 2002:196)

Of particular importance for union activities in the reform era was the strong re-assertionof the authority of Party organizations. Unions were warned to, “Guard against andprevent any tendencies toward throwing off or weakening the leadership of the Party,guard against and prevent some people with ulterior motives from destroying stabilityand unity.” (ibid:197) Additionally, there was a rhetorical shift away from recognizingthe “particular” interests of the working class, and back towards consideration of the“national interests.” In an October 1988 address to the ACFTU’s National Congress,Party Secretary Zhao Ziyang had argued, “In the past, the union’s protection of workers’specific interests has been overlooked, which has brought a negative influence to unionconstruction which should not exist.” (Li and Chen 1988:295) But in the aftermath ofJune 4th, the tone from the Party was markedly different:

At the same time [as protecting particular interests] in the practical operationsworkers, youth, and women should be lead to self-consciously subordinate individualinterests to the nation’s interests, subordinate particular interests to general interests,subordinate short-term interests to long-term interests. (Wang 2002:199)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (62)


Additionally, the Party’s single-minded pursuit of economic growth was re-affirmed as acommon goal for the union:

In our current period of governance rectification and deepening of reform, [tradeunions] should educate workers to understand and make allowances for the nation’sdifficulty, and work hard to increase production and economize, increase income andreduce expenses, and continuously increase economic benefits. (ibid:201)

Following the release of this Party notice, ACFTU leadership distributed an“advice” throughout the trade union structure on how to interpret the new marchingorders. Of paramount importance was their role as first line of defense in crushingindependent worker organization:

When worker organizations are discovered that oppose the four basic principles40

and that harm state power, unions must immediately inform same-level Party organizationsand higher levels of the trade union. Additionally, they must be resolutely exposed anddisintegrated, and the government should be asked to ban them according to law. As forspontaneous worker organizations that arise for reasons of economic interests, trade unionsshould convince and lead them to disperse and stop their activities. (ibid:251)

In addition to re-affirming the commitment to oppose independent organization, ACFTUleadership nearly went so far as to claim that workers’ particular interests were irrelevant:

When the union protects the legal rights and interests of workers and participatesin interest mediation, it must begin with benefiting social stability and developing theproductive forces. [Unions] must conduct their work according to the law and relevantpolicies, and actively lead workers to self-consciously subordinate individual interests tonational interests, subordinate partial interests to general interests, subordinate short-terminterests to long-term interests. (ibid:254)

In the space of less than one year, ACFTU leadership had gone from openly calling forgreater reform and for the promotion of distinct interests of the working class to fullsubmission to the authority of the Party and a renewed focus on increasing economicgrowth at all costs.

The fallout from 1989 was to have profound implications for the activities of theACFTU and its subordinate unions as marketization in China accelerated over the nexttwo decades. When the state determined that pursuit of “national interests” requiredreform of the state owned sector, unions stood to the side as innumerable millions ofworkers were laid off (Solinger 2001), had their pensions stolen by corrupt officials, orwere forced to accept lower wages, longer hours, and reduced job security. At the sametime, a new class of migrant workers was taking shape in the export zones of thesoutheast. Although subject to high rates of exploitation, long hours, dangerous workingconditions, etc., the official union structure had essentially no presence in these privatelyowned factories. With a strengthening of the alliance between state and capital at the

40 The four basic principles were established by Deng Xiaoping in the wake of the 1978

Democracy Wall movement. The principles are to, 1) uphold socialism; 2) uphold the

people’s democratic dictatorship; 3) uphold the leadership of the Communist Party; 4)

uphold Marxism-Leninism-Mao Zedong Thought.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (63)


local level, unions – even those that may have wanted to take a more aggressive position– were left with few means for advancing worker interests.

Although there had been a number of attempts by union officials to enhanceorganizational autonomy in the wake of expanded labor conflict and protest, each effortended with the Party re-asserting its dominance. In this sense, we can see the clearorganizational consequences of having representative power arrogated to the ACFTU bythe state rather than through delegation by membership. The salient point is that in thesocialist period and up to the present, ACFTU subordinate unions have – at theinstitutional level – operated with an extremely high level of oligarchy (as defined in theprevious chapter) by continually excluding workers from participation. Unionheteronomy combined with the Party’s single-minded pursuit of economic developmentresulted in a profound rupture between represented and representers in the socialisteconomy. In a sense, there was never a rupture at all between the union and the newworking class in the private sector, since these workers had never been organized andmobilized. But how precisely is this oligarchic union par excellence structured today?We now turn to a more detailed explanation of the organizational form of the Chinesetrade union system.

The Structure of the Trade Union System in Contemporary China

Unions in China are officially referred to as “mass organizations,” putting them inthe same category as the Youth League and Women’s Federation. Officially, the functionof these organizations is to link the Party to various social groupings in society (workers,women, and youth, respectively). Following the Leninst conception, the union structure isconceived of as a “transmission belt” to carry information back and forth between theParty and workers (Unger and Chan, 1995:37). The ACFTU claimed a membership of226 million in 2010,41 which if true would make it the largest national union federation inthe world by an extremely large margin. In this section, we will see quite clearly how theunion’s only source of legitimacy derives from the state, often times at the expense of therelationship to workers.

National Organization and Relationship to the Party

At every level of the union hierarchy, union organizations are subject to “dualcontrol.” The first and primary form of control comes from the horizontally located Partyorganization, if one exists. Party organizations do exist for all regionally based unionfederations, in state-owned enterprises, and in many large private enterprises. Secondly,union organizations are subordinate to the immediately superior trade union organization.The primacy of Party control was strongly re-asserted after the 1989 political crisis, whenit was announced that unions at various levels could “listen to the advice” of

41 Both the accuracy and the meaning of this number must be interrogated closely. While

no quantitative data exist, it is very likely that an incredibly large number of these people

would not identify themselves as a member of a union and do not pay union dues. A

somewhat smaller percentage would not have any idea what a union is.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (64)


hierarchically superior union bodies (Wang 2002:198), but that final decisions must bemade by the Party organization.42 Such an arrangement has not been questioned since,and for a union official to do so publicly would certainly lead to severe consequences.

The original logic behind such an arrangement was that with the successfulrealization of the dictatorship of the proletariat, independent labor organizations such asexist in capitalist societies were no longer necessary. Since the Party represents theinterests of the nation as a whole, there can be no antagonism between the working classand the Party. As was argued in each successive moment when unions attempted greaterautonomy, confrontational independent worker struggles run counter to the national andlong-term interest.

Even if trade union leaders in China will not frequently use such Marxist-inflectedlanguage anymore, they are quick to point out that what they lack in independence, theymake up for in access to state power. When visiting American unionists beam with envyover the rapidly burgeoning membership of the ACFTU, their Chinese counterpartsfrequently take the opportunity to describe how the Party can bring enormous pressure tobear on recalcitrant employers. Perhaps most importantly, the union at various levels hasdirect access to the state’s legislative bodies. When labor laws are being considered, it isnecessary protocol to consult with the unions, in not an entirely pro-forma way. The mostprominent recent example of the ACFTU mobilizing its political strength in legislationwas the passage of the 2008 Labor Contract Law, supposedly over the objections of themore conservative Ministry of Human Resources and Social Security.

But even if unions claim high levels of access to state power, their relative statuswithin the state is still quite low. As a “mass organization,” the union does not wield thesame power as a government ministry. Although union cadres claim that their status hasbeen in the rise over the past decade, the union system has typically not even outrankedthe Youth League, the organization from which Hu Jintao originally hails. To the extentthat their power has increased in recent years, union officials are explicit that this isbecause the problem of labor unrest has become more acute.

Although not an official government agency, ACFTU-subordinate unions are partof the state, a fact that must not be overlooked. Nowhere is this fact more apparent thanin the method of cadre recruitment and leadership selection.

Union Leadership

The most significant way that the Party keeps control over the union (and notcoincidentally, alienated from workers) is through tight control over union leadership.Although formally subjected to “internal democracy,” the selection of union federationchairs at all levels is a fully non-transparent procedure,43 and is almost certainly dictatedby the Party. In order to win any position of power within the union system, an officialmust be a Party member and must have a track record of allegiance to Party-determinedgoals. While there are exceptions, the rule is for high-level leaders to not be drawn fromwithin the ranks of cadres who have worked their way up through the union, but rather to

42 See below for a more detailed description of union structures.43 There is greater variation in the selection of union chairs at the enterprise level, which

is discussed below.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (65)


rotate them in from other mass organizations or government agencies. This is of course toprevent leadership from developing broad-based support from within the ranks of theunion, a scenario which could potentially lead to a power struggle at the top. Because ofthis policy, union leadership at every level is likely to have much stronger allegiance tothose who gave them the position rather than to their own membership.

This frequently makes for career trajectories for unions leadership that differssignificantly from those in most other countries, as it is not at all unusual for people withno experience whatsoever in trade unionism to be appointed to very high-level positions.For instance, prior to becoming ACFTU chair in 2002, Wang Zhaoguo’s longest officialappointment was in the Taiwan Affairs Department, where he served from 1990 to 1997.Wang Yupu, appointed first ACFTU vice-chair and Party organization secretary in 2010,was an executive and Party committee member at the Daqing Oilfield Company from1999 until 2003, after which he served as CEO and general manager until 2009.44

Another ACFTU vice-chair, Xu Deming (the man personally responsible for bargainingthe first collective contract in a Chinese Wal-Mart) previously served as the director ofLiaoning province’s Mines and Geology Department. Upon leaving the ACFTU in 2008,Xu was appointed director of the State Bureau of Surveying and Mapping.45 Theunsurprising consequence of such a career structure within the union is that leadership isfrequently unfamiliar with, and often times uninterested in, labor issues.

Another reason that union leadership is often unconcerned with their ostensiblerole as representative of the working class is the system of joint appointments. Unionchairs of regional federations, from the municipal level all the way to the national, havein recent years been given joint appointments as the vice-director of the standingcommittee of the relevant People’s Congress (the legislative body of the Chinese state).So for instance, Chen Weiguang is both the chairman of the Guangzhou Federation ofTrade Unions and also the vice-director of the standing committee of the GuangzhouMunicipal People’s Congress. The logic behind such an arrangement is that the tradeunion will have a greater say in legislative affairs at all levels of the state. While this iscertainly true to some extent, it also means that many trade union chairs are much moreconcerned with their position in the People’s Congress, and are consequently less fullyengaged in union work.

In addition to these steps aimed at controlling union leadership, the state has alsotaken extensive measures to ensure the allegiance of regular union cadres. While it is notrequired that lower level officers be Party members, it is certainly welcomed (and isnecessary for long-term career advancement). Union cadres must pass the same civilservice exam that other government officers take. In the 1990s the state changed the rulesdictating pay and benefits for union cadres, such that they are the same as governmentofficials at an equal level of the hierarchy. Additionally, unions have recently beenpursuing “professionalization,” which implies recruiting more university graduates ratherthan recruiting internally.

44 In 2002, the Daqing oilfields were rocked by huge protests by workers angered over the

theft of their pensions. While it is unclear what precisely Wang’s role may have been in

this crisis, as a high level executive and Party committee member, he was certainly

involved in some manner.45 Biographical information comes from

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (66)


While some have argued that union cadres are thus subjected to a “doubleidentity” (Chen 2003), i.e. as both worker representatives and agents of the state, myresearch supports Raymond Lau’s contention that these officials think of themselves as,and behave like, government officials (Lau 2003). Lau relies on the Bourdieusian conceptof habitus, which is defined as, “not only a structuring structure, which organizespractices and the perception of practices, but also a structured structure.” (Bourdieu1984:170). The habitus of union officials is structured by the same set of institutionalparameters as those of government officials, and it therefore generates (i.e. structures)similar social practices. That union officials have a state habitus is of course a difficultproposition to “prove,” but there is much evidence to support such a claim. As has beendiscussed, they undergo the process of selection and socialization that is entailed bygaining Party membership. They are subject to the direct and unquestioned dominationby horizontally located Party organizations, and higher-level union organizations. Manyof them have extensive formal experience as government and Party officials. And from apractical standpoint, they have much in common with formal agents of the state. Whetherit is their diction and reliance on official slogans, hosting lavish dinners, beingchauffeured around in black luxury cars with tinted windows,46 the stiffness andformality of interaction, professed enthusiasm for national events (e.g. Olympics, WorldExpo), etc., the practices of union officials are clearly structured by the same habitus asthat of government officials.

But where this practical alignment is most obvious, and perhaps mostproblematic, is in dealing with labor conflicts. Union officers’ first response to strikes isnever to rush to the side of workers and strategize on building collective power to forceconcessions from capital. Their response is that of an agent of the state: intervene,“rationally” encourage dialogue, convince the workers to make “reasonable” demands(i.e. lower their demands), and perhaps try to persuade (through non-coercive means)management to meet some of the workers’ demands. Until major changes in 2010(discussed in Chapter six) union officials’ primary concern when discussing strikes hadalways been how to avoid them altogether, rather than seeing them as a key tactic inadvancing working class interests. In general, union officers believe that the best way toavoid labor conflict (which is their wish) is through legislation and administration, butnot through organizing workers. The parallels with the state’s “logic of practice” areapparent enough.

Regional, Industrial, Sectoral, and Enterprise Unions

The primary form of organization within the ACFTU-controlled hierarchy is theregional federation. These federations officially represent all of the other unionorganizations (industrial, sectoral, enterprise) within their given jurisdiction. At the apexof this hierarchy is the ACFTU, followed by union federations at the provincial,municipal, and depending on particular jurisdictional arrangements, district (qu), county(xian), township (zhen), and street (jiedao) levels. They participate in labor-related

46 The vehicles of choice for government officials are somewhat dependent on city, but

include Audis, Buicks, large Toyotas, etc. But it is the black paint and tinted windows

that are the defining characteristics.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (67)


legislation, are first-responders to severe labor conflicts, and – particularly at themunicipal level – experiment with new forms of union organization and coordinateunionization drives. Regionally-based federations have attained the most power withinthe union hierarchy, in part because they strictly mirror the organization of the Party (andtherefore are seen as relatively safe).

Parallel to this geographically-based structure is the system of industrial unions.There are ten industrial unions within the ACFTU, including the Educational,S60c60i60e60n60t60i60f60i60c60,6060 60C60u60l60t60u60r60a60l60,6060H60e60a60l60t60h60 60a60n60d60 60S60p60o60r60t60s6060W60o60r60k60e60r60s’ Union; Seamen and Construction Workers’ Union; Energyand Chemical Workers’ Union; Machinery, Metallurgical, and Building MaterialWorkers’ Union; Defense Industry, Postal and Telecommunications Workers’ Union;Financial, Commercial, Light Industry, Textile and Tobacco Workers’ Union;Agricultural Forestry and Water Conservancy Workers’ Union; Railway Workers’Union; Aviation Workers’ Union; and Financial Workers’ Union.47 These unions aresubject to dual control, first by the horizontally located geographically based unionfederation, and secondly by the immediately superior entity within the industrial union. Inaddition to the national-level organizations, they are generally established at theprovincial and municipal levels, but not below. These unions are relatively weak, andindeed were specifically undermined by the Party in the 1950s for fear that they wouldseek greater autonomy. They do not sign collective contracts on behalf of workers, orseek to organize new members, but rather engage in legislative activities, consultationwith government agencies, and interactions with large companies in the relevantindustries.

In recent years, “sectoral unions” (hangye gonghui) have become increasinglywidespread in China. These unions, generally organized by municipal level unionfederations, seek to organize all of the employees in a given industry (e.g. construction,shoes, eyeglasses, hotels, etc.) within the municipality. The aim of such an organizationvaries somewhat depending on the industry, but the general intention is to attempt to setsome standards that apply to all employers within the specific geographic region. Themost common practice is to try to establish unified pay rates, although we will see thatthere are immense challenges in actually implementing sectoral wage agreements. Theability to represent workers in signing collective contracts is what distinguishes themmost clearly from the industrial unions. Additionally, they are not directly integrated intonational-level hierarchies. Sectoral unions are generally staffed, financed, and controlledby municipal union federations. For a variety of reasons which will be discussed in laterchapters, this organizational form has become heavily promoted by many different levelsof the union structure.

Finally, there are the “grassroots” (jiceng) or enterprise-level unions. It is withinthese organizations that workers directly encounter the union. The enterprise-level unionactivities include collecting dues, conducting various forms of entertainment activities(including birthday and holiday parties, field trips, etc.), distributing gifts of cooking oil,

47 Some readers may notice that there are two industrial unions responsible for financial

workers. Because I did not undertake a study of these unions I cannot explain this

redundancy, this is how the industrial unions are listed in official ACFTU literature.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (68)


mooncakes, and sometimes cash at holidays, running “sending warmth” activities(providing assistance to workers facing severe economic difficulties), mediating labordisputes, and perhaps most importantly, representing workers in collective negotiations.In practice, many (if not all) of these activities never take place, and the enterprise unionexists on paper only. But in many industries in China, a workplace-based system ofindustrial relations is emerging, so these enterprise-level unions continue to be of greatpotential importance.

Relationship to Management

Even if union federations have gained better access to state power at the nationaland regional level, union organizations in the workplace remain incredibly weak andincapable of ensuring even the most basic legal enforcement (Chen 2009). If theautonomy of higher-level unions is deeply circ*mscribed by the Party, at the enterpriselevel it is management that dominates the union. In most cases, management’s control ofthe enterprise union is completely transparent, as it is quite common to have humanresource managers or relatives of the enterprise owner serve as union chair. For example,in 2009 the Guangzhou Federation of Trade Unions (GZFTU) revealed that more than50% of union chairs in private enterprises were from management.48 This wasparticularly alarming because the announcement came more than a year after Guangzhouhad passed a new regulation specifically banning the practice. A 2004 nationwide surveyby Hishida, et. al of 1,811 enterprise level trade union chairs revealed even moretroubling trends. Most immediately significant was the finding that only 3.7% of thosesurveyed listed “staff and worker” as the position held immediately prior to assuming thepost of union chair (Hishida, Kojima, Ishii and Qiao 2010:121). But even in the unlikelyevent that a union chair is elected (or appointed) who is interested in confrontingmanagement, their ability to do so is highly constrained by organizational factors. Unionsare primarily financed by a 2% payroll tax which is paid by the enterprise, rather thanbeing supported by membership dues. Most importantly, the wages for enterprise unionchairs are paid directly by management. There are countless examples of activist unionchairs being summarily fired for antagonizing management, with few (if any)repercussions for such retaliatory behavior. The consequence is that enterprise-levelunions’ capacity to effectively represent their membership in collective negotiation andother spheres is severely curtailed (Clarke, Lee, and Li 2004).

There have been some exceptional cases in which workers have been allowed todirectly elect their own union chairs at the enterprise level. In at least two cases, suchelections came about as a result of pressure from foreign brands (Chan 2008), while inothers, activists had the support of higher levels of the trade union (Chan 2007). TheACFTU has been interested in conducting experiments with direct elections in variousplaces around the country, but so far there is little evidence of long-term success (Howell2008). As we will see subsequently, in the rare instances that an activist union chair is

48 July 24, 2009. “guangzhou san nian nei jiejue qiye fuzeren jianren gonghui zhuxi

wenti.” [Guangzhou will resolve the problem of enterprise managers have joint positions

as union chairs within three years]. Guangzhou Ribao. Of course, the real number is

likely much higher, but this is what was reported publicly.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (69)


elected, they continually are hamstrung by threats of retaliation from management andacquiescent higher levels of the unions. Thus, while much of the criticism of Chineseunions focuses on their lack of independence from the Party-state, lack of independencefrom capital is an equally vexing problem.


In this chapter we have traced the trajectory of the ACFTU from its revolutionaryheyday in the 1920s up to the present. The common theme uniting each of the seeminglyvery different iterations of Communist trade unionism has been the inter-related goals ofsecuring ethno-national autonomy and expanding production. During the general strikesin 1925-1927, unions mobilized the working class in an attempt to resist foreignimperialism and to advance the national revolution. While class-based demands werecertainly included at this time, they were secondary to anti-imperialism. During theMaoist era, union leadership occasionally struggled for greater autonomy, but each timewas defeated and Party dominance was strongly re-asserted. Operating under the theorythat socialization of the relations of production would result in a liberation of the forcesof production, labor in this period came to be highly decommodified as industry wasnationalized and the “iron rice bowl” was institutionalized. As market reforms began inthe 1980s, there was a push from within the ACFTU for organizational reform; howeverthis was quickly crushed in the aftermath of the Tiananmen Square massacre.

Throughout the course of development of the ACFTU, there have been a series ofshort-lived experiments with expanding internal democracy. Since the end of the 1920sthe ACFTU has collapsed and been resurrected twice – once in 1948 following the civilwar, and once in 1978 following the Cultural Revolution. These resurrected versions ofthe ACFTU bared only the faintest resemblance to the revolutionary organization of the1920s, as they were (and still are) entirely creations of the Party. These two instances ofresurrection where key in institutionalizing appropriated representation for the workingclass and in stamping out forms of representation based on membership delegation. As aresult, the extent to which the unions have been able to engage in serious democraticpractice has been severely circ*mscribed since the establishment of the People’sRepublic. This oligarchy is most evident in ends formulation; the ultimate ends of actionfor the union have been determined heteronomously, i.e. workers have not been seriouslyengaged in the question, “what should the union do?” The means employed by the unionwere quite democratic in the 1920s, with high levels of worker involvement. But since1948, there have been only limited examples of the union directly involving membershipin trying to secure pre-determined ends. The Party viewed the few attempts for amodicum of democracy within the union in 1951, 1957, and 1989 as serious threats to itsmonopoly on political power and were therefore crushed.

Then there is the question of pursuit of analytically-determined ends – i.e. has theunion promoted decommodification for the working class? It is beyond a doubt thatduring the post-revolutionary era labor in China became significantly less commodified.But it is not clear at all that the union as representative of the working class played adecisive role in this process. As is evidenced by the purges of Li Lisan and Ruo Laiyu,the Party was unwilling to accept union autonomy; thus it appears that it was state power,not the collective power of worker organization, that was the decisive factor in state-

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (70)


socialist decommodification. This expansion of decommodification was anchored in thedanwei system of workplace organization, which served to incorporate urban workersinto the state. However, the union’s subordination to the Party in terms of endsformulation meant that when the transition to capitalism began in the 1980s, there washardly any possibility for organized resistance to the incursion of the market. And indeed,during the process of “smashing the iron rice bowl,” unions were completelyunable/unwilling to develop the power of their membership to counter this attack on theirlivelihood. Thus, while conditions of appropriated representation may not have been sucha severe problem for the working class in an era in which decommodification andnationalization of industry were being backed up by the coercive power of the state, in aperiod when the state became increasingly committed to capitalist development itresulted in a political crisis.

Finally, we can see how the post-1948 unions came to function as a part of thestate, as a factor of cohesion. The Party has continually exerted efforts at maintainingtight control over the union, particular when it comes to organizational structure and theselection of leadership and cadres. Unions have come to behave as mediators, rather thaninitiators, of labor conflicts. In particular, they have always devoted great efforts towardssymbolically constituting worker demands as purely individual and economic, rather thancollective and political, in nature; this is what is meant by “isolation effect.” The ACFTUhas assumed responsibility for dealing with the problem of labor conflict, and through itspromotion of “harmonious labor relations,” it has attempted to maintain the unequalequilibrium of compromise necessary for continual expansion of productive capacity.However, when we turn to the case of contemporary Guangzhou, we will see that even inthe most likely case, the unions’ ability to realize compromise between labor and capitalis extremely limited.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (71)


Chapter 4

The Most Progressive Union in China

We now shift our focus from the historical development of the ACFTU to itscurrent activities in Guangdong’s capital of Guangzhou. This metropolis of over tenmillion49 anchors the Pearl River Delta, the most significant manufacturing center inChina, and indeed the world. There are a number of good reasons to believe that this citywould be most likely to experience the emergence of the institutional moment of thecountermovement: its relatively long experience with marketization, high levels ofworker insurgency, supposed openness to international cultural and intellectual currents,and most importantly, the progressiveness of the leadership of the municipal unionfederation, the Guangzhou Federation of Trade Unions (GZFTU).

By 1980, three of China’s four Special Economic Zones (SEZ) were located inGuangdong province. Guangzhou itself was subsequently designated as a “Coastal OpenCity” in 1984, allowing for the large inflows of foreign direct investment (FDI) andprivate enterprise that emerged a few years earlier in the SEZs. Guangdong quicklybecame the country’s premier destination for FDI (Chen, Chang, and Zhang 1995:694)and private enterprise more broadly. As many Chinese trade unionists now openlyproclaim, the old model of regulating labor relations through the “iron rice bowl” anddanwei system is inoperable in private enterprises where the workforce is made uplargely of migrant labor (Solinger 1995). In other words, one might reasonably expectthere to be great impetus for unions to play a role in representing workers in places wherecommodification was the most advanced.

It has been a key argument thus far that unions and states do not pursuedecommodification and incorporation of labor out of any necessary or mechanicalresponse to commodification, but rather because continual accumulation is threatened byworker insurgency. And just as one would expect from Polanyian theory, as the placewith the highest levels of capitalist industrialization, Guangdong has also experienced thehighest levels of worker resistance (Leung and Pun 2009; Pun, Chan, and Chan 2009).The province accounted for more than 17% of all labor disputes in the country in 2009

49 (accessed January 18, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (72)


with 118,155, well above the 74,637 cases in second-place Jiangsu.50 This instability inlabor relations is grasped subjectively by union and government officials who are pre-occupied with reducing such conflicts. Thus we see that the insurgent moment of thecountermovement is pushing union officials to respond with the existing institutionalmachinery they have at their disposal.

Finally, we might expect that unions in Guangzhou might be more pro-laborbecause of the unusual political disposition of the city’s pre-eminent labor leader, ChenWeiguang. There is significant literature on unions both in China (Liu 2009) and in theWest (Stepan-Norris and Zeitlin 1991; 1995; 2002) which suggests that the particularqualities and political commitments of leadership can have a significant impact onoutcomes.51 What’s more, there is reason to believe that in China, specific leadershipqualities could have an even greater impact than in liberal democracies. Despite theChinese state’s propagation of “rule of law” (fazhi) in the post-Mao era (Keith 1991;1994; Lubman 1999), it is popularly believed that China remains a country “ruled bypeople” (renzhi). That is to say, laws are not universally applied by an impartial andobjective judiciary, but rather their proper application is adjudicated in each particularinstance by the “cultivated man” (Weber 1946:243).52 To the extent that administrativedecisions in the union are not being made according to Weberian principles ofinstrumental rationality, there is the possibility for an increase in the autonomy of theindividual officer (though of course the flipside to this is that any given officer cannotcount on their subordinates to rationally carry out orders). In any event, GZFTU chairChen Weiguang is a truly exceptional character within the ACFTU-controlled unionstructure, and his outspoken pro-worker positions have helped him gain notoriety withinChina as well as internationally. If the political outlook of leadership can in fact have asignificant impact on unions playing a role in institutionalizing decommodification andincorporation of labor in China, then the workplaces within Chen Weiguang’s jurisdictionwould be the most likely place to witness such a phenomenon.

In this chapter we will begin with an extended introduction to Chen Weiguang.We will see that his rise to power is in itself an indication of tentative first steps towardclass compromise, as are the series of pro-worker policies that have been adopted underhis guidance. However, when we analyze labor conditions in the enterprise withsupposedly the best union in Guangzhou, Hitachi Elevator, we find that the limited levelsof decommodification that have been secured came as a result of state action rather thanfrom the union acting as worker representative. Oligarchy within the union prevents moreencompassing decommodification, as is most clearly evidenced by the continualprevalence of highly contingent intern-workers. Since workers were not involved at any

50 2010. Zhongguo laodong tongji nianjian. Beijing: Zhongguo tongji chubanshe.51 Additionally, some literature suggests that some combination of “bottom-up” and “top-down” mobilization can be most effective (Milkman 2006).52 Weber uses the term “cultivated man” in a, “completely value-neutral sense; it isunderstood to mean solely that the goal of education consists in the quality of a man’sbearing in life which was considered ‘cultivated,’ rather than in a specialized training forexpertness.” (1946:243) This has resonance for those who study Chinese unions, asleaders are generally chosen not for the “expertness” in labor issues, but rather their“cultivation” within the Party.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (73)


point in advancing decommodification, the union has failed to secure recognition andincorporation is tenuous. On the other hand, when there is bottom-up initiative to useenterprise unions to advance worker interests, we find that the leadership of the GZFTUmaintains a “passive repressive” position when grassroots activists are subject tounchecked managerial retaliation. This continual resistance to bottom-up initiative isclearly demonstrated in the case of Liu Yongyi, the union chair from the GuangdongUnion Hotel – owned in whole by the Guangdong Federation of Trade Unions. We willsee that there are attempts to secure decommodification within the auspices of the unionstructure – however, we will also see that there are severe limitations as to what can beaccomplished given highly durable union oligarchy.

Chen Weiguang – Maverick Labor Leader

I first met GZFTU chair Chen Weiguang in late 2007, when – in a highly unusualviolation of protocol for an official of his stature – Chen came to my family’s apartmentfor dinner. At the time, my parents had been working as visiting lecturers at Sun Yat-senUniversity in Guangzhou. My mother, a long time labor activist in the U.S., had firstcommunicated with the GZFTU via contacts in Hong Kong. That he was so willing tomeet and engage in substantive dialogue with foreigners is quite exceptional amongChinese union leaders, particularly in instances when there is no symbolic payoff (whichis certainly the case for an informal dinner at home). Chen arrived that evening with thedirector of the GZFTU’s international department, as well as his official translator.

My previous experience with ACFTU officials had been in highly scripted andstiff interactions with union leaders in Beijing and Shanghai. The informality with whichChen spoke was the first noticeable marker of difference. What’s more, he was engagedand interested in discussing labor politics in China and internationally, and he ended upstaying for nearly five hours. During this time, conversation ranged across a number oftopics, and I was taken aback by his obvious dedication to his work and in his highlyforthcoming critique of the existing set-up. Although I came to understand the limitationsof Chen’s political position as the research project advanced, he did not disappoint inliving up to his reputation as the most progressive union leader in China.

His dedication to union work, while not totally unique, set him apart from mostunion leaders. Zhou Ling,53 a staff member of a union federation in X city (a majorindustrial city on China’s east coast),54 compared Chen quite favorably to the chair of herown union. According to Ms. Zhou, most union chairs are selected from the Party andgovernment rather than starting as workers and moving up. In her view, the chairman inX city had felt “sidelined” when he failed to win a more prestigious post in citygovernment, and maintained a resentful attitude towards union work.55 Another worker-activist who had dealings with unions in China’s interior before coming to Guangzhoucommented that most union chairs rarely come to the union offices (preferring to spendtime at the People’s Congress where most of them hold joint appointments), and that they

53 Pseudonym.54 I have kept the city ambiguous to protect the identity of the informant, who openlycriticized her superior.55 Field Notes, May 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (74)


will only show up for important meetings. This activist thought that Chen was morededicated to his work because he spends the majority of his time at the unionheadquarters, even though he also holds a more prestigious position as vice-chair of thestanding committee of the Guangzhou Municipal People’s Congress.56

Chen’s views on labor politics were shaped by his experiences as a young workerin a state owned textile factory during the 1970s. With only a hint of nostalgia he recalledthe social bonds and lack of disparities between workers and managers in the factorywhere he worked as a youth. Even though workers only earned 30 RMB a month, thedirector’s salary was 50 RMB, which meant that, “the interpersonal relationships weregood, and we could talk about anything with the mangers.”57 His career trajectory whichbegan as a regular worker and ended as the top labor leader in one of China’s largestcities (and perhaps most important in terms of industry and manufacturing) is unusualwithin the ACFTU. While experience as a worker is certainly no guarantee againstoligarchic behavior once in office, in Chen’s case he continued to draw on this past forinspiration about how to effect progressive change in the present.

Chen referred to himself as a “resolute socialist,”58 (with no mention of “Chinesecharacteristics”) an identity which most government officials in today’s China embraceawkwardly at best.59 For him, the realization of socialism was more important even thanfollowing the leadership of the Party. Indeed, he once commented that if the Party everturned its back on socialism, he would leave the Party, though conspicuously failing tostrictly define how he would be able to determine if such an event came to pass. Havingcome of age in the era of state socialism, he continued to cling to the hope that the Partycould guide society towards a more just and equitable future. However, in a meeting in2009 he could not suppress a trace of melancholy when saying that there was a realdanger of the “socialist road” ceasing to have any meaning.60

There are of course limits to how far any union leader can go in criticizing theACFTU and other subordinate unions. That being said, Chairman Chen was more vocalin his criticism of the ineffectualness of Chinese unions than any other official in thecountry. One of the most notable of such instances was when he openly differed withnone other than ACFTU chair Wang Zhaoguo at the union federation’s 15th Congress inBeijing during the fall of 2008. Wang had given a speech where he claimed that“Western”61 and Chinese unions share some “common ground,” but that their “essence”

56 Field Notes, April 200957 Field Notes, December 200758 Field Notes, December 200759 I initially asked union officials in interviews to provide me with a definition ofsocialism, but respondents fumbled for words resulting in a highly embarrassingsituation. I decided to drop the question since it provided me with more of a “gotcha”moment rather than illuminating important phenomena, and raised the risk of alienatingpotentially important contacts.60 Field Notes, November 200961 I use quotes here because the terms “Western” and “foreign” are often consciouslyconflated by union and government officials in China. This conflation derives from theessentialist claim that “Asian” culture is fundamentally different from “Western” culture,particularly in that it is less amenable to democracy and confrontation.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (75)


(benzhi) is different.62 Chen took issue with this characterization, and in a speech at thevery same congress rejected the views of his superior, claiming that the essence of unionsin both places is to “organize and provide solidarity to workers in order to protect theirinterests.”63 Rather, it is merely the methods employed which are different, and this isdue to different national conditions (guoqing), and the political and social systems.Perhaps in part due to such comments, other municipal union leaders at the congresschided Chen for being “too much like a Western union [leader].”64

Not infrequently, Chen’s critique of the status quo and calls for reform were evenmore explicit. In a meeting with instructors from Sun Yat-sen University, he outlinedwhat he thought the primary problems were in what he termed, “officially-run”(guanban) unions: “adminsitrativeness (xingzhenghua), bureaucratism, and separationfrom the masses.”65 His solution to such a state of affairs was for reform and “moreindependence,” and that, “we can’t completely follow the Western model, but we can’tnot reform.”66 Though it is not unusual to hear union leaders use the word “reform,”Chen spoke of changing union practices with greater conviction, and as we will seeshortly, actively advocated changing existing laws and regulations in an attempt to hastensuch a process.

As may be evident from his critique of officially-run unions, Chen was at theforefront in calling for greater union accountability to its members, as well as forexpanded grassroots strength and worker involvement in the union. In his view, “Theunion is there to represent the workers but the union doesn’t have any strength unless theworkers are strong too.”67 And in a highly surprising move for a one-party state such asChina, he was quoted in the media as saying that legal action should be taken againstderelict cadres:

When an enterprise violates the rights and interests of employees and the unionchair doesn’t do anything about it but takes the opposite side from the workers andrepresents the enterprise in legal cases against workers, how can such a union chair protectthe rights and interests of workers? I support workers in suing this kind of union chair.68

Following the publication of this article he encountered some blowback, but a yearlater continued to hold his ground:

I was quoted as saying that if union cadres in Guangzhou are not doing what theworkers want them to and are being too bureaucratic, that the workers should sue thecadres. I got a lot of pressure for this, and many cadres were very worried about beingsued. But it’s important to make them represent the workers.69

62 Field Notes, October 200863 Field Notes, October 200864 Field Notes, October 200865 Field Notes, November 200966 Field Notes, November 200967 Field Notes, December 200768 November 16, 2006. “gonghui zhuxi bu zuowei wo zhichi gongren gao ta.” [I supportworkers suing union chairs in dereliction of duty] Guangzhou Ribao.69 Field Notes, December 2007

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (76)


He additionally expressed opprobrium for the method of establishing unions throughpurely administrative means, arguing, “we just want to let workers organize trade unionsby themselves.”70

This support – on the rhetorical level, at least – for worker initiative in unionactivity was paired with a relative willingness to compromise the interests of capital.Chen not infrequently would maintain that improving conditions for workers could resultin the “win-win” outcomes so insisted upon by ACFTU leadership. But he also talkedabout forcing concessions from capital if necessary. When questioned by a group ofAmerican unionists about fears of capital flight resulting from passing pro-laborlegislation, he responded that companies that were just searching for the lowest cost laborwere not contributing to the economy, and if they wanted to “move to Vietnam,” that wasfine with him.71 While in the thick of a campaign to establish unions in all Fortune 500companies operating in Guangzhou, Chen remarked that the current unionization drivewas different from previous ones for two reasons. The first was that they did not care ifGDP growth was negatively affected, and the second was that they did not care if theemployers wanted a union or not: “I’ve said in the media that we don’t need theagreement of employers to organize trade unions.”72 This lack of concern for protectingGDP growth is of course in part due to the fact that Guangzhou is one of the wealthiestcities in China. However many union leaders in other developed regions of Chinahighlight economic growth and strength in attracting investment just as much if not morethan their union achievements. For instance, in a meeting between visiting U.S. tradeunionists and the leadership of the union federation in Shanghai’s Pudong district (one ofthe wealthiest areas in the entire country), the vice-chair spoke for nearly half an hourabout economic growth and then added a few comments about labor rights almost as anafterthought. Chen did not feel that capital’s success within his district was somethingworth highlighting to visiting foreign trade unionists, and he was unafraid to discussovercoming resistance from employers in defense of worker rights.

If his willingness to criticize employers and other trade unionists is surprising inthe context of an authoritarian state, his open interrogation of the appropriate relationshipbetween Party and union is even more remarkable. Chen was always careful to establishthat the Party exercises leadership over the union, and that this relationship is just andappropriate. But he was not afraid to openly question what the precise content of such arelationship should be. Following a presentation he made to an industrial social workclass at Sun Yat-sen University73, one student bluntly asked him to detail the main factorsinhibiting union reform. As part of his response, Chen referred to the metaphor of ananny teaching a baby to walk in order to describe what sort of leadership the Partyshould exercise over the union. The argument was that while in China, nannies alwayshold the child’s hand to prevent them from falling, in the West, nannies allow children totry to walk on their own. Even if babies in the West sometimes fall, they can learn from

70 Field Notes, August 200871 Field Notes, May 200872 Field Notes, August 200873 It is quite unusual for union officials to come to university classes, and nearlyunthinkable that they would accept spontaneously generated questions from students.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (77)


the experience.74 The straightforward implication was that Chinese unions should be ableto “walk by themselves,” even if it means that the “nanny” is still watching over them. Chen’s view of strikes – one of the most politically sensitive topics in China –was more accepting than any other union leader in the country. As the number ofconflicts and strikes in China skyrocketed throughout the 2000s, most trade unioniststried to bury their heads in the sand or attribute it to “unreasonable” behavior on the partof workers. Legal strikes were seen as “Western” and not in line with purported “Asian”values of harmony and non-confrontation. On the very first night I met Chen, he arguedthat while some strikes are illegitimate,

…many strikes are reasonable and the workers are just trying to get their rights.Actually, I’m secretly happy when workers go on strike and I support it, because it givesmore pressure to the government. I think when workers go on strike for good reasons wecan take an open attitude in dealing with it.

In nearly every meeting I attended, he would speak directly and frankly about thequestion of strikes, never attempting to gloss over the expanding discontent amongGuangzhou’s working class. There were limits to his support of strikes, and he wasexplicit that he could not accept when strikers turned violent or when explicitly politicalaction was taken. But even before the watershed auto industry strikes in the spring of2010 (discussed in chapter six), Chen believed that some forms of strike action should belegalized.

Finally, Chen was very open to interactions with foreign activists, if stoppingshort of coordinating cross-border solidarity actions. In November 2007, when nearly allinternational union exchanges were quite limited in scope and channeled through thenational leadership (though there were some exchanges with the Shanghai Federation ofTrade Unions), my mother was afforded the incredibly rare opportunity to attend theGZFTU Congress, held once every five years. This was one of the few, if not only, timesthat a foreigner has been allowed to attend such a high-level union congress in the post-Mao era. In somewhat of a violation of protocol which holds that national-level foreignunions should be funneled through Beijing, Chen hosted Service Employees InternationalUnion president Andy Stern in August of 2008. With a slight chuckle, Chen noted thatACFTU leadership was “surprised” to learn of this exchange.75 Additionally, InDecember 2008, the GZFTU established a formal relationship with the San FranciscoLabor Council, one of the first such agreements with an American union federation.Finally, my own research would not have been possible had Chen not, completely of hisown accord, invited me to Guangzhou to conduct research.

In sum, Chairman Chen is quite an exceptional figure in the world of Chineseunion politics. Especially when compared to the typical leaders that see union work asjust one step in advancing their political career, Chen’s dedication to the work,willingness to confront capital, engage with foreign activists, and even to openly criticizecolleagues and the state sets him apart. In all of these regards, we might reasonablyexpect that a union under the command of such a chair might be more active in fightingfor the interests of its membership and more likely to win the allegiance of these workers,thereby advancing the institutional moment of the countermovement. But before we

74 Field Notes, December 200875 Field Notes, October 2008.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (78)


descend to the shopfloor to analyze the lives of Guangzhou’s workers, let’s first look atthe organizational context of the GZFTU as well as some of the policies that Chen wassuccessful in pushing for.

GZFTU at the Forefront of Reform?

Before getting to specific policy initiatives, it is first important to understandsomething about the political character of the union federation as a whole. Onesignificant failing of my own research is that I was never able to get a satisfactory answerto the question of how Chen became union chair. When I asked him directly, Chenmerely responded that the union has internal democratic procedures and that he waselected according to these procedures. Other officers generally attributed it to ameritocratic selection process, saying that his performance as a lower-level officer wascommendable and therefore he continued to rise through the ranks. Given thepreponderance of career politicians with little interest in labor politics that end up asunion leaders (including within the GZFTU, as we will see immediately), the meritocraticand democratic explanations do not seem sufficient. An alternative explanation is thatGuangzhou is more economically developed and has more acute class conflict than otherplaces in China, and so the Party leadership saw Chen as more likely to effect somedegree of compromise. If this were the case, then Chen’s election could be seen as aninstitutional response to insurgency. But other municipalities that are similar in theseregards (e.g. Shanghai, Dongguan) have the conservative and highly oligarchic unionleadership typical of ACFTU unions, so there must be some political factors at play aswell. Given the lack of transparency in Chinese unions, it will likely remain difficult togain greater insight into the process of leadership selection.

Chen’s ascension was certainly not due to prevailing progressivism among otherleaders within the union federation. As Mingwei Liu has noted, national leadership hadcriticized the GZFTU in 2000 for its general ineffectualness (Liu 2009). At that timeACFTU chair Wei Jianxing had pushed the unions in Guangdong to reform in order tobetter handle rising labor conflicts, and Chen’s election in 2003 may have been related tothis effort. Chen’s experience as a factory worker remained unusual in the federation, andas of 2010 only two of the six vice-chairs had similar working experience (andsignificantly these were the two oldest members). But while Chen had some minorsuccess at bringing in more overtly pro-labor officers, such people were still generallyconfined to union headquarters and were not engaged in actual worker organizing.

Under Chen’s leadership the GZFTU had made a concerted effort to bring inyounger, highly educated officers. One such officer spoke excellent English and hadstudied abroad, but had no interest in, or experience with, labor issues prior to coming tothe union. When asked why he came to the GZFTU, he mentioned a friend he had whohad worked there previously and that, “I must find a job and the union will [need to]employ somebody. And so I tried, and I got in.”76 Another such young officer wanted tocome to Guangzhou because her boyfriend had found work there. When asked why shewanted to work for the union, she responded, “because I met their conditions.”77 These

76 Interview, October 18, 200877 Interview, October 2008

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (79)


staff members certainly brought additional cultural capital and a greater worldliness intothe organization, but they were neither experienced with, nor necessarily dedicated to,union work.

There was one instance in which Chen took something of a risk to bring in a staffmember who had exhibited strong organizing skills in earlier union work. Gao Haitao hadbeen hailed in the national media as the most capable organizer to emerge in the ACFTU-directed Wal-Mart campaign of 2006. While studying for the legal exam, Gao had takenon a job at a Wal-Mart store in Nanchang, capital of Jiangxi province, to cover livingexpenses. Once the campaign in Nanchang got off the ground (against the stridentobjections of management) Gao quickly came to be recognized as having strongleadership and organizing skills, and played a crucial role in the successful establishmentof the store-level union in Nanchang. In the relatively democratic elections that followed,he received the overwhelming support of his colleagues and was elected union chair.Over the subsequent months, Gao was fearless in pushing hard for the interests of rankand file, and successfully out-maneuvered management on several occasions. However,he eventually pushed the envelope a bit too far leading to, “some divergence with thehigher levels of the union,”78 as he put it, and he was forced out of office.

At this point, Gao had received national and even international attention. But hisdeparture from the position as union chair clearly signaled that he had lost the support ofthe leadership in Beijing that he had previously enjoyed. In early 2009, a foreign laborscholar approached Chairman Chen to see if he would consider hiring Gao, as Chen hadexpressed interest in recruiting officers with more organizing experience.79 Much to thesurprise of Gao and others, Chen decided to take him on as staff in one of the union’sstreet level legal clinics. While Gao’s position was not a particularly prominent one, it isstill highly indicative of Chen’s willingness to take some risks in pushing reform withinthe GZFTU.

Despite these efforts, most GZFTU officers maintained the highly bureaucratic-conservative dispositions typical of other union officials. Many were hardly interested inlabor issues at all, and would stick to official slogans such as “harmonious laborrelations,” “win-win labor relations,” or “scientific development” when talking aboutunion business. Leadership sometimes expressed crude neoliberal conceptions of politicaleconomy, including that “[unions] cannot interfere with the objective operation of themarket.” One of the younger staff members argued that,

The most important thing is development. Maybe during the development somepeople or some organizations will pay a price, but the most important thing is

78 Interview, December 200879 To fully disclose my own involvement in this process, I originally tracked Gao downthrough one of the reporters that had written an article about the Nanchang Wal-Mart.When I met him in December 2008, he had recently left his position as union chair andwas trying to figure out how to stay involved in labor activism. I then introduced Gao tothe foreign labor scholar mentioned, who then made the suggestion to Chen. In this senseI am guilty of “contaminating” the study, however this fact in no way diminishes theremarkableness of Chen hiring an activist who was thought of as a trouble-maker byother union federations.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (80)


development. Once we’ve reached a step when we’ve developed enough… everybody caneat the cake. Then everybody can eat the cake together, but the cake must be big enough.80

Another young officer claimed that China has a, “Marxist government and so they canmediate between the two groups on equal ground.” She continued by saying, “if we wantto do union work well, we need to develop the enterprise… to allow the workers to havea good life.” The belief that economic growth and the health of the enterprise were bothof immediate benefit to workers and therefore must be a central part of union activity isfundamental to the ideology of ACFTU unions, and the GZFTU was no exception.

But if Chen’s efforts at recruiting more overtly pro-labor officers had failed toeffect a sea change within the federation’s personnel, he could sometimes use hisauthority to push for somewhat more pro-worker policies at the municipal level. Oneindication of his success in this regard was Guangzhou’s relatively high minimum wage.By 2010, Guangzhou had the second highest minimum wage in the country at RMB1100/month, behind only Shanghai where the wage standard was an additional RMB20/month (a difference of USD $3). When the global economic crisis intensified in late2008, many top officials called for a freeze or temporary suspension of the minimumwage. Chen on the other hand said, “we are unwilling to accept the [central]government’s announcement of temporarily putting off raises in the minimum wage… Ibelieve we still need to raise the minimum wage to get through the economic crisis.”81 Itis true that following a 10% increase of the minimum wage in Guangzhou in spring 2008the Guangzhou Labor Department announced that there would be no such raise in 2009,though they rebuffed calls to eliminate the wage standard.82 However, it had been thepractice for many years previously to only raise the minimum wage once every twoyears, and levels were once again increased in 2010, this time by nearly 30%. It isimpossible to assess with any great precision how important the role of Chairman Chenor the GZFTU was in such a process. However, unions are consulted in the determinationof minimum wages, and we do know that he consistently expressed strong support forincreasing minimum wages. Finally, in the spring of 2010, GZFTU vice-chair LiuXiaogang submitted a proposal to the Guangzhou People’s Congress entitled“Guangzhou Municipality Regulations on Collective Bargaining in Labor Relations.”83

This proposed regulation sought to link increases in workers’ wages to GDP, therebyattempting to halt the trend of declining real wages.

Perhaps the clearest example of the GZFTU being at the forefront of union reformwas in their attempts to win greater autonomy – from management, not from the state –for enterprise union chairs. Chen had frequently expressed frustration with the fact thatthe norm was for such officers, crucial as they are for union work, to come from high-level management and the Human Resource department rather than from rank and file.Indeed, according to the publicly announced results of a 2007 survey conducted by theGZFTU, more than 50% of all enterprise union chairs were concurrently held by 80 Interview, October 200881 Field Notes, December 200982 February 18, 2009. “guangzhou zuidi gongzi biaozhun jinnian bu tiaozheng.”[Guangzhou’s minimum wage will not be adjusted this year] Nanfang Ribao.83 April 13, 2010. “gongzi zhangfu zhui GDP fazhi zhi shi qianti.” [for wages to followGDP, rule of law is a precondition] Da Yang Wang.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (81)


someone from management.84 An additional problem Chen (and others) identified is thatthose enterprise union chairs that did try to fight on behalf of membership were subject tounchecked retaliation by management.

In an attempt to address these problems, in December 2007 the union announcedthat they would be enacting the “Guangzhou Municipality Measures for theImplementation of the ‘People’s Republic of China Trade Union Law’.”85 The first andmost important feature of this measure was to explicitly ban enterprise managers, humanresource officers, other mid-high level managers, and any of their relatives from servingas chair or vice chair of the enterprise level union. Managers that already were serving asunion chairs would not be removed from office, but would not be able to serve anadditional term (most of which last for three years). Additionally, the measure requiredthat higher levels of the trade union (most likely the district union federation) beconsulted before any trade union chair could be dismissed prior to the end of their term.This was seen as a method for reducing the likelihood of retaliation by managementagainst activist union chairs.

A year and a half later, the GZFTU felt that the “Measures for Implementation,”had not gone far enough, and passed a new resolution in July 2009. The “GZFTUResolution on Advancing Union Reform and Construction” aimed to reinforce the earliercommitment to eliminating managers from enterprise union leadership, and to furtherprotect union activists that suffered from retaliation. The union aimed to “resolve theproblem of enterprise leaders holding concurrent posts as union chairs within threeyears.”86 Of course, the need to re-affirm such a commitment was an implicit recognitionof the fact that the “Measures for Implementation” from a year and a half prior had in factnot resolved the issue of management controlling enterprise level unions. But such apublic re-affirmation was directed both at managers and at other officers within theunion, to let them know that this was something the leadership was taking quite seriously.An additional feature of this 2009 resolution was for the union to start a fund to supportunion chairs that had been fired in retaliation for their activism. But this fund was seen asa stop-gap measure for the relatively lawless environment in which activist union chairsfaced essentially unchecked retaliation. Chairman Chen was quoted in the newspaper assaying, “National laws and Guangzhou Municipal laws clearly protect the rights of unionchairs. If [union chairs] are pushed out or fired, this is only a short-term behavior. Oncethe union finds out about such activities, it will intervene and handle the matter to protectthe rights of the union chair."87 The implication is that the long-term intention is to ensure

84 December 10, 2007. “xingzheng fuzeren jianren zhuxi de gonghui hai jiao gonghui

ma?” [Is a union with a manager holding concurrent post as chairman still called aunion?] Xinhua Wang.85 Such “measures for implementation” are common in the Chinese legal system, andallow local governments to adapt national laws to particular conditions.86 July 24, 2009. “guangzhou tuijin gonghui gaige he jianshe, qiye lingdao bu ren

gonghui zhuxi.” [Guangzhou promotes union reform and construction, enterprise leaderscan’t serve as union chair] Guangzhou Ribao.87 July 24, 2009. “guangzhou tuichu xin guiding: qiye lingdao bude jianren gonghui

zhuxi.” [Guangzhou unveils new regulation: enterprise leaders cannot hold jointappointments as union chair] Xinxi Shibao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (82)


that union chairs are not fired in the first place, rather than have them rely on the unionsupport fund.

Once the government’s worst fears about the economic crisis passed, the uniononce again went on the offensive. In the spring of 2010, GZFTU vice-chair Liu Xiaogangsubmitted a draft regulation to the Guangzhou Municipal People’s Congress in hiscapacity as a congressional representative. The draft regulation, entitled “GuangzhouMunicipal Labor Relations Collective Negotiation Regulations,” sought to addressdifficulties the union federation has experienced in the course of collective negotiations.As stated in the official introduction to the draft regulation, failures in collectivebargaining had resulted in negative consequences for Guangzhou’s workforce: “Increasesin workers wages have not been ideal. From 2003 to 2007, workers’ wages did notdevelop along with the economy and society, they did not grow along with GDP.”88 Inthe hopes of overcoming the issue of frequent management recalcitrance, the newregulations called for a fine of up to RMB 20,000 for failure to respond to a request forcollective negotiations within twenty days. However, GZFTU legal department chairZhang Ruizhou was explicit that management could not simply pay the fine to avoidnegotiations: “The fine doesn’t mean you don’t have to negotiate, bargaining still has totake place as before.”89 While RMB 20,000 may not be a huge sum for most companies,the regulations were intended to send a clear message about the importance the union wasattaching to promoting collective negotiations.

How then should we assess these moves by the GZFTU under Chen Weiguang’sleadership? In particular, is it possible to detect the emergence of the institutionalmoment of the countermovement? One area in which Chen has had little success is inrecruiting more union officers who have an interest in, and are dedicated to, advancingthe interests of the working class. While there are a few exceptions, in general the movehas been towards greater professionalization, i.e. recruiting officers with advanceddegrees and legal training, but with little or no working or organizing experience. Part ofthe problem is simply that there is no social movement field in China, and so there arevery few candidates. But the explicit decision to hire college grads rather than to recruitfrom among rank and file of course has important implications for the political standpointof union organization.

In some ways, the attempt to prohibit management from serving as union chairswas not a breakthrough at all. Indeed, in 2006 (more than one year before the “Measuresfor Implementation” were enacted) the ACFTU had inserted the following language intothe national-level “Enterprise Union Work Regulations”: “Enterprise administrators, theirpartners, or close relatives cannot stand as union committee members in the enterprise.”90

88; accessedJanuary 6, 2011.89 March 18, 2010. “gongzi jiti xieshang mei nian zhishao yici zhigong xiang zhangxin

qiye bu licai jiang fakuan.” [collective wage negotiations at least once a year, ifemployees want to raise wages and the enterprise ignores it, they will be fined] Dushi

Kuai Bao.90 Qiye gonghui gongzuo tiaoli (shixing). [Enterprise Union Work Regulations (trial)]. (accessed January 6,2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (83)


Since enterprise union chairs are necessarily members of the union committee, thisregulation would seem to ban the practice of managers serving as union chairs. TheGuangzhou regulation gives greater specificity as to which people count as“administrators” and expands the scope to include vice-managers and officers fromhuman resources departments. Indeed, six months after the “Measures forImplementation,” the ACFTU issued a more detailed set of regulations on the“production of union chairs” which closely followed the example set in Guangzhou. Butthe most important difference is how the enactment of these regulations differed on arhetorical level. China has many excellent laws and regulations – including theconstitutional rights to freedom of association and speech – that officials never mentionpublicly and which are widely understood to be inoperable.91 Agents of the state mustleverage their symbolic power to summon such regulations into efficacy, something theyare frequently indisposed towards. While the ACFTU quietly inserted its restriction(totaling one line) deep into a lengthy set of regulations, Chen Weiguang and otherofficers of the GZFTU made it a point to talk openly, forcefully, and frequently about thenew ban. The resolution that was passed a year and a half later was in essence justanother re-assertion of the union’s intention to enforce existing regulations. In theseregards, the regulations, measures, and resolutions passed in Guangzhou represent onlymarginal improvements on a formal level; rather it is the way in which they are talked

about by those in power which lends them greater credence, and which distinguishesChen Weiguang’s GZFTU from other municipal union federations.

These regulations, along with those to protect activist union chairs, were meant toensure greater autonomy for enterprise level unions from capital, if not from the state.Greater autonomy from capital would be necessary to make collective negotiation moremeaningful, and the draft regulation from 2010 was meant to ensure that managementcould not refuse to bargain. The draft regulations on collective negotiation weresubsequently tabled when the 2010 spring strike wave in the auto industry put collectivenegotiation high on the provincial and national agenda. However, that the GZFTU wasconsidering this even before the strike wave is an indication that they were ahead ofwhere other union federations were on this issue.

At this point it should be clear that Chen Weiguang is a highly unusual characterin the world of China’s labor politics. He has used his position of power within theGZFTU and Guangzhou’s government to push for more pro-labor policies, and has triedto force union cadres in his own jurisdiction to be more beholden to workers. Thequestion now becomes, to what extent has relatively pro-labor leadership and policy beentranslated into decommodification and incorporation of labor? To answer this question,we must leave the halls of power and enter the workplace.

Hitachi – The Best Union in the Country?

From the perspective of Chen Weiguang and the GZFTU, the union at the HitachiElevator (China) plant in Dashi town was perhaps the best in the municipality. As we will

91 O’Brien and Li’s (1999) argument about “selective policy implementation” wasdeveloped to describe rural politics, but holds analytical value for urban labor politics aswell.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (84)


see, the Hitachi union had been showered with praise from the GZFTU and was ashowcase for visiting foreign trade unionists. While conditions for many – but certainlynot all – Hitachi employees were better than for most manufacturing workers in the PearlRiver Delta, it is not clear that the enterprise union played a decisive role in bringing thisabout.

Hitachi Elevator (China) was established in 1995 as a joint venture betweenHitachi and the state-owned Guangzhou Guangri. The initial investment was USD $90million, with the Japanese holding a 70% stake in the company and the Chinese theremaining 30%.92 The joint-venture produces elevators, escalators, and moving sidewalksat production facilities in Guangzhou, Shanghai, and Tianjian. It has consistently beenone of the top producers of elevators in the country.

I first had the opportunity to visit the Guangzhou Hitachi facilities whileaccompanying a delegation of visiting American trade unionists in December of 2008.We toured the incredibly clean and modern facilities on electric golf carts, and then wentto a conference room to hear about the union. We learned that the enterprise union hadwon a series of official accolades over the previous five years, including being deemedby the GZFTU as a, “model workers’ home” (mofan zhigong zhi jia), “superior tradeunion unit,” and an “advanced trade union unit.” Additionally, the Guangzhou Hitachiplant was one of only eight companies to be awarded the “AAA Harmonious LaborRelations Enterprise” rating by a committee consisting of the municipal labor department,employers’ association, and the GZFTU. Enterprise Party secretary and union chair HuFeng93 spoke in glowing terms about the accomplishments of the union, and specificallymentioned that their union was better equipped to serve their membership than those ofother joint ventures in Guangzhou. He was very proud of the collective contract they hadnegotiated on behalf of workers, and said that it required a huge amount of effort. Thisincluded seeking the advice of rank and file and a series of challenging negotiations withmanagement.

And there were reasons to believe that Hu Feng took his role as union chairreasonably seriously. Unlike many other union officials I encountered during myfieldwork, he was not only willing, but quite enthusiastic to meet with me to discussunion work. Although I was invited to join him at a fancy seafood restaurant for theinterview, he arrived wearing the same outfit as production workers, adorned with a smallhammer and sickle lapel pin. Hu’s background was similar to Chen Weiguang’s in that heworked his way up from the shopfloor through the union structure. Starting in 1979 hewas a union committee member in the Chinese Navy, and switched to the state-ownedelevator company in 1994 where he continued with union work. And in sharp contrast tomany enterprise level union officials, he showed a keen interest in the labor politics of theU.S. and questioned me intensely during our lunch.

While not quite as politically astute as Chen Weiguang, he displayed eminentlypro-worker instincts and had only positive things to say about Chen’s leadership. Inparticular, he thought that Chen had been effective in winning support from regularworkers: “[Chen] has done some important things [for workers], and they have welcomedhim… He is very easygoing with workers, and gets along with them. He is an outstanding

92, accessed January 7, 201193 Pseudonym.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (85)


representative of workers.”94 Hu related a story where leaders from the GZFTU askedhim what the union’s greatest accomplishment had been in recent years, and heresponded, “It is that… the crisis had an influence on the company [in terms of reducedorders] but we didn’t lay off workers.”95

But perhaps most significantly, Hu was asupporter of (relative) independence from management. He mentioned how Chen statedvery clearly that enterprise union chairs cannot come from management, and said that,“This method is correct… If management is the union, how can this work? Your positionwon’t be correct, it will be hard [to do your work].”96 Of course he did not advocate ahostile relationship to management, arguing that, “in China, the union has to closelycoordinate with the administration, it isn’t antagonistic,” and, “The enterprise mustdevelop a lot, and this depends on the workers.”97 Additionally, he had high words ofpraise for the company president, with whom he had a close working relationship.

But if the leadership of the Hitachi union was relatively progressive, they hadfailed to make such an impression among the workforce. After my official introduction toHitachi management and union officers, I returned to the district where the plant waslocated to meet workers independently from official channels. While not all workers hada negative impression of the union, they were hard-pressed to find something positive tosay.

Many, but not all, workers were at least aware of the fact that they had a union,something which is not true at a very significant number of unionized workplaces. Uponhiring, they received some introductory materials which described the union’s work. Butthis introduction would be unlikely to attract the interest or enthusiasm of the young menand women that make up the workforce, as union activities were described in the stiff andclichéd official language of the state. The employee handbook introduced the union asfollows:

As our nation’s mass organization of the working class and the bridge and linkbetween the Party and working masses, the basic task of the union in this new period is tolead the working masses in implementing “Three Represents” important thought.Representing China’s advanced productive forces, representing China’s advanced culture,and representing the basic interests of the broad masses, causes us to more deeply clarifythe basic responsibilities and other social functions of the union… The union organizationmust fully implement its social role, and at the same time as upholding the general interestsof the entire nation, express and uphold the specific interests of workers. This is to upholdthe stable employment of the ranks of workers and to ensure political and social stability.

The document does go on to talk about union successes in improving health coverage,wage payments systems, etc. Additionally it claims that, “enterprise employees recognizethe active function of the union in promoting occupational safety, labor protection, andworker benefits.”98

94 Interview, April 200995 Interview, April 200996 Interview, April 200997 Interview, April 200998 This document was provided to me by Hitachi workers.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (86)


And yet, interviews with rank and file workers that I met through non-officialchannels revealed quite a different perspective. First, and perhaps most crucially, is thatnone of these workers were aware that their employment was (theoretically) regulated bya collective contract. Workers did receive individual contracts, one of which wasprovided to me by a worker at my request. But there was no understanding that theirsupposed representative, the enterprise union, had played any role in negotiating theterms of employment. This stood in strong contrast to the claims of union leadership thatmembership had been contacted through the shopfloor organizations (fenhui) and theiropinions solicited.

Even if workers were not active participants in the union, things for a certainportion of employees were relatively good when compared with other workers in theregion. For the “regular” workers (as opposed to the “interns,” which will be discussed ina moment) they could count on a degree of job security, wages which were aboveminimum wage, one day of rest a week, and some employer contributions to socialinsurance. Regular workers had to undergo a six-month trial period, during which timethey received a reduced salary of RMB 1000 (at the time, just above minimum wage). Ifthey successfully made it past the trial period, they would receive a raise of RMB 400.With overtime, most workers were able to earn around RMB 2000/month. The averagemonthly wages in Guangzhou in 2009 were RMB 922 for rural residents and 2,301 forurban residents (and since most Hitachi employees were from the countryside, this meanstheir wages were significantly higher than other migrant workers in the city).99 Workersconfirmed that they had always received their wages on the 15th of the month asstipulated in their contract. And it appeared as if the company was fulfilling itsresponsibilities in paying for social insurance. One worker who suffered appendicitis saidthat he got time off from work and the company had paid for 60% of his operation,thereby greatly reducing the financial burden. Most regular workers characterized thebenefits at Hitachi as “somewhat better” than other places in Guangzhou.

But if things were reasonably good for the regular workers, the “interns”(shixisheng) which constituted up to half of the workforce in some workshops weretreated as flexible and highly exploitable labor. Following the model of labor forcedualism so prevalent in China’s capital intensive industries (Zhang 2008), Hitachisurrounded the core of regular workers with a cadre of interns which were hired fromtechnical schools. Many of these interns were illegally employed for periods of up to andexceeding one year (which is beyond the legally mandated amount of time). Workersfrom the electronic equipment department claimed that nearly half of the employees intheir shop were interns, who despite doing similar work received much lower pay (aroundminimum wage) and benefits than regular workers (also a legal violation), and enjoyedno job security. One such intern expressed great dismay at her continued informal status,and was worried that even after working for more than a year they would not sign aformal contract with her. Additionally, since they were not regular employees, theseinterns were not allowed to join the union. When asked if the union could be of any helpto her in trying to secure a regular contract, one long-term intern merely replied, “us

99 (accessed January 25, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (87)


employees haven’t encountered them [the union].”100 One worker who had made thetransition from internship to a regular contract said, “I don’t know if the union protectsthem [interns]… they just protect the regular workers.” The only union functions internswere aware of were the “entertainment” (yule) activities, which include collectivebirthday parties and the occasional field trip to nearby attractions. When questioned aboutthe union, one intern could only say, “I’ve heard that they take you to some fun placesduring holidays.” Although union chair Hu had said that there were no layoffs during thefinancial crisis, he did admit that there were some “individual adjustments.” It was notuntil I spoke with interns that I discovered that he was referring to firing interns (whereas“layoffs” would refer only to regular workers). In short, the interns at Hitachi weremaintained as a reserve army of cheap and flexible labor to be disposed of at will,without any of the contractual or legal protections afforded to regular workers.

But even the contracts extended to the regular workers were, on a formal level,severely lacking in content. As has already been mentioned, regular workers were notaware of any collective contract, and the union denied my request to see such adocument. Therefore, it is the individual contracts that are the only formal recourseworkers have when involved in workplace disputes with management. The contract doesvery little to specify what sorts of conditions and benefits the workers are entitled to, andalthough it does list a monthly wage, there is no indication of how many hours are to beworked within a month. In nearly every section of the contract it is merely stated that“relevant laws and regulations” will be followed, without specifying what those are orhow a worker might find out what those are. Below are two such sections, in theirentirety, which would seem to be crucial to any labor contract:

Section IV. Social Insurance

Party A [employer] and Party B [employee] will participate in and contribute to socialinsurance according to relevant national, provincial, and municipal laws and regulations.According to the law, Party B will enjoy the relevant social insurance benefits.Section V. Labor Protection and Labor Conditions

(1)Party A will, according to the work needs of Party B, establish the standardimplemented work-time system.(2)Party A will implement relevant national, provincial, and municipal regulations onwork, rest, vacations, and labor protection, and will provide Party B with labor conditionsand safety equipment in accordance with national regulations.

Nowhere in the contract is any mention made of a collective contract, of the unionrepresenting the worker, or how to receive information on relevant laws. Again, nomention is made of the union in the section on resolving labor disputes. Rather theindividualizing legal procedures (Friedman and Lee 2010) are briefly outlined (firstinformal negotiations, then filing with the enterprise level labor dispute committee, andfinally filing for arbitration with the labor bureau), without providing any details abouthow to proceed with such a process. In general, the reference to unspecified legalstandards, the vagueness of the contract, and the failure to mention the potential avenuesfor resolution of collective problems leave the individual worker – even one with aregular contract – in a highly precarious position even at the formal level.

100 Interview, April 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (88)


This invisibility of the union qua collective voice of the workforce was reflectedin workers’ responses when asked how they go about resolving problems on the job. Thetemp workers were all unhappy with their unequal treatment, but none of them hadconsidered seeking the help of the union. Some of these workers had no idea what aunion was, with one of them stating, “I don’t have any understanding of it,” whileresponding with an emphatic “no!” when asked if she had ever encountered a unionofficer. Another regular worker had a relatively positive response when initially askedabout the union, saying, “The Hitachi union will protect your basic rights”101 and thatthey always get their appropriate days of rest and are paid on time. But when pushed toprovide some more details about how specifically the union had helped them, he said,“Most of us employees don’t understand specific things about the union.” As theconversation progressed we began to discuss some specific dissatisfactions he had atwork, most of which centered on an abusive manager, and the methods he had forresolving such grievances:

That manager was rather coercive, and I was dissatisfied with his managementmethod, so I said something to higher leaders. [I said] ‘that manager has a big temper…lots of people are afraid of him.’ But the higher leaders said that they didn’t have anythingto say about this manager… The union in the enterprise doesn’t give us a lot of help. Theunion doesn’t have any use. In form it says that they will help us, but in reality there isn’tany protection.102

While levels of satisfaction with their jobs varied for different workers, none of them saidthat they would seek the help of the union in resolving problems. If one of the corefeatures of a trade union is to represent employees in resolving individual and collectiveissues in the workplace, the Hitachi union failed to accomplish such a task – and thisfailure was even more spectacular when the intern workers are taken into consideration.

In sum, overall conditions for regular workers at Hitachi were certainly somewhatbetter than for other manufacturing workers in Guangzhou and the Pearl River Deltamore broadly. The extent to which labor was decommodified for these workers and whatthe union’s role was in bringing this about will be addressed below. Regardless, theexistence of labor market dualism and the pervasive legal violations in long-termemployment of interns represented a serious problem at Hitachi, and is indicative of theinability/unwillingness of enterprise unions to think expansively about who makes uptheir constituency. The union suffered from extremely low legitimacy among interns andregular workers alike, and they consequently would not turn to the union to resolveworkplace problems. Thus, even though overall working conditions in the Hitachi factorywere not as degraded as many other factories, the “model” union had failed to make apositive impression on their membership.

We will now turn to another “most likely” case in a different sector: the union-owned Guangdong Union Hotel.

The Union as Employer

101 Interview, April 2009.

102 Interview, April 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (89)


One of the most noteworthy features of reform-era China’s political economy –and something which distinguishes it sharply from liberal democracies – has beenincreasing demands for government agencies to generate revenue through marketmechanisms rather than taxation (Duckett 2001). This impetus towards profit is perhapsbest exemplified by the business practices of the People’s Liberation Army (Bickford1994), but also extends to union federations. Such businesses are administratively distinctfrom state-owned enterprises (SOE), and thus far there is no scholarly research on laborrelations in these enterprises. Unfortunately, the case of the union-owned GuangdongUnion Hotel does not provide much hope that workers in these enterprises enjoy anythinglike the privileged position that SOE workers do.

I first came to understand the logic behind such state-owned businesses in asomewhat humorous moment while accompanying a delegation of American unionleaders in Shanghai. While driving to a banquet with the leaders of the ShanghaiFederation of Trade Unions (SHFTU), our guide – an officer in the internationaldepartment of the SHFTU – was beaming with pride as he told us that we would bedining in the first union-owned five star hotel in the country. One of the members of theU.S. delegation asked the guide why it is that a union would own a hotel. Looking quiteperplexed the guide responded, “Why, to make money of course!” This was met withlaughter from the Americans, but a profound point about the logic of governmentagencies was thus revealed.

There were several union-owned hotels in Guangzhou, though none as luxuriousas those of the SHFTU. These hotels are open to the public, and as was the case inShanghai, they are used for generating profit according to market principles. Trivialthough it may seem, one indication of the managerial logic applied in these businessesappeared in the form of a piece of art hanging in the restaurant of a GZFTU-owned hotel.Done in a traditional style, the piece of calligraphy read, “If you don’t work hard today,you will work hard looking for a job tomorrow.” [see image 1] The specific hotel we areconcerned with here – the Guangdong Union Hotel (Guangdong gonghui dasha) – isowned in whole by the Guangdong Federation of Trade Unions (GDFTU) and it occupiesthe western side of the same building that houses the union headquarters. The 75 room,three star hotel is administered by the GDFTU’s property management committee.

Image 1

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (90)


“If you don’t work hard today, you will work hard looking for a job tomorrow.” Photo by theauthor.

The central actor in the Guangdong Union Hotel case was a woman named LiuYongyi, a 52 year old hotel employee and chair of the enterprise level union. Liu beganworking in the hotel in 2000, and shortly thereafter was involved in establishing a unionorganization. She gained considerable support among other hotel employees, and thefollowing year was elected to a five-year term as union chair. Although it is unclear justhow competitive and transparent the election process was, in 2006 she was re-elected foranother five-year term.

It was in the first year of her second term as chair that Liu began to attract the ireof hotel management. The company’s policy had been to only purchase medicalinsurance for employees with a Guangzhou hukou (household registration), which meantthat migrant workers in the hotel were on their own when it came to medical expenses.103

Not coincidentally, the ten employees that had Guangzhou hukou were in management,meaning that all of the other workers were without insurance. Liu brought this up withmanagement and after several months won health insurance for all employees. The nextstruggle she took up centered on legal violations in non-payment of overtime wages.Management was refusing to pay employees the additional wages to which they werelegally entitled for working extra shifts on weekends and on holidays. Once again, Liuwas successful in remedying this problem, and gained increased stature with her

103 Information in this paragraph is derived from media sources.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (91)


constituency as a result. Finally, in late 2007, a number of employees came to the unionto complain about how increased rents in the city were creating financial hardship forthem. Liu began a multi-month struggle to establish a housing fund for employees, whichended victoriously the following spring. These three victories firmly established herreputation as a competent representative of the hotel’s employees.

But if employees at the Guangdong Union Hotel were supportive of Liu’saggressiveness, management was increasingly frustrated. Following a rapid succession ofgeneral managers at the hotel, Jiang Lingquan was brought in at the end of 2008. InDecember of that year, one of his first moves as general manager was to seal up Liu’soffice and notify her that she was not to discuss union business. As Liu fought this overthe next few months, management decided to take more decisive action, and on April 1st

notified her that she was being dismissed at the end of the month. Losing her employmentat the hotel also meant that she would not be able to continue to serve as union chair.Two other members of the union committee were also notified of their dismissal. Justover one week before her contract was to be terminated, the story blew up in the media,receiving extensive coverage both in Guangzhou and nationally.

After a number of articles sympathetic to the union chair’s cause appeared,management went on the counter-offensive and held an emergency press conference.General manager Jiang’s explanation for her dismissal focused on two main issues. Onthe one hand, he portrayed Liu as corrupt, having “problems with her character,” and inneglect of her work duties. But it was clear that dereliction of duty was not the only issueJiang was concerned with. Rather, it became apparent that difficulties she created formanagement were a motivation: “[she] has created a negative environment here. She’shad really bad relationships with the few prior general managers, and the cancellation ofher contract was done according to the rules.”104 Becoming somewhat exasperated duringthe press conference, Jiang perhaps unwittingly revealed that her dismissal was highlycalculated and involved union leadership: “We the GDFTU should explain this to theGuangzhou media: last year when the Party organization sent me to this hotel, it wasprecisely because she had created such a bad environment.”105 Thus, while trying to castaspersions about Liu’s character, it became clear that the primary motivation for thefiring was related to her activism.

The general manager additionally invited the media to “come talk with ouremployees when you have time, many of them don’t like her.” But in a spurt of quiteaggressive journalism, a number of reporters spoke with regular employees about theirviews of the union chair. Without exception, they supported Liu, with one worker saying,“she has a strong sense of justice, and won overtime and housing subsidies for us.”106

Additionally, employees believed that Liu’s firing was directly related to her activities asunion chair rather than any defect in character, with some describing her as “morallyupright.”107

104 April 23, 2009. “Gonghui zhuxi bang yuangong weiquan jing zao jiegu” [Union chairhelps employees protect their rights, is fired] Nanfang Ribao.105 Ibid.106 April 23, 2009. “Gonghui zhuxi bang yuangong weiquan jing zao jiegu” [Union chairhelps employees protect their rights, is fired] Nanfang Ribao.107 Ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (92)


Higher levels of the union refused to come out in support of the fired union chair,as the district and municipal union federations remained silent on the matter. Provincialunion federation vice-chair Kong Xianghong, considered by many to be quite progressivefor his advocacy of reforms in collective bargaining laws, joined Jiang Lingquan at thepress conference and declared that the GDFTU would “not side with either party.”108

Somewhat surprisingly, the most sympathetic union official was the GDFTU’s propertymanagement committee chair Wu Zhaoquan. Wu was quoted publicly as saying, “Theunion is originally a unit for helping workers protect their rights. Enterprises owned bythe union should strictly follow the Labor Contract Law in employment relations.”109 Heeven went so far as to say, “the way the hotel handled the termination of the laborcontract was clearly inappropriate.”110 But the tenor from people within the union thathad the stature to change the outcome was decidedly less supportive.

The media, for its part, pushed the ambiguous borders of permissibility (Hassid2008) and provided the public with extensive and highly critical coverage of the story.Following a number of sympathetic pieces from prominent outlets such as Nanfang Ribao

and Guangzhou Ribao, even more critical opinion pieces appeared on the internet. Onesuch piece was provocatively titled, “How many more union organizations will be‘raped’ by employers?” and provided an incisive critique not just of managementretaliation, but also of the union:

If companies can do whatever they want right under the nose of the highest trade unionorganization in the province, and they haven’t paid any attention to workers’ legal rightsand interests, who would dare hope that the union can represent workers’ interests…? Evenmore frightening, how many other union organizations, union committee members, andunion members have in reality been ‘raped’ by employers?111

Another opinion piece argued that the system of management funding union activitiesand union chair’s salaries was at the heart of the problem and that, “unions should be paidindependently.”112 Unfortunately, the huge amount of attention generated by the coverage

108 April 22, 2009. “guangdong gonghui dasha gonghui bang yuangong weiquan dezuilingdao bei cai.” [Guangdong Union Hotel union chair helps employees protect rights,offends leaders and is fired] Yangcheng Wanbao.109 April 23, 2009. “‘gonghui zhuxi zao jiegu’ xu, dasha zongjingli cheng zhong lequantao.” [‘Union chair fired’ continued, hotel manager claims he was set up] Nanfang

Ribao.110 April 23, 2009. “guangdong gonghui dasha gonghui zhuxi bei chao.” [GuangdongUnion Hotel chair is fired] Guangzhou Ribao.111 April 24, 2009. “hai you duoshao gonghui zuzhi bei qiye ‘qingjian’?” [How manymore union organizations are “raped” by employers?] Xibu Wang.(, accessed January 13,2011)112 April 23, 2009. “gonghui zhuxi zao jiegu qianyin houguo, gongzi fafang duli cai youdiqi.” [The cause and effects of union chairs being fired, salary should be paidindependently in order to have courage] Longhu Wang.(, accessedJanuary 13, 2011)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (93)


upset the authorities, and three days after the story broke all reporting on the caseconspicuously disappeared.

The final outcome of the case was not reported in the media. But there is no doubtthat Liu Yongyi was successfully fired and removed from her post as union chair. Itwould have taken a very strong intervention on the part of labor heavyweights such asChen Weiguang or Kong Xianghong to reverse management’s decision. Additionally,union leadership would have been eager to take credit for such an intervention had ittaken place, as it would have been a prime opportunity to burnish their credentials asdefenders of the working class. But the GZFTU and GDFTU’s emphatic neutrality – a“passive repressive” response – was tantamount to siding with management, a fact thatwas surely not lost on Liu Yongyi or other labor activists following the story. TheGuangdong Union Hotel case thus took pole position as one of the most egregious andtragically ironic instances of unchecked managerial retaliation against union activists.

The Institutional Moment in Guangdong?

Thus far, we have analyzed two workplaces within the jurisdiction of ChenWeiguang’s GZFTU where we should be most likely to find the union playing animportant role in bringing about institutionalized decommodification and incorporation oflabor: The award winning Hitachi Elevator joint venture, and the GDFTU-ownedGuangdong Union Hotel. How should we assess the outcomes of union activity in theseenterprises?

It is first important to establish the extent to which labor was decommodified foremployees. Hitachi was characterized by a dual labor market in which a portion ofworkers received relatively good wages, job security, and benefits. For these regularworkers, their labor was comparatively decommodified. None of them were laid off whenorders dropped in late 2008-2009, thus indicating a degree of workplace security. Socialinsurance payments were made on time, and workers were able to get significantsubsidies for medical expenses, both indications of social protection. Wages were goodfor migrant workers, even if they were far short of what would be required to make a lifein the city. There were no indications of participation in production, as regular workersremained completely disengaged from union activity, save its entertainment functionssuch as birthday parties. And instances in which workers were unhappy with oppressivemanagerial styles were not resolved.

But if regular workers enjoyed a degree of protection from the market, thetemporary interns remained fully subject to the vagaries of managerial autonomy andmarket fluctuations. These workers were by and large unaware of the union or any otherchannel for resolving grievances. They did not receive contracts, had wages significantlybelow those of regular workers, and did not receive benefits. Employees were often keptas interns for much longer than is legally permitted, often times for more than a year. Andwhen the economy stumbled during the economic crisis, they were the first ones to losetheir jobs. In short, any decommodification that regular workers were the beneficiary ofwas not extended to the interns who made up a large segment of the workforce.

Then we have the question of whether union organizations either at the enterpriseor municipal level played a significant role in bringing about the limiteddecommodification that was secured for regular workers. It is impossible to strictly

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (94)


determine causality in this case, as the union and management both operate with littletransparency. I was not allowed to see the collective contract, and my only account of theprocess of collective bargaining comes from union officials themselves. Given workers’complete ignorance of the collective contract, it is impossible to independently verifywhat the union’s role was, or to develop a counterfactual as to what might have happenedwithout the union. But there are a number of reasons to believe that the relatively goodconditions for regular workers derives more from Hitachi Elevator’s capital-intensivenature of production, the relatively strong market position of its workers, and mostimportantly that it is partly held by a state-owned enterprise rather than from the activitiesof the union as representative of the workers. In this sense, the union has not moved inthe direction of incorporation.

Although managers in state-owned enterprises in China have continually movedtowards an embrace of market principles (Gallagher 2004; 2005; Zhao and Nichols1996), conditions for workers are in general still better than in the private sector – at leastfor those who are lucky enough to work for profitable firms (Chan and Unger 2009). Abrief comparison with three other enterprises in Guangzhou, the foreign-owned OtisElevator, foreign owned-Nanhai Honda (both of which will be discussed in greater detailin chapter six), and the joint venture Guangzhou Honda is instructive. All four of theseenterprises are capital intensive, and employ relatively skilled workers. All four of themhave union organizations. However, workers’ conditions (most importantly wages) aresignificantly worse at the two privately owned companies, Otis and Nanhai Honda. Aswe will see in the next chapter, such conditions eventually caused workers in these twofactories to revolt, resulting in highly militant strike action. Despite the strike wave of2010 which largely affected Japanese-owned enterprises (mostly in auto, but in otherindustries as well), neither Guangzhou Honda nor Hitachi experienced any labordisturbances. Although conditions for workers in these plants certainly leave somethingto be desired, the ability of regular workers to consistently make over RMB 2000 and tohave social insurance paid on time was something that workers in the privately heldcompanies could not enjoy. Although such a tiny sample prevents us from drawing anystrong conclusions, it is quite likely that conditions for workers improve when the state isa major shareholder in a company, particularly in capital intensive industries with askilled workforce – precisely the type of enterprise we have with Hitachi Elevator.

Perhaps the objection may be raised that Hitachi is not a state-owned company,but rather a joint venture which is only 30% owned by the Chinese partner. Indeed, thestate’s minority share does indicate that the Japanese investors retain a strong hand inmajor managerial decisions. However, the managerial hierarchy in the company highlyfavors the Chinese. First and most importantly is that the CEO, Pan Shengshen, isChinese and has had an active political career, holding prominent positions inGuangzhou’s municipal legislative body as well as the People’s Political ConsultativeCongress. As the Hitachi union chair put it, “Because he has these positions, the Japanesereally trust him.”113 Pan maintains close relations with many prominent political figuresin Guangzhou, and attended a banquet held by Chen Weiguang in honor of visitingAmerican union leaders. Additionally, all twelve of the “high level managers” at Hitachiare Chinese, whereas the Japanese only are vice-managers. In short, while the Japanese

113 Interview, April 2009.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (95)


maintain a majority share in the company, it is clear that primary control is in the handsof the Chinese managers that have close ties to the Party and local government. Withthese manager/politicians in key positions in the enterprise, the state (broadly conceived)is able to exercise significant control over key decisions, including treatment of workers.

It is because the limited decommodification at Hitachi came about as a result ofstate action, rather than democratic mobilization of the workforce, that the “success” is solimited in scope. Most significantly, the continual heavy-reliance on interns at Hitachirepresents a significant failure, and a potential source of instability. Most union officialsargue that interns, temporary, and “dispatch” (paiqian) workers are not formal employeesof the enterprise and therefore are not a concern of the union. This is a reflection of thenarrowly focused service-based orientation of ACFTU-subordinate unions. And in asituation when even regular workers remain highly atomized, ignorant of their collectivecontract, and disengaged from the union in general, their gains remain tenuous. As theChinese state always reserves the right to violate its own laws, it would not be surprisingto find that the collective contract could be easily ignored by management if deemednecessary. In a slight modification of Feng Chen’s argument about individual rights inChina being undermined by a lack of collective rights (Chen 2007), we see here thatindividual material gains can potentially be undermined by lack of collective power. Inshort, failure to incorporate the workforce implies that material gains are segmented (onlycertain workers enjoy better treatment) and potentially reversible.

If in Hitachi a degree of decommodification came about for a segment of theworkforce as a result of state action, the Guangdong Union Hotel presents us with a verydifferent scenario. Here we saw an activist enterprise union chair who had strong rapportwith and support from regular employees. These workers knew Liu Yongyi and weregrateful for her leadership within the union. In this sense, the enterprise union wasoperating as worker representative, rather than as an unmediated expression of statepower, and was able to solve problems through rationalized channels. However, it waslikely because Liu had begun to consolidate a base of support among the membership thatthe higher levels of the union took a “passive repressive” position in response to Liu’sfiring. That is to say, it was precisely at the moment when pro-worker leadership fromChen Weiguang (or other high level union officials in Guangzhou) could have made a

difference that such leadership evaporated. Simply by doing nothing (i.e. behavingpassively), the inherent repressive capacity of capital to hire and fire was unleashed.Given this structural power asymmetry at the point of production, passivity from unionleadership is tantamount to repression.

The Guangdong Union Hotel raises a few additional salient points. The first isthat Liu’s struggles against management were a clear example of the Polanyiancountermovement in action. Her constituency demanded social protection (housingsubsidy, social insurance) and workplace security (appropriate overtime wages) to protectthem from market fluctuations, the clearest example of which was increased housingcosts. Liu mobilized the organizational resources at her disposal to fight for, andeventually win, these protections. In this sense we can see that the union structures at theenterprise level maintain the potential to be directed towards decommodifying andincorporating ends. Liu’s victories are an indication of the first hints of the institutionalmoment of the countermovement.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (96)


But the fate that befell Liu Yongyi was not a chance occurrence. Indeed, activistenterprise-level union chairs have emerged in a variety of different industries all over thecountry, and have on many occasions successfully challenged capital in advancing theinterests of membership. But time and time again, these union chairs are summarily fired,and almost always in violation of the Trade Union Law (which requires a majority recallvote among a membership congress).114 The reason that the Guangdong Union Hotel casestruck a chord in Chinese society is because it was such a common occurrence, but onewhich many were surprised and angered to see repeated at a union-owned enterprise. Thecased revealed the profound vulnerability of these union chairs to unchecked managerialretaliation (Chen 2009), even in the enterprises one would least likely expect to encounterit.

This then is a very clear indication of union oligarchy threatening tenuous stepstoward institutionalization of the countermovement. To a certain extent, union oligarchyat the enterprise level was diminished particularly in terms of ends formulation: Liupursued particular goals in response to the wishes of membership. It is important to notethat these ends were pursued oligarchically, i.e. rank and file were not actively mobilizedin pursuit of these ends. And it was because of this inability to adequately develop thecollective power of workers that the victories in the Guangdong Union Hotel are sotenuous. But ultimately it was the oligarchy and heteronomy of the municipal andprovincial level unions that allowed for the activism of Liu to be effectively crushed. Theunion was heteronymous in this instance both vis-à-vis the state and in relationship to themarket. That is to say, the impetus for the union to tamp down grassroots activism and togenerate profit through market mechanisms overrode any capacity they might have todefend activist chairs within their jurisdiction. And oligarchic in the sense that high levelleaders refused to use their political and symbolic power to intervene in a situation inwhich they could have made a difference. Their passive repressive response waseminently political and held important symbolic consequences. Since this case was beingclosely watched they knew other potential activists would be aware of the outcome. Themessage was clear: we will not intervene on behalf of activist union chairs. Thisunwillingness to support grassroots activism is a strong indication of oligarchy, and acontinual threat to the construction of an institutional response.

But perhaps it unfair to blame Chen Weiguang for his failure to act in the LiuYongyi case. Maybe he wanted to intervene, but because it involved his direct superiorsfrom the provincial federation he was unable to. There is of course no way to know whatChen’s personal feelings were on the case. Given what I know about his character andpolitical disposition, it is quite likely he was upset by the outcome. But his feelings on thematter are immaterial, for what I am interested is political action. And if in fact he wasunable to act out of fears that he would upset his superiors, this reveals the fundamentalweakness of the entire Chinese trade union structure, predicated as it is on appropriatedrepresentation: officers are not beholden to their constituency, but to the political system.In this sense, we can see strong limitations on what progressive leadership, even those as

114 To be more specific, there are three steps in recalling a current union chair: 1) Morethan 1/3 of members must request a recall; 2) A congress of members or memberrepresentatives must be convened, with at least 2/3 in attendance; 3) More than half of thecongress participants must vote in favor of the recall.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (97)


prominent as Chen, will be able to achieve in terms of promoting the interests of theirmembership.

Finally, it is worth noting the mixed results that the GZFTU has had in removingmanagement from trade union leadership (one of Chen’s top priorities). The clearestexample of this comes from the formation of the prominent Tianhe District Retail SectorUnion Federation. As we will see in more detail in chapter five, the formation of sectoralunions has been at the top of Chen Weiguang’s agenda for a number of years, and he hasbeen personally involved in their formation. It then came as quite a surprise when theformation of the Tianhe district federation was announced in Novemeber of 2010 that thevice-chair was a man named Wang Honggang, the human resource director for the mega-chain Trust-mart (haoyouduo). It is also worth noting that Wang had been, andpresumably still is, chair of the enterprise-level union at the Trust-mart store. With ahuman resource manager as vice-chair of the new federation, management will have astrong voice in the development of the sectoral union as it expands from Tianhe district tocover the other districts in Guangzhou (as is planned). The selection of Wang as vice-chair was in clear violation of the 2007 “Measures for Implementation” which explicitlyban managers from serving as union chairs or vice-chairs. If three years after the ban waspassed management was still able to secure powerful positions in prominent unions beingformed under the direct supervision of the GZFTU, it is quite likely that such practicescontinue unabated in other enterprises. This is a strong indication that Chen Weiguang’sattempts at winning greater union autonomy from management continue to beconfounded by structural oligarchy.


In this chapter we have seen some of internal dynamics of the GZFTU, widelyconsidered to be the most progressive trade union federation in China. The ascension ofChen Weiguang to the position of federation chair is itself an indication that labor andpolitical leaders were searching for a response to rapidly growing worker insurgency.Under Chen’s guidance, the city has enacted a number of pro-worker initiatives, mostimportantly related to trying to win greater autonomy from management for enterpriseunion chairs.

However, as far as the institutional moment of the countermovement isconcerned, the results in Guangzhou are mixed at best. In the Hitachi plant, we see thatstate ownership, combined with capital intensive production and a skilled workforce, hasthe potential to result in a degree of decommodification for regular workers. However,the failure of the Hitachi union to win job security, decent pay, or benefits for internworkers, paired with workers’ complete ignorance of core union activities, are strongindications of the limits of such an approach. On the other hand, in the Guangdong UnionHotel, we have the impetus for decommodification coming from the bottom up in theform of enterprise-level union chair activism. However, when enterprise chair Liuencountered retaliation from management for pushing decommodifying initiatives, ChenWeiguang and other union leaders maintained a passive repressive position, and she wasremoved from her post. Thus, the material victories she had gained for her membershipmay come to be undermined by lack of a countervailing collective force on the shopfloor,i.e. failure to incorporate threatens economic advances.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (98)


In sum, it is evident that traces of the institutional moment of thecountermovement are emerging, even within a union structure which at first glanceappears hopelessly ossified and mired in stultifying oligarchy. In other words, we can seean attempt to respond to the insurgent moment of the countermovement within existinginstitutional parameters. However, when we look at two of the enterprises in whichrelatively decommodified and incorporated labor should be most solidly institutionalized,we see that the gains that have been made are modest, tenuous, and subject to ongoingthreats. This is due in large part to the state’s categorical ban on the development ofworker collective power at the point of production. Thus we can see the consequences ofthe contradictory nature of labor in China – after Chen Weiguang engaged in widelypublicized legislative attempts to reduce retaliation against activists union chairs, even achair in a union-owned hotel could not be saved from dismissal.

In this next chapter, we will see that sectoral unions are another attempt by theACFTU to circumvent the problem of worker power as they attempt to dictate the contentof regional collective contracts without engaging in worker organizing – what I will term“oligarchic decommodification.”

Chapter 5

Oligarchic Decommodification? Sectoral Unions and Crises of Representation

Thus far, we have seen several cases where ACFTU-subordinate unions have,despite great pressure from their membership, not played an active role in the realizationof decommodification. While the details of the specific cases vary, the general point isthat oligarchy has blocked incorporation and therefore stood in the way of the unionsbeing a participant in institutionalizing the countermovement. Worker unrest has failed toproduce a re-alignment of political power in society, and so the countermovementremains stalled at the “insurgent moment.”

When we turn to the case of Zhejiang, we find very different economic andpolitical dynamics than those in our previous cases. Specifically, the model ofdevelopment employed in the southeastern part of the province has relied on localentrepreneurial activity, and a majority of economic output is derived from suchbusinesses. These conditions have given the municipal governments and unions greatercapacity than those in Guangdong to organize entire industries in an attempt to rationalize

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (99)


employment relations. At first glance, it appears as if these conditions have allowed forunions to play a key institutional role in bringing about decommodification, the clearestindication of which is the conclusion of sectoral-level collective wage agreements. Thereason that this has been possible is because certain segments of capital demandedrationalization and needed the union’s partnership to bring it about. However, while theZhejiang and Guangdong cases are quite dissimilar in terms of the institutional capacityof the state/union to respond to instability in labor relations, the final outcome (i.e.decommodification and incorporation) is quite similar in both places. Although thesectoral agreements reached in Zhejiang seem to imply a degree of decommodification ona formal level, the lack of incorporation led to contract non-enforcement and ongoinghigh levels of commodification. This suggests that the state and union’s attempts tolegislate labor conflicts out of existence – what I will call “oligarchicdecommodification” – may fail, as illegitimate representative organizations cannotconvince their members to abide by agreements reached without their participation orconsent.

In this chapter I seek to first explain why it was that the Rui’an Eyeglass Unionand Wenling Wool Knitwear Union were able to play a key role in the conclusion ofsectoral-level agreements, while unions in Guangzhou have failed in such endeavors. Theanswer to this problem lies in an analysis of the distinct models of economicdevelopment, and the type of politics that such an economy generates. Finally, I willshow under conditions of appropriated representation, even if labor is decommodified atthe formal, contractual level, employers and workers alike frequently ignore suchconstraints on the market in practice. Thus, the union’s attempts at realizing “oligarchicdecommodification” have failed.

To set the stage, some background in China’s paths to development will benecessary.

Development and Labor Politics in China

The role of foreign direct investment (FDI) has been widely discussed in bothacademic and mainstream accounts of Chinese economic development over the past threedecades. Although there are some exceptions, the dominant narrative is that the Chinesegovernment has been able to secure hundreds of billions in foreign investment, whiledomestic industry has played a secondary and relatively minor role in promoting growth.Statements such as the following are representative of this line of argumentation:“…China's export-led manufacturing boom is largely a creation of foreign directinvestment (FDI), which effectively serves as a substitute for domestic entrepreneurship.During the last 20 years, the Chinese economy has taken off, but few local firms havefollowed….” (Huang and Khanna 2003:75) While some recognize the spatial and sectoralunevenness of FDI (Broadman and Sun 1997), the overall consensus is that foreigninvestment is good for growth (Chen, Chang, and Zhang 1995) and, despite possiblepolitical drawbacks (Eng 1997; Gallagher 2002; 2005), that it is a fundamental part of thestory of development in China (Tseng and Zebregs 2003).

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (100)


I do not wish to debate the validity of these claims. Indeed, total utilized FDImore than doubled from US$40.3 billion in 1999 to $92.4 billion in 2008,115 makingChina the largest FDI recipient in the world among developing countries. However, Iwould like to draw attention to Zhejiang province, an economically vibrant part of thecountry that has relied relatively little on FDI in its quest for growth. There are significantconsequences of this path to development for politics in the province.

The focus of my research is on labor politics, not on determinants of growth. Thereason I would like to introduce Zhejiang into the conversation on the Chinese labormovement is to highlight the relationship between models of development and thepolitical possibilities for labor. The overwhelming majority of literature on Chinese laborpolitics has focused either on the state-owned sector (Blecher 2002; Cai 2002; Chen2000; Chen 2003; Hurst 2004; 2009; Lee 2002) or the FDI-fueled Southeast (Chan 2001;Chan and Pun 2009; Lee 1995; 1998; Pun 2005; Thireau and Hua 2003), with ChingKwan Lee (2007) producing the only significant comparison of the two regions. Thedominant narrative about the role of FDI in Chinese development is really a story aboutthe “sunbelt” which is centered on the Pearl River Delta. Footloose transnational capitalhas been drawn to this region by its well-developed infrastructure, low labor costs,relatively healthy and well-educated workforce, and categorical ban on independentworker organization. This model of development has produced a certain set ofpossibilities for labor politics that I will not go into great detail about at the moment. Butvery briefly, workers in this area have engaged in frequent, cellular forms of protests,which have yet to result in significant decommodification of labor or a re-alignment ofpower relations in the determination of the labor process. The local state is committed toattracting and retaining foreign investment, and therefore its ability to push for theinterests of labor or to discipline capital is highly curtailed. As has been discussedpreviously, the inability/unwillingness of trade unions to win recognition from theworking class is indicative of this dilemma.

Now, when we turn our attention to Zhejiang, we see that the model ofdevelopment there has produced a somewhat different political environment, and that thepossibilities for the labor movement are therefore distinct from what one might thinkfrom a reading of the extant literature. While the local state in Zhejiang maintains thealliance with capital that we see in Guangdong, it has a significantly greater potential toorganize, if not necessarily coerce, employers. This is because a much greater percentageof employers in Zhejiang are locals and are organized into representative associations. Asa result, the state has many more political, administrative, and social tools at its disposal.This means that the state has greater capacity to violate the immediate interests of certaincapitalists in the service of greater long-term competitiveness for the sector as a whole.Worker insurgency in many sectors has pushed the state to search for a more rationalizedapproach to accumulation, the clearest example of which is the greater propensity forconclusion of sectoral-level collective wage agreements. While we should maintain noillusions that this represents a libratory or even particularly positive opportunity for thedevelopment of the Chinese labor movement, it does indicate that there are momentswhen the interests of the state and at least some segment of capital overlap with those of

115 PRC Ministry of Finance, cited on: (accessed 9/10/09)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (101)


workers, and that, under the right conditions, Chinese unions can play an importantinstitutional role in the realization of such contracts. Despite these important politicaldifferences, much as was the case in Guangdong the weakness of trade unions at thegrassroots level means that what the union wants to happen and what actually happensare two very different things.

Zhejiang and the Wenzhou Model

Although not nearly as well known internationally as Guangdong, Zhejiangprovince is in fact one of the most economically vibrant areas in China. Along with thePearl River Delta, it ought to be considered ground zero for the birthplace of post-MaoChinese capitalism. The success of the region in promoting economic growth is quiteastonishing. Although Zhejiang’s output of RMB 313 per capita in 1978 made it only the13th wealthiest province in China,116 by 2006 that number had skyrocketed to 31,874.With Jiangsu clocking in at 28,814 and Guangdong at 28,332, Zhejiang could claim to bethe most economically successful province in the country by a wide margin.117 WhileGuangdong’s official population of 86 million dwarfs the 46 million in Zhejiang,118 thesignificance of Zhejiang’s vault to economic pre-eminence among Chinese provincesshould not be overlooked.

The question now is, how is it that Zhejiang was able to accomplish such a task,and in particular, how did this path diverge from the typical story about reliance on FDI?Perhaps the most significant indicator of the extent to which Zhejiang differs fromGuangdong is the percentage of output that derives from foreign invested firms.119 In2003, a modest 20.1% of output in Zhejiang came from foreign firms, and that numbercrept up to 26.64% by 2007.120 While one quarter of the economy is clearly ofsignificance, it pales in comparison to Guangdong’s massive reliance on foreigninvestment. In the year 2000, 58.28% of output in Guangdong came from foreign firms,and it was up to 61.05% by 2007.121 For the three largest cities in the Pearl River Delta(the focus of my Guangdong study), the numbers are even more astonishing: 64.44% inGuangzhou, 67.88% in Shenzhen, and 77.85 in Dongguan.122 In Guangdong, foreigninvestment is not a just a significant part of the economy; it the foundation of theeconomy.

This tremendous divergence is of note in and of itself. But when we dip below theprovincial level of aggregation in Zheajing, we see that the story is a bit more complex.Zhejiang has traditionally been divided into a flat, prosperous, and centrally-locatedNortheast and a mountainous, poor, and remote Southwest. However, as Ye and Wei(2005) argue, over the past 20 years regional inequality within Zhejiang has increased

116 on 9/10/09)117 gonghui tongji nianjian 2007. Beijing: zhongguo tongji chubanshe.118 This of course includes firms from Hong Kong, Taiwan, and Macau.120 Zhejiang tongji nianjian 2008. Beijing: zhongguo tongji chubanshe.121 Guangdong tongji nianjian 2008. Beijing: zhongguo tongji chubanshe.122 Ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (102)


between the interior and coastal regions while decreasing between the traditionalNortheast/Southwest regions. Because I am interested in the most industrially advancedregions, I will focus for the moment on the Northeast and Southeast regions of Zhejiang.The Northeast includes the cluster of municipalities of Hangzhou, Jiaxing, Shaoxing, andNingbo, while the Southeast consists of Wenzhou, Taizhou, and Jinhua [see image 1].But why is this categorization of regions within Zhejiang important? Since the reason Iam interested in the province in the first place is because it is representative of a differentmodel of development from the Pearl River Delta, I of course would like to focus on theregions which best exemplify this model. In this case, the southeast, and in particularWenzhou, are representative of an indigenous, entrepreneurial model of economicdevelopment that does not rely on FDI.

This is not to say that the Northeast of Zhejiang, centered on the provincial capitalof Hangzhou, has been lackluster in its growth. Rather, it is that this region has manyadvantages that the southeast of the province does not. Hangzhou and its environs havebeen a major engine of economic growth for many centuries. The Grand Canal, firstcompleted in the 7th century, was built to link Hangzhou with Beijing in the north.Ningbo, just to the east of Hangzhou on the East China Sea, was the premier port city inthe region until it was eclipsed by Shanghai in the 19th century. Today, this regionbenefits tremendously from its highly developed infrastructure and proximity to both theinvestors and markets of Shanghai. These are advantages that municipalities in thesoutheast have not enjoyed as they have pursued growth.

In particular, the amount of foreign investment in the two regions is remarkablydifferent:

Year 2007123 Value of Utilization of

Foreign Capital (US$)

Value of Utilization ofForeign Capital per capita


Hangzhou 5,580,590,000 830Ningbo 4,501,070,000 797

Shaoxing 2,365,160,000 542Jiaxing 3,455,120,000 1,026


Wenzhou 1,049,444,000 137Jinhua 373,352,000 81

Taizhou 816,810,000 144

123 Zhejiang tongji nianjian 2008. Beijing: zhongguo tongji chubanshe. p.564; Zhejiang

Economic Census Yearbook 2004. Beijing: China Statistics Press. pp.7-16.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (103)


From the above data we see there is an unmistakable divergence between northeast andsoutheast in terms of both the absolute and the relative amount of foreign investment. Thetop recipient of foreign investment in per-capita terms is Jiaxing. While less glamorousthan the better known Hangzhou, Jiaxing municipality borders Shanghai. As a suburbansatellite of China’s largest metropolis, Jiaxing has a prime location for attracting foreigncapital, most of which has poured into manufacturing. Ningbo is also quite close toShanghai, and with the completion of the Hangzhou Bay Bridge in 2008, travel timebetween the two cities has been significantly reduced. And then there is Hangzhou, theprovincial capital with well-developed transportation infrastructure and easy access to thecapital and consumer markets of Shanghai and other cities in the Yangzi River Delta. Allof these cities in the northeast have the advantage of access to Shanghai’s enormous andhighly modernized Yangshan deep-water port, which is now vying for the title of busiestport in the world.

When we look to the southeastern municipalities, it is clear that the role of foreigninvestment has been considerably smaller than in the northeast. Although Wenzhou is aport city, it does not have the international stature of Shanghai, and is not even as large asNingbo. Physical separation, poorly developed infrastructure, and a relatively lowinternational profile make it unsurprising that the city that has had the most success withFDI, Taizhou, receives only 1/10 the amount of foreign investment per capita as Jiaxing.

These data hold important implications for a Guangdong-Zhejiang comparison. Itis true that aggregated at the provincial level, a much smaller percentage of Zhejiang’soutput comes from foreign invested enterprises than is the case for Guangdong. However,the three southeastern municipalities are much more exemplary of the local-drivendevelopment that I am interested in. Thus, for the purposes of this research I will focus onthe three municipalities of Jinhua, Taizhou, and Wenzhou.

The rise of Zhejiang’s Local Entrepreneurialism

The distinctive model of development found in southeastern Zhejiang started inthe city of Wenzhou. Although the city leadership was condemned for harboring pro-capitalist tendencies during the Mao era (Forster and Yao 1999:61), the so-called“Wenzhou Model” received support from the central government in the 1980s andquickly gained national fame. While the model was promoted as feasible for imitationnation-wide, it is really only in the neighboring municipalities in southeast Zhejiangwhere such emulation has proven highly successful. Despite the Wenzhou Model’sundeniable success in capital accumulation, its political feasibility in the formallysocialist China of the 1980s was far from guaranteed. But what is this “Wenzhou Model”which gained so much prominence in both the Chinese mainstream media and inscholarly research during the 1980s and 90s?

As mentioned previously, Wenzhou was quite isolated from the major hubs ofChinese social life for much of history. In the Mao era, state investment in the region wasminuscule, as the city’s proximity to Taiwan made it appear vulnerable to military attack.Additionally, Yia-Ling Liu (1992b) argues that the relative independence of the WenzhouCommunist Party in the revolutionary era often put local officials at odds with, andwilling to violate the directives of, the central leadership. Although these may only bepartial explanations, the fact remains that up until 1978, only 1% of provincial fixed

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (104)


capital investments went to Wenzhou, despite the fact that the municipality contains 15%of the province’s population (Forester 1990:56). While one should take into account thenegative effects of the Cultural Revolution for economic growth, in the years 1966-1978Wenzhou measured only a .1% average industrial growth rate (Parris 1993:244). In short,Wenzhou was a place with a rebellious political streak that had been, by and large,excluded from the fruits of the Maoist command economy.

At the end of the Cultural Revolution, things in Wenzhou started to change. Withpolitical leadership unable and/or unwilling to stop it, experiments with household-basedcommodity production began to take off. Whereas in 1980, only 1% of industrial outputin Wenzhou came from private firms, by 1988 that number jumped to an astonishing 41%(Liu 1992a:703). The impact on the labor market was immense: the percentage of therural labor force employed in agriculture dropped from 89% in 1978 to 37% by 1985(Dong 1989:79). Aside from a “privatization” of the economy and a partialproletarianization of the local populace, incomes jumped remarkably: “Average ruralincomes grew from 55 yuan in 1978 to over 447 yuan in 1985, still below the provincialaverage of 548 yuan but above the national average of 397 yuan.” (Parris 1993:249)Between 1981 and 1985 the municipality’s gross industrial output increased by 130%Forster 1990:57). An area that had been poor, not just by Zhejiangese standards, but bynational standards, was turning into an economic dynamo, and it was local privateenterprise that was fueling the transformation.

But what precisely was the structure of these firms that were quickly gainingrenown throughout the country? Calvin Chen (2008), who has done perhaps the onlyethnography of enterprise development in southeast Zhejiang (both in Wenzhou andJinhua), talks about closely-knit, relatively egalitarian organization of production in theearly years of the enterprises he studied. In particular, high levels of social capital werecrucial in promoting the development of the firm: “These young workers, even theirparents if need be, could speak directly with factory officials if they felt dissatisfied.Many did so with ease. These meetings were not simply exchanges between employerand employee; rather, they were discussions between familiar parties who interactedsocially outside the workplace.” (Chen 2008:73) With firms rarely exceeding 100employees and more typically employing around 35 (Parris 1993: 247), all of whomwould be hired from the locality, there was deep social integration between workers andmanagement in Wenzhou firms. This was a double edged sword: sometimes it meant thatworkers would accept delayed payments of wages “for the good of the enterprise,” butother times it meant that managers would work on the assembly line to help quickly fillorders. Chen argues that the socially embedded nature of production allowed, “membersof the workplace [to] understand and willingly fulfill the duties and responsibilities theyare assigned.” In the view of Burawoy (1982), this might sound like the utilization ofdense social ties in the construction of a hegemonic labor regime; but for Chen it allowedfor healthy enterprise development. Regardless of one’s normative assessment of such asituation, the point is that, during the early-mid 1980s, small scale, socially embeddedenterprises that enjoyed a high degree of managerial autonomy (Bramall 1989) quicklybecame the bedrock of Wenzhou’s local economy.

In addition to a reliance on existing local social networks for the recruitment ofworkers, entrepreneurs also depended greatly on support and protection from localgovernment officials. The first benefit the state provided these entrepreneurs was actually

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (105)


to do nothing: they looked the other way as informal, and indeed illegal, capital marketsbegan to develop in the municipality (Tsai 2002). Informal “credit associations,” usuriousmiddle men, and family borrowing all emerged on the scene in the 1980s, clearlyviolating the legal monopoly of the state banks. Then, in 1986, someone formally openedup a credit union with variable interest rates, and was only able to overcome oppositionfrom state banks because of protection offered by local government and Party officials(Parris 1993:248). Government officials’ support of private enterprise also extended toassisting them with the difficult task of registering their businesses. Since originally therewas no such thing as a “private enterprise,” entrepreneurs had to be very creative withtheir formal status. The practice of “wearing the red hat” referred to a procedure wherevillage governments (which are below the municipal level) would register privateenterprises as “collectives.” This would allow the enterprises to appear politically correct,and also generated revenue for local government officials. Yia-Ling Liu has argued thatthis coincidence of interests between entrepreneurs and local officials made it possible tofirst resist pressure from higher levels of the state, and then to sustain high levels ofgrowth for many years.

But when this support from the local state became most important was in thepolitical struggle that emerged as the Wenzhou Model gained national attention. WhileDeng Xiaoping had initiated market reforms in 1978, these centrally mandatedexperiments spatially quarantined capitalist relations of production into a few specialeconomic zones. Wenzhou presented a much more serious challenge to the still-dominantsocialist ideology in the sense that capitalist-style credit, labor, and commodity marketshad sprung into existence without any management or affirmation from Beijing.Although Wenzhou had received visits from central authorities as early as 1983, it wasnot until the 1985 visit of reform-minded Premier Zhao Ziyang that locals felt they had astrong sign of support from Central leadership. That being said, there was still heateddebate about the emergence of what was undeniably an experiment with capitalist laborrelations. Supporters argued that Wenzhou’s strength in realizing the development of theforces of production justified the increased wealth disparities and exploitation thatinevitably arose. One well-known economist foreshadowed an increasingly uneasyrelationship between the Party and Marxism by arguing that exploitation did not exist inprivate firms in Wenzhou because the firms were still regulated and taxed by a socialiststate. Detractors held that the Wenzhou model was tantamount to the reintroduction ofcapitalism into China. 25 years later, there is little doubt that they were correct in thatassessment. However, the key point is that Wenzhou officials’ political protection andsupport allowed local enterprises to flourish in such a way as to prove their advantages atpromoting capital accumulation in poor, remote areas. The material success of Wenzhouentrepreneurs resulted in a victory for pro-market Party leaders. The consequences forChina’s transition to full-blown capitalist labor relations have been profound.

What became of Wenzhou?

After Wenzhou received a flurry of scholarly attention in the late 1980s and early90s, the city and its model of development fell out of the spotlight. Perhaps this wasbecause capitalist labor relations were no longer controversial, and the more traditionalhubs of Chinese economic and political power put economic reform into high gear. In

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (106)


1984, 14 major coastal cities were opened up to foreign investment, and Shanghai’sPudong district was established as a special economic zone (SEZ) in 1990 (Ge1999:1282). After Deng Xiaoping’s “Southern Tour” in 1992, reform continued toaccelerate, and the emergence of private enterprises continued throughout the country.Soon, national and international attention was focused on the economic juggernauts ofthe Guangzhou-anchored Pearl River Delta124 and the Shanghai-anchored Yangzi RiverDelta. Wenzhou’s dependence on private enterprise for economic development was nolonger controversial, nor did it seem particularly extraordinary.

On the one hand, the Wenzhou elite had great cause for celebration: theirresistance to political pressure and perseverance in “following the capitalist road” hadended victoriously. The entrepreneurs who bucked Maoist orthodoxy ended up lookinglike heroic trailblazers by the 1990s, and their once-derided political sponsors were seenas visionaries. But it was a bitter-sweet victory: the flipside of China’s full embrace of themarket was that Wenzhou-based enterprises suddenly faced much stiffer competition thanthey had when they first started out. The family-style firms that were innovated inWenzhou soon spread to neighboring municipalities in southeastern Zhejiang, mostnotably Jinhua and Taizhou. Foreign invested factories, equipped with deeper reserves ofcapital, more advanced technology, more rationalized forms of management, and largereconomies of scale began encroaching on their markets. While many Wenzhouenterprises had a first-mover advantage in their respective sectors, they would have tochange in order to survive.

The 1990s have been referred to as a time of expansion (Chen 2008) andrestructuring (Wei, Li, and Wang 2007) for Wenzhou industry. The very small,household-based production that characterized the Wenzhou Model in the early daysgave way to increasingly large, differentiated, and rationalized firms in the 1990s.Although many firms re-organized into share-holding enterprises, in most cases singlefamilies maintained majority stakes. Companies began setting up branches throughoutChina, and in some cases even relocated their headquarters to Hangzhou and Shanghai inorder to secure better access to human and financial capital (Wei, Li, and Wang2007:438). The informal kinship networks that were crucial for marketing in the 1980sneeded to be updated and rationalized. In order to survive, most firms had to conquer newdomestic markets and begin to expand into international ones.

At the same time, internal labor process reforms radically reformulated employee-employer relations. Speaking of the two firms he studied in Wenzhou and in neighboringJinhua, Calvin Chen has said that,

… enterprise leaders moved away from the norms of reciprocity, social trust, andcollective well-being that had been integral components of the company’s previousstrategy… The rough sense of egalitarianism that had previously existed was soon replacedby an ever-widening gulf between the skilled and the unskilled, the managers and themanaged. (2008:89)

124 Of course Guangzhou’s role in “anchoring” the Pearl River Delta is less absolute thanthat of Shanghai’s in the Yangzi River Delta. In the PRD Shenzhen was the first SEZ,and Hong Kong has provided the lion’s share of the investment. Such “competitors” donot exist in the same way in the Yangzi River Delta, where the supremacy of Shanghai isunquestioned.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (107)


Additionally, Wenzhou locals were less and less interested in the relatively low-paid,exhausting, and frequently dangerous work on the shopfloor. As was the case in othercoastal cities in China, these firms began recruiting from the waves of migrant workerswho arrived in the city. The rationalization and concomitant dissolution of bonds ofsocial solidarity within the factory meant that capital-labor relations became increasinglytense. Open conflict, the type of which was very rare in the household-productionenterprises, burst into the open, and strikes, legal disputes, etc., all increased infrequency.

The final section of Chen’s 2008 book Some Assembly Required focuses on theproblem of “reintegration” in such firms: “The core task of reintegration is tosimultaneously consolidate improvements in production capabilities made during theexpansion stage and restore the tight-knit, trust-filled relationships of the [earlier] era.”(2008:126) In both Chen’s research and my own investigation, enterprise managers andstate actors expressed a strong desire for some sort of re-integration. And the desire tohave such re-integration unmistakably derives from anxiety over increasing laborconflicts and social instability. In marked contrast to their foreign counterparts in thePearl River Delta, entrepreneurs in Wenzhou and other areas in southeast Zhejiang beganactively seeking the assistance of the Communist Party and other party organizations,most notably the trade union, in an attempt to regain stability in production.

Labor Relations in Zhejiang

I was first alerted to the activities of Zhejiang trade unions in casual conversationswith officials from the Guangzhou Federation of Trade Unions (GZFTU). In September2008, Chairman Chen Weiguang led a delegation from the GZFTU in visiting five citiesin eastern China (including Wenzhou) to learn about that region’s experiments withindustrial (chanye) and sectoral (hangye) trade unions. Upon returning, members of thedelegation were glowing in their praise for Zhejiang unions, with the official internalreport saying, “In thirty years, Guangzhou unions have been at the forefront of reform,have developed their own characteristics, and have had enormous success. But we cannotbe proud and complacent. This visit has allowed us to see our shortcomings andgaps….”125 In the view of one GZFTU official, there was no question that the success ofZhejiang unions in establishing sectoral unions had to do with greater governmentsupport for their activities. This was a mutually beneficial arrangement for union andgovernment as, “The union there knows how to help the government do things.”126 Theofficial contrasted this with the situation in Guangzhou, where he unequivocallyexpressed that the local government was not nearly as supportive.

This subjective assessment of the superiority of Zhejiangese trade unions inpromoting the development of sectoral level unions and wage negotiation is supported byofficial statistics:

Data on








#workerscoveredby WCC






125 This report was provided to me by an officer of the GZFTU.126 Interview, October 2008

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (108)






coveredby WCC

by WCC coveredby




Zhejiang 31,723,800 84,878 3,689,895 6311 346,413Guangdong 52,500,900 41,101 2,075,095 4062 84,889

We can see that, even in absolute terms, the number of workers covered by a WCC or asectoral-level WCC is significantly higher in Zhejiang than in Guangdong. But in relativeterms, the disparity is even more striking. Whereas in Zhejiang 11.6% of workers arecovered by WCCs and 1.1% by sectoral-level WCCs, in Guangdong it is only 4.0% and0.16%, respectively. It is true that a relatively small number of workers are covered bysectoral-level WCCs in both provinces. However, at over 1%, a significant number ofZhejiangese workers are covered by such contracts, while the 84,889 workers inGuangdong under sectoral-level WCCs is a truly paltry sum. In addition to the greater success at concluding WCCs, there are several otherindications that Zhejiangese unions wield more power than their counterparts inGuangdong:

Legal and


activities of

the union,


# of workers inthe legal aidorganizations

of union

# of unioncadres with


# of casesaccepted by

legal aidorganizations

of union

# of localstatutes

participated inby union abovegrassroots level

Zhejiang 1006 127 479737 (1st incountry)

Guangdong 683 74 147215 (4th incountry)

What I am specifically interested in is the relationship between oligarchy anddecommodification. Although wage levels, pensions, benefits, etc., are one potentialindicator of decommodification, I want to focus on the conclusion of sectoral-levelcollective contracts and their economic and political consequences.

Sectoral-level collective negotiation requires at least two representative parties,namely labor and capital. In order to understand what has happened in Zhejiang, it isimportant to first understand something about the unique nature of employer associationsin the region.

Zhejiangese employer associations

Zhejiang, and in particular the southeastern area centered on Wenzhou, is famousthroughout China for its entrepreneurial spirit and success at developing the local privateeconomy. Although largely ignored in the English-language literature, a significant

127 zhongguo gonghui tongji nianjian 2007. Beijing: zhongguo tongji chubanshe128 ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (109)


amount of research has been conducted by Chinese scholars on the noteworthydevelopment of independent employer associations in the region, with the best examplesto be found in Wenzhou (Wang 2004; Wu 2004; Yu, Xu, and Jiang 2007). Both thesestudies and my own research reveal that the emergence of these associations hasimportant consequences for politics in the region, and that they are also a strong indicatorof the highly classed nature of political participation in China. These relativelydemocratic organizations allow for member participation in organizing and expressingcapitalist class interests, an opportunity that is not available to any of the dominatedclasses in China. However, they also provide a bargaining partner for trade unions in theirattempts to engage in sectoral-level wage negotiations. While this is not necessarilypositive from the perspective of workers, it does mean that it is easier for the state torationalize and organize production within a given sector in a way that is simply notpossible in areas such as the Pearl River Delta (where employer associations do exist, butwith much lower density and legitimacy in any given sector).

In typical Leninist fashion, China has a nationally organized employerassociation, the All China Federation of Industry and Commerce (ACFIC). As with othermass organizations, such as the trade union, it is subordinate to the Party and its structuremirrors that of its parent organization. As is the case for workers and the ACFTU, mostemployers have little regard for the ACFIC. In most instances, ACFIC organizations willbe dominated by large state-owned enterprises (SOEs), with government or SOE officialsinstalled as the chair. These are typically set up to try to establish and enforce certainstandards within an industry (which will frequently benefit the SOEs), and therefore aremuch less beneficial to smaller private employers. While there have been moments whensuch organizations behave in a corporatist manner (Pearson 1994), they are certainly notthe type of representative organizations found in other countries.

In Zhejiang, on the other hand, a new type of employer association (hangye xiehui

or shanghui) has emerged, which from a formal level, looks a lot more like therepresentative, class-based organization one might find in a liberal democracy. Whilethese organizations still must be formally “attached” (guakao) to a parent state organ,their organizational structure and decision-making procedures are formally autonomous.They do not receive any funding from the state, but rather are supported by member dues.While there still are some instances in which the government will try to appoint employerassociation chairs, their leadership selection process is usually internally determined. Asone indication of this trend, in 2003 77% of employer associations in Wenzhou reportedthat their chairs were selected according to their own internal rules (Yu, Huang, and Fang2004:37). While such organizations are of course interested in having good relationshipswith the government, and will often invite government officials to hold honorarypositions, they are more directly concerned with their membership’s interests than is thecase for unions.

As has already been pointed out, industrial development in southeast Zhejiang hasbeen highly dependent on the emergence and success of local, privately ownedenterprises. Although today there is more foreign investment compared to the 1980s, it isstill a relatively small amount. This fact is reflected in the membership of employerorganizations, which are endowed with very high levels of social capital. In describingthe process of setting up the eyeglasses employer association in Rui’an, theorganization’s director said that since everybody already knew everybody, it was a

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (110)


relatively straightforward procedure.129 Of the nearly 100 eyeglasses manufacturers in thecounty-level city, only one of them is a joint venture (with Taiwanese investors), andthere are no fully foreign owned firms. Even more noteworthy, I was told that only two ofthe Chinese-owned enterprises are owned by people who come from different cities; thus,the sector is overwhelmingly dominated by locals.130 A trade union official in Rui’ancommented that members of the employer associations would listen to and obey theirchair since, “their relationships are like brothers.”131 The thick social networks tietogether not just employers to other employers, but also the business associations to thestate. With retired government officials sometimes serving as officers in the employerassociations, representatives from each group frequently have long-standing relationshipswith each other. What this means is that the state and union may have not justadministrative, but also social, resources at its disposal in its interactions with localcapital.

There has been debate about the political implications of these organizations, andwhether they herald the coming of a robust civil society in China (Fewsmith 2005). Withwidely disparate definitions of “civil society” this is not a straightforward question toanswer. But as regards my own research, the relevant political question is, to what extentdoes the formation of employer associations in Zhejiang present an opportunity for thestate to more effectively organize capital? At one time, there was great debate in the Westabout unequal capacities for labor and capital to expresses class interests throughrepresentative organizations (Offe and Wiesenthal 1980; Streeck 1991; Traxler 1993).But given that labor is characterized by appropriated representation and that politicalspace is highly constrained, this cannot be the question in China. In Zhejiang, we have acase where we see that the organization of capitalist interests is the prerequisite for thestate, through the auspices of the trade union, to try to establish sector-wide standards.The formation of representative employer associations in Zhejiang has importantimplications for labor politics because it provides the union with a bargaining partner. Onthe other hand, the state in the Pearl River Delta is confronted with a disorganized andhighly mobile set of capitalists; under such conditions, it is incredibly difficult toestablish wage agreements for an entire sector. This makes it unlikely that conscioushuman action, rather than the whims of the free market, will determine the substance ofthe labor process and the price of labor power. In southeast Zhejiang, on the other hand,we have a situation where certain segments of capital decide that a more rationalizedapproach to production is in their interest, and they are organized into associations whichallow for this interest to be expressed. This allows capital, in cooperation with the state(again, through the auspices of the union) to establish a more coordinated system ofproduction which could result in decommodification of labor. Let’s now turn to the datato see how this process has played out empirically.

Trade Unions and Decommodification in Southeast Zhejiang: The Official Story

129 Interview, July 2009130 Interview, December 2009131 Field Notes, July 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (111)


Starting in 2004, trade unions in certain industries in southeast Zhejiang began tosuccessfully conclude sector-wide wage agreements that cover nearly all enterpriseswithin a given city. This model has been most successful in relatively labor-intensiveindustries, where there are a large number of small local employers concentrated in aparticular district. In all of the cases that I am aware of, the final agreement on sector-wide wage standards was only possible through the coordinated efforts of agencies of thestate, as well as the trade union and relevant employers association. While the trade unionhas played a key role in signing such agreements on behalf of workers, the impetus forsetting wage standards has come largely from employers and the state after they havebecome concerned about high turnover and excessive labor conflicts. In order to addressthis problem, an organizational innovation in the form of sectoral trade unions wasnecessary. The establishment of sectoral unions mirrored the already existing employersassociation, the primary difference being that employer associations are actuallymembership based, while the unions remain an extension of the state. The end result hasbeen the successful conclusion of sector-wide agreements in several places in southeastZhejiang. In this next section we will turn to two cases – the previously unstudied Rui’aneyeglass union, and the much-ballyhooed (within China) Wenling wool union – in orderto understand the political dynamics behind that state’s attempt at “oligarchicdecommodification.”

Given its prominence in the media and Chinese scholarly literature, the data onthe Wenling case come from secondary sources. The general economic and politicalconditions are very similar between the two cases, but Rui’an has been less celebrated inthe media. In this sense it is a better case because the exceptional attention from the stateand media likely would have an effect on the operation of the trade union in Wenling.However, I include data from the Wenling case to demonstrate that regional conditionsallow for a particular type of trade unionism, and because it represents the highestaspirations of the ACFTU. Data on Rui’an come from interviews with workers andmanagers as well as officers from the Eyeglass Employers Association, Eyeglass SectoralUnion, Rui’an Federation of Trade Unions, and the Rui’an Labor Bureau.

Collective Bargaining in Wenling

Perhaps the earliest, and certainly the most widely discussed, experiment withsectoral-level collective bargaining took place in the county-level municipality ofWenling, less than an hour away from Rui’an by car. Just as Rui’an is encompassed bythe municipality of Wenzhou, Wenling is within the administrative jurisdiction ofTaizhou municipality. As was described earlier in the chapter, Taizhou’s model ofdevelopment has been very similar to that of Wenzhou, with a preponderance of localenterprises serving as the foundation of the economy. The establishment and functioningof the Wenling Wool Knitwear Union has been quite similar to that of the Rui’anEyeglasses Union.

The Wenling wool industry grew very rapidly in the 1990s, and many localentrepreneurs jumped into the market. As of 2004, the industry in Wenling employed

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (112)


12,000 workers in 113 enterprises, with a total yearly output of 1 billion RMB.132 Most ofthese enterprises are rather small in size. By the first part of this decade, nearly all of theworkers in the industry were migrants and approximately 90% of them were female.Although the industry developed quickly, it encountered fierce worker resistance andinstability in the labor market. Speaking of the first few years of the 2000s, one laborofficial said, “At that time, workers were constantly arguing with their bosses aboutunreasonable wages, and every day there would be employees jumping ship (tiaocao).”133

In 2002, the industry experienced approximately one conflict involving ten or moreworkers every single day.134 The highly seasonal nature of the work meant that skilledworkers enjoyed strong market position in the May-December high season, during whichtime they could force employers into bidding wars. As one worker in the wool industryput it, “Jumping ship became the only method workers had to resist and defend theirrights.” 135 Faced with short lead times and slim profit margins, these high rates ofturnover and frequent conflicts were of concern to a significant number of employers.Though not framed in precisely such terms, there was a strong incentive to bring aboutrationalization in the labor market. Both large employers and the state (who, as was thecase in Rui’an, was concerned with the development of the sector as a whole) wanted totake action.

The process of setting up a sectoral level wage agreement was not straightforwardat all, despite the fact that conditions in Wenling were much more propitious than mostplaces. In early 2003, the Wool Textiles Employers Association met with leaders fromthe Wenling Federation of Trade Unions to discuss the matter. The idea of trying toestablish a sectoral level wage agreement gained support, and the two parties began tomove forward with full support from the local government. With such support from stateand capital, the establishment of the Wool Knitwear Union was a rather simpleadministrative procedure. However, it is of little surprise that some individual employerswere not so enthusiastic about this development. The trade union official who was taskedwith investigating wages in the sector (with the end goal of being able to establishcommonly agreed upon piece-rates) was originally turned away by employers when hecame to ask to look at their books. In the end, however, the much more powerful localParty committee told these employers that they had to cooperate with the union, and theissue was thus resolved. At the conclusion of the investigation, the union delineated fivemajor categories of work and 59 individual work procedures. However, the piece ratesbeing offered for identical procedures varied greatly between enterprises, with thedifference sometimes as high as 1 RMB. The next task of the union was to try to figureout a piece-rate system that all parties involved could agree upon.

Wage negotiations were, of course, complex. In June of the same year, workerrepresentatives were selected by management to participate in an officially sponsored“frank discussion” on establishing uniform piece rates. There were differences in opinion

132 December 31, 2003. “wenling xieshang hangye gongzi wending laodong guanxi.”(Wenling sectoral bargaining stabilizes labor relations). Gongren ribao.133 September 22, 2008. “wenling gongzi xieshang jieya laozi maodun.” (Wenling wagenegotiations relieve pressure from labor-capital conflicts). Minzhu yu fazhi shibao.134 Ibid.135 Ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (113)


on many items, but after distributing more than 500 survey questionnaires, holding sixnegotiation sessions, ten “frank discussions,” and conducting three wage adjustments, afinal agreement was reached between the union and the employers association.136 OnAugust 8th, the union and the employers association signed the “Wool Knitwear SectoralWage (piece-rate Agreement),” with the agreement that wages would be re-negotiatedonce every year.

From the government’s perspective, the sectoral collective wage agreement wasan unqualified success, as they claimed that turnover and conflict in the industry droppeddramatically. In the final months of 2003 after the new wage agreement wasimplemented, there was a dramatic drop in official complaints (shangfang) coming fromthe wool industry, with only 11 such incidents involving a total of 120 workers. From2004 to 2005 there were only three complaints involving 11 workers, and from 2005 to2006 there was only one complaint involving three workers. Then, in 2007, workers inthis sector did not file a single official complaint.137 Additionally, since 2006 there hasnot been a single labor dispute (jiufen) over wages. With the government claiming thatlabor strife was way down from its highs earlier in the decade, the Wenling governmentcould make a strong claim that their model of industrial relations was a perfectembodiment of “harmonious society.”

In fact, it was not long before the Wenling union won national recognition for itsactivities. As early as 2003, Wenling’s accomplishments had been recognized by itsnational-level parent industrial union.138 After gaining repute within the trade union, theso-called “Wenling model” was officially approved by Premier Wen Jiabao, whoproclaimed in a work report that, “Wenling’s approach can be summarized andpopularized.”139 The following month, Zhejiang’s Party secretary advised that PremierWen’s advice should be followed, and that the popularization of the Wenling modelwould start first and foremost in Zhejiang. Constantly on the lookout for practicalexamples to demonstrate that “harmonious society” and “win-win” labor relations aremore than just slogans, both the ACFTU and the Party-state more broadly was veryimpressed that the Wenling union had managed to reduce conflict without affectingproductivity.

The Rui’an Eyeglass Union

136 April 14, 2008. “zhashi tuijin gongzi jiti xieshang: Zhejiang wenling hangye gongzijiti xieshang jishi.” (Solidly promote collective wage negotiation: an account ofZhejiang’s Wenling sectoral collective negotiations) gongren ribao.137 March 31, 2009. “Zhejiang jiceng minzhu de zhidu chuangxin.” (Breakthroughs inZhejiang’s grass-roots democratic system) xuexi shibao.138 August 4, 2009. “Yan Jiamin: nuli kuoda hangyexing jiti xieshang fugaimian.” (YanJiamin: work hard to increase the coverage of sectoral collective negotiations) Gongren

ribao.139 April 16, 2008. “wenling zhigong gongzi yu qiye xiaoyi tongbu zengzhang.”(Wenling’s employee salaries and enterprise productivity increase together) taizhou


Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (114)


Though largely unknown outside the borders of the city, the case of the Rui’anEyglass Union is very similar to the Wenling model. Rui’an is a county-level city of 1.17million residents that is administratively subordinate to the Wenzhou municipality140. Itsper-capita output of 31,525 RMB141 is significantly higher than most cities in China,though it is almost precisely the same as the provincial average for Zhejiang. As one ofthe more industrialized areas within Wenzhou, it exemplifies the “Wenzhou model” ofdependence on small, family owned enterprises. Rui’an is best known for its productiveindustries including shoes, textiles, bags, small consumer goods, etc. However, the focusof my research is the eyeglass sector that is centered in the small township of Mayu andemploys more than 12,000 workers. For a number of years the city government has beenintent on becoming the pre-eminent eyeglass manufacturing center in the world, and theyhave begun to actively recruit international eyeglass brands to establish offices in Rui’an.

After the first eyeglass factory was established in 1978, the eyeglass sector inRui’an grew quite rapidly for a number of years.142 Many enterprises that started ashousehold-based units expanded in the 1980s and 1990s, and the total number of eyeglassand eyeglass accessory manufacturers increased as well. Relying on the renownedWenzhou-style family-based marketing networks, these producers first conquereddomestic markets, and then expanded internationally, eventually gaining customersthroughout the world. By the late 2000s, sales from the Rui’an eyeglass industry totaledmore than 10 billion RMB, accounting for half of all nationwide eyeglass sales, and theEyeglass Employers Association estimated that 60% of all eyeglasses worn in China wereproduced in Rui’an.

But it was the very success of the industry in the 1990s which began to lead to adeterioration of employment stability in the industry. More and more entrants to themarket increased competition for skilled labor. Armed with relatively strong marketplacebargaining power (Wright 2000), skilled workers began pitting employers against eachother in an effort to increase their wages. Smaller, younger enterprises did not have thefinancial or knowledge-based resources to conduct the comprehensive worker trainingprograms that the larger and more established companies held. One result of this was thatthese smaller enterprises began to pilfer skilled workers from other enterprises byoffering marginally higher piece rates. In a formal address, the chair of the EyeglassEmployers Association presented this state of affairs as a very serious problem:

… the need for technical personnel and skilled workers has been steadily increasing withinthe industry, and using high salaries in hiring has been an inevitable decision forenterprises. “Cutting the ground out” from each other is already not a new phenomenon,and this led to lack of order and chaos in enterprise employment. Workers have beenjumping ship without order for personal gain and lured by the promise of [higher wages].This has disturbed order in the industry and destroyed the normal functioning ofenterprises. Labor conflicts are obviously up, and there have been instances of using

140 2008 nian rui’an shi guomin jingji he shehui fazhan tongji gongbao (2008 report onRui’an’s economic and social development). on 10/15/09)141 ibid.142 The data in the proceeding section are derived primarily from official uniondocuments.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (115)


certain articles of the Labor Contract Law to continuously engage in appeals, file lawsuits,and commit blackmail [yaoxie]. According to the social security authorities, labor conflictshave been occurring continuously in the industry, which has led to a worsening of theemployment environment and the positive development of the industry has been restricted.

While such a situation is not necessarily disadvantageous for skilled workers, largeemployers were much more concerned. Thus, in 2001 a few players in the industry tookthe first tentative steps toward addressing these problems. Two of the largest eyeglassfirms in Rui’an took the initiative in trying to establish sector-wide standards, and theybegan to appeal to the Eyeglass Employers Association for assistance.

The Rui’an Eyeglass Employers Association was previously a formal state-sponsored organization. Eventually, the local government decided to “give themfreedom,”143 which meant both greater organizational autonomy but also impliedresponsibility for raising their own operating expenses. Since that time, the associationhas operated as an independent civil (minjian) organization, 100% dependent onmembership dues for its operations. The organization has served as a platform (pingtai)for eyeglass manufacturers in Rui’an, allowing them to communicate common intereststo the government, and also allowing the government to have an efficient way ofcommunicating with the business owners. Often new laws and regulations that are ofrelevance to the industry are disseminated through the employer association. As oneofficial from the local labor department said, “These are things the government used todo but they gave this responsibility to the association.”144

The push for a sector-wide standardization of wages has been one of the mostsignificant tasks taken on by the employers association, and surely required intensenegotiations between firms which may have had conflicting interests. It is notcoincidental that the CEO of one of the largest manufacturers in the city, Chen Chengwei,was also the chairman of the Eyeglass Employers Association. The instability in the labormarket was most problematic from the point of view of large employers, since smallerenterprises would frequently poach their skilled workers. In coordination with a few otherlarge manufacturers, Chen used his position within the association to begin to push forwage standardization, presumably against protests from smaller employers. Eventually,conversations within the association and with government agencies led them to the ideaof collective wage negotiations. As a sympathetic employer argued, “Every year in thebusy season, workers would go on strike, and nearby manufacturers would not hesitate touse high salaries to poach skilled workers. If we did not sit down and bargain, it would bedifficult for the entire eyeglass industry to continue on.”145 The problem now was that theemployers needed a bargaining partner.

It was at this point that officials from the Rui’an Federation of Trade Unions(RAFTU) became involved. Following the lead of the Wenling Wool Knitwear Union,the decision was made to establish an eyeglass sectoral union. Leaving absolutely nodoubt about where the initiative for such a thing came from, the official document

143 Interview, July 2009144 Interview, July 2009145 February 5, 2008. “Zhejiang goujian hexie laozi guanxi chutan.” (first explorations inZhejiang’s construction of harmonious labor relations) ban yue tan.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (116)


approving the establishment of the Rui’an Eyeglass Union is addressed to the EyeglassEmployers Association. The first two lines of the document read as follows:

Rui’an City Eyeglass Employers Association:After reviewing your request to establish an eyeglass sectoral union, we approve of

the establishment.

The remainder of the document lists the names of the officials who will staff the unioncommittee, and says that they will hold office from August 2005 until July 2010.

The next item of business was to conduct collective wage negotiations. The unionconducted an investigation into wage levels in the various eyeglass enterprises, speakingwith managers and workers and conducting a survey. The information was analyzed andthe union put forth a proposal for unified piece rates for 318 specific work operations.After soliciting comments and a holding a few negotiation sessions, in September 2005 afinal contract was signed by the respective chairmen of the employers association and thesectoral union. The contract was approved by the labor department, and thus becamelegally binding.

As already mentioned, internal tensions between different manufacturers are notentirely clear. Officials from the Eyeglass Employers Association have tried to put apositive spin on such intra-sector competition. Although appearing reluctant to discussthe issue, the current chairman of the association did admit that some small employersrefused to participate in the sectoral wage agreement. Since the employer association isnot a government agency, they could not “force them to sign.”146 Additionally, headmitted that of the 130 eyeglass manufacturers in Rui’an, about 30 of them are notmembers of the association. That being said, cooperation between the employersassociation, trade union, and perhaps most importantly, government and Party officials,eventually resulted in a binding collective wage agreement.

The actual content of the contract is very limited, as there are only five articles.Article 1 simply states the two year time period during which the contract is in effect.Article 2 is the meat of the contract, and says that workers will be remunerated accordingto the piece rate established in the “2009 Rui’an City Eyeglasses Employers AssociationEnterprise Piece Rate Form.” Article 3 says that workers must be paid on time, but that ifan enterprise is having “difficulty,” payment of wages can be delayed after consultationwith the union. Article 4 states that the contract is legally binding, while article 5 saysonly that disputes should be handled according to the law.

Since the conclusion of the sectoral wage agreement, union leadership says thatthey have continued to develop their organizational capacity within the industry. At thetime of the establishment of the eyeglass sectoral union, many manufacturers in the cityhad not yet established enterprise-level trade union organizations. With the help of theemployer association, union officials have been making efforts to address this issue, withthe belief that the sectoral union will be more effective with enterprise-level branches inplace. Additionally, in June of 2009 the preparatory committee for an eyeglass sector“staff and workers’ congress” (zhidaihui) was established. Particularly in state-ownedenterprises in China, the staff and workers’ congress has been, in theory, a mechanismwhereby employees can directly participate in the determination of the labor process.

146 Interview, July 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (117)


Aside from clichéd rhetoric about “democratic participation,” it is unclear precisely whatfunction the eyeglass congress will serve. But given that Chen Chengwei, former chair ofthe employers association, was appointed as director of the preparatory committee for thestaff and workers’ congress, it is almost certain that it will remain under the control ofemployers.

Four years after the conclusion of the collective wage agreement, the unions,government, and employers were very pleased with the results. The official story is thatprevious illegal actions on the part of employers, such as holding employees’ ID cards ortaking deposits have been eliminated. Turnover was greatly reduced, which stabilizedproduction. Additionally, bosses can no longer unilaterally determine wages, but ratherhave to engage in the official process of negotiation. Finally, and perhaps mostsignificantly, the number of labor conflicts has dropped significantly.

In official accounts, both Wenling and Rui’an are great success cases where laborconflicts have been reduced and production has been rationalized, all of which wereaccomplished through administrative means. The state and higher levels of the tradeunions have been quite impressed with such reports, and have gone out of their way toshower praise on such experiments. Zhejiang: exemplary model of labor relations?

As was revealed in the descriptive statistics earlier in the chapter, Zhejiang hassignificantly more workers than Guangdong that are covered by collective wageagreements, and an even greater (relative) edge in sectoral-level agreements. This successhas not gone unnoticed, as the state has proclaimed for many years the implementation ofcollective contracts as a primary goal. Indeed, the legal framework for theimplementation of collective contracts in China was firmly established with the labor lawreforms of 1994 (Clarke, Lee, and Li 2004; Warner 1995), but the union’s ability tobargain and enforce such contracts has been highly circ*mscribed. Thus, the apparentsuccess of Zhejiangese trade unions at concluding collective contracts was received withgreat enthusiasm from both the Party-state and from the national leadership of theACFTU.

On a formal level, at least, the Zhejiang government made a strong commitmentto promoting collective wage negotiations. In late summer of 2008, the provincialgovernment announced the goal that 70% of private enterprises and 100% of state-ownedand collective enterprises be covered by such agreements by the end of 2010.147 The stateappeared to be taking this task seriously, as all government agencies and Partycommittees have been exhorted to support this work. To put some muscle behind thedemand, the provincial Party Committee added success with collective negotiations intotheir procedure for evaluating the accomplishments of government officials. Of coursethe relative weighting of such accomplishments in comparison to GDP, attractinginvestment, maintaining social stability, etc., is unclear, but the fact that it exists at all issignificant. While some of these plans were likely temporarily derailed by the collapse of

147 September 8, 2008. “Zhejiang to promote collective wage negotiation.” (accessed October 2009)

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (118)


the export economy in late 2008 and 2009, it is undeniable that the experiences of placeslike Rui’an and Wenling convinced the provincial authorities of the viability anddesirability of collective negotiation.

Zhejiang’s apparent commitment to this project garnered support from the highestlevels of the trade union as well as attention from academic researchers. Aside from thefrequent affirmations from Party leadership that have already been mentioned, theprovincial Party committee and provincial government held a conference in Wenling inMarch of 2008 to learn about their experiences with sectoral bargaining. The followingmonth, the ACFTU sponsored a national-level conference on collective negotiation thatwas held in Zhejiang’s capital city of Hangzhou.148 Entitled “National Union CollectiveWage Negotiation Exchange,” the conference was attended by the ACFTU’s number twoin command, Sun Chunlan, as well as the vice-chair of the province’s Party Committee.The fact that a national-level conference was held outside of Beijing is indicative of theextent to which ACFTU leadership was willing to try to elevate the experience ofZhejiang unions. In addition, collective negotiations in Zhejiang have begun to attract theattention of semi-official scholarship in China (Chen and Huang 2008; Zhu 2008), andeven foreign researchers (Liu 2009).

It is clear that the state and unions were enthusiastic about the perceived successwith collective bargaining in Zhejiang. But the question remains: to what extent was theofficial rhetoric realized in practice? In order to answer these questions, I went beyondthe official statements and interviews with union and employer association leadership totalk with factory management and workers in the Rui’an eyeglass industry.


During my first brief foray into the field in Rui’an, I only interviewed officialsfrom the labor department, trade union, and the employers association. Betweenconversations with these officials, and the documents they provided me with, the story Ipieced together was compelling in its innovativeness. Faced with instability inemployment relations, the union, government, and employers came together and throughnegotiation and compromise, reached an agreement on wage level standards for the entiresector. It was a true win-win-win for workers, the state, and employers. But uponspending a bit more time in the field, I discovered there was just one problem: thecontract was not being enforced.

My first realization that things on the ground were not as I had been led to believecame in a very awkward interaction with a manager from a small factory in MayuTownship. I had been brought to this factory by my closest contact in the eyeglassindustry, Ms. Du, a human resource manager at Zhilian Eyeglasses who was also thechair of the enterprise’s trade union branch. I had first met Ms. Du in a meeting with thechair of the Rui’an Eyeglass Employer Association, when I had been given an official,and overwhelmingly positive, account of the development and results of sectoral levelbargaining in Rui’an’s eyeglass industry. In previous conversations, she had frequently

148 April 11, 2008. “quanguo gonghui gongzi jiti xieshang gongzuo jingyan jiaoliu hui zaihang zhaokai.” (National union collective wage negotiation exchange conferencecommences in Hangzhou). Zhejiang ribao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (119)


touted the achievements of both the employer association and the union in bringing aboutthe breakthrough of a sectoral-level wage agreement. However, when she brought me tovisit the neighboring Huangwei factory (directly across the street from Zhilian), adifferent picture began to emerge. With Ms. Du sitting at my side, I began chatting withHuangwei’s general manager Mr. Guo about the factory’s operations. When I inquiredabout the sectoral-level collective wage agreement, he looked puzzled and asked me whatI was talking about. I tried rephrasing the question, but to no avail. At this point Ms. Duinterrupted and said, “He wants to know how you set your piece rates.”149 Finally Mr.Guo’s face lit up in recognition, but his response had absolutely nothing to do with thecollective agreement to which his factory was supposedly bound. In fact, Mr. Guorevealed that Huangwei doesn’t use piece rates at all (as dictated by the collectiveagreement), but rather uses hourly rates. Fearful that I had just embarrassed my mostimportant contact in the industry, I quickly moved to change the topic.

Mr. Guo – ignorant as he was of the wage agreement – was by no means anoutlier among those in the Rui’an eyeglass industry. As I interviewed manager aftermanager over the next several weeks, I discovered that not only were very few employersabiding by the agreement, hardly any of them had even heard of it. Managers from thelargest enterprises with up to 500 workers, and those from firms with only 80 employeeswere equally confused by my inquiries about collective wage negotiations. Evenmanagers from Zhilian Eyeglasses (one of the two large enterprises that promoted theidea of a sectoral collective wage agreement) told me during a factory visit that theirwages were higher than those of other neighboring factories, indicating that they were notabiding by any industry standards.

It will come as little surprise that workers were equally unaware of the existenceof such an agreement. In my many evenings hanging out with workers at the pool tablesin Mayu’s industrial zone, I quickly learned that the topics of collective contracts andtrade unions would elicit no response. On my first night in Mayu, I innocently enoughasked a few workers from a nearby glasses factory about the collective contract. Not onlywere they unaware of the existence of a collective contract, they said that in their factoryworkers did not sign contracts at all, but rather it was based on “trust.”150 Some workersdid know what a trade union was, but nobody had any direct experience with them.

The employers in the eyeglass industry were familiar with the concept of “tradeunion,” but – in contrast to what I had been told by the Rui’an union officials – very fewof the enterprises had established a union branch. Many employers claimed that unionswere only set up in the larger enterprises, because they are more “by the books”(zhenggui). One manager from a small enterprise with fewer than 100 employees had thisto say about a union branch: “At the moment we don’t have one, we can’t set it up,”151

indicating that establishing a union requires resources that smaller enterprises do nothave. Somewhat curiously, he then articulated a position widespread among Chineseworkers and intellectuals: “In China, I’m not sure if the union has any function(zuoyong),”152 and went on to say that unions in Europe accomplish more things. A

149 Interview, December 2009150 Field Notes, December 2009151 Interview, January 2010152 ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (120)


manager from one of the largest firms told me in response to an inquiry about a unionbranch that they did not have one, but that they had set up a Party branch. When askedwhat the Party does in the factory, he merely laughed and said, “they can help somepeople,”153 but nothing more. While the municipal level union officials had claimed thatthey used the negotiations over the collective wage agreement as an opportunity toestablish more enterprise level union branches, I did not detect even a suggestion of neworganizing initiatives.

As mentioned previously, there are a small number of enterprises in Rui’an thatdo not participate in the employers association, and therefore would not be covered bythe wage agreement. However, this was not the case with the firms that I had contactwith, as all of them confirmed that they were members of the employer association. Thisis not surprising, as I was based in the epicenter of Rui’an’s eyeglass industry, Mayutownship. It was in Mayu were there were the highest levels of social capital, and wherethe activities of the employer association were focused. Employers’ assessments of theassociation varied from mild to enthusiastic support. Most commented on the associationhelping them to learn about new policies and new market opportunities, but nobodymentioned their role in establishing industry standards. All members had to pay dues thatwere adjusted based on the size of the enterprise. Thus, employers’ lack of familiaritywith the collective wage agreement cannot be attributed to their non-participation in theemployers association.

The lack of enforcement of sectoral wage standards was further highlighted byemployers’ frequently expressed angst over one of the main problems that the agreementhad originally set out to resolve: high turnover of skilled workers. Senior managers fromboth small and large enterprises complained about workers jumping ship. One managerfrom an enterprise with 110 workers described the problem as “very serious,” and “verytroublesome,” but that, “there is nothing to be done… [workers] have freedom.”154 Asone would have expected, large employers were even more agitated about the pilfering ofworkers. Mr. Wu, a veteran of more than 20 years in the eyeglass industry, had workedfor the two largest eyeglass manufacturers in the city, Zhilian and Huakai. Both Zhilianand Huakai generally have workforces exceeding 300, though both have at times reachedabove 500. Mr. Wu described in detail how these two enterprises are some of the onlyfactories that have extensive worker training programs. Additionally, they provideworkers with a base wage even in the off-season in order to retain talent, something thatthe small enterprises cannot afford to do. However, during the busy season, smallenterprises will try to offer marginally higher piece rates to their skilled workers to lurethem away. According to Mr. Wu, the first question these small enterprises will askpotential employees is, “Have you worked for Zhilian or Huakai?”155 as this is a strongindication of how well-trained they are. In general, nobody claimed that the problem ofjumping ship had improved in recent years (in the time since the conclusion of thecollective wage agreement), with some saying that things hadn’t changed and othersexpressing that the situation had gotten much worse.

153 Interview, December 2009154 Interview, January 2010155 Field Notes, December 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (121)


Though not directly related to either union activities or the wage agreement, thelabor market dynamics in Rui’an exerted a significant influence on employers’ capacityto retain skilled labor. Perhaps the single most pervasive complaint from employers wasnot that the economic crisis had reduced the number of orders they received, but ratherthat they could not hire and retain enough workers. In some cases, manufacturers wereunable to accept orders because production was running so far below capacity. On onefactory visit, I strolled through a shopfloor on which less than half of the machines werebeing used. The manager that was accompanying me motioned to the machines and said,“as a manager when you look at this… it’s such a waste.”156 Employers both large andsmall were distraught at their inability to recruit workers, and the industrial zone in Mayuwas plastered with “help wanted ads” everywhere.

In sum, it is clear that the sectoral-level wage agreement that was so highly toutedby officials from the Rui’an Federation of Trade Unions and the Eyeglass EmployersAssociation was not being enforced. This, despite the fact that the enforcement of such anagreement could help resolve the serious problem the industry as a whole faces, namelyhigh turnover of skilled workers and labor conflicts. Despite a strong possibility ofbuilding a cross-class alliance, and the exertion of extensive efforts among officialrepresentatives, the agreement exists on paper only. And yet, when we turn to theexperiments with sectoral unions in Guangzhou, we see that the state and union’sresponse is, at the formal level, very different.

Shortcomings of Sectoral Level Trade Unionism in Guangzhou

As mentioned previously, union officials in Guangzhou held the activities ofZhejiang sectoral unions in high regard, and frequently expressed frustration at their owninability to achieve similar success. Guangzhou Federation of Trade Unions (GZFTU)Chairman Chen Weiguang said that the development of sectoral unions was one of theprimary tasks of the GZFTU in the years going forward. Reform-minded as he is, Chenwas particularly enamored of sectoral unions because they presented an opportunity tocreate greater independence between unions and employers.157 The chair of a sectoralunion would not be employed in a particular enterprise (as is the case for most primaryunits of the trade union), a scenario which could potentially give them greater freedom toimpose demands on employers. And yet, Chairman Chen complained that the work ofsectoral unions in Guangzhou was greatly constrained by the fact that they do not have a“partner,” by which he meant there is no party with which they can negotiate collectivecontracts. One of the vice-chairs of the GZFTU was quite explicit about the problemwhen describing why in Guangzhou the unions were only focused on collectivebargaining at the enterprise level: “our sectoral unions are not mature, and… theemployer associations are not mature. They [employer associations] are even less mature!So we don’t have an opponent.”158

During my time in Guangzhou I was able to learn about the activities of the twosectoral unions that the GZFTU considered the most successful: the construction and

156 Field Notes, December 2009157 Field Notes, December 2008158 Field Notes, October 2009

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (122)


sanitation worker unions. The construction union leadership claimed that they were themost developed sectoral union in the city, with their organizing committee officiallyestablished in October 2007. They originally set up a union only in the central Liwandistrict, but then expanded to cover the remaining districts in Guangzhou. With 2,000construction sites and around 400,000 workers in the city, this was no small task.However, when asked about what specific activities they conducted, it was not clear thatthey were performing any sort of a representative function. The union cadres did say thatthey provided workers with legal assistance, and disseminated information on workers’legal rights. But the activities beyond this remain quite limited. Construction union cadresconfirmed that the GZFTU had suggested they look into sectoral-level bargaining, butthey admitted that no progress had been made on this front. Their explanation was thatdifferent employers had their own systems for calculating wages, and that unifying payscales would be very difficult. Reverting to free market ideology as an explanation, onecadre argued that, “The market economy is a presupposition, so there isn’t any need tointervene here [by negotiating wages].”159 Even more alarming, the construction uniondoes not even have the capacity to collect dues, but rather is fully supported by theGZFTU. Indicating the difficulties Chinese unions have faced in the process ofmarketization, the chair of the construction union complained, “In the past theconstruction sector was all state-owned. And we got the dues. But the system reformed,and it is very relaxed [i.e. unregulated] now and we can’t collect the dues.”160

Deregulation also created significant problems for the union’s attempts toorganize sanitation (huanwei) workers161 in the city. Originally all sanitation workers inGuangzhou were directly employed by the city. However, after entry to the WTO in2001, the government deregulated the industry and contracts were given to a variety ofprivate companies. By 2008, there were more than 600 companies involved in sanitationwork in Guangzhou, many of them owned by companies from Hong Kong, Taiwan,Malaysia, and Singapore.162 With privatization came worsening conditions for manyworkers, and in the spring of 2008 street cleaners in one of the industrial districts ofGuangzhou went on strike. Shortly thereafter, the GZFTU decided to establish a streetcleaners’ sectoral union, starting where the workers had gone on strike in Baiyun district.However, by December of the same year, little progress had been reported, withChairman Chen saying that they needed the “government’s strength” to organize theemployers into a representative organization.163 The following October, the chair of thecleaners union still had little claim to success, saying only that they had continued toexpand their presence into more districts of the city. However, in the summer of 2009,several months after I left my field site in Guangzhou, it was reported in the media thatthe Luogang district union concluded the city’s first sectoral level collective contract. Thecontract was reported to cover only six employers, which is less than 1% of the 600+

159 Interview, April 2009160 Interview, April 2009161 The category of “sanitation worker” is broader in China that would be the case in theAmerican context, as it includes not just garbage collectors, but other workers involved incleaning activities such as street sweepers, janitors, etc.162 Field Notes, December 2008163 Field Notes, December 2008

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (123)


employers operating in the city. Although information remains highly incomplete since Iwas not able to independently verify the content of the contract, the media report did notmention anything about increases in wages or benefits, but emphasized that relevant lawsand regulations would be better enforced.164

It is worth emphasizing that all of the sectoral unions in Guangzhou that were inoperation or in the planning stages (e.g. restaurant and hotel unions) were in the servicesector. In contrast to Zhejiang, there has not even been an attempt to establish sectoralunions in manufacturing sectors. In general, one would expect organizing sectoralagreements in manufacturing to be even more difficult that in the service sector, sinceservice work is place-specific. Even given this seeming advantage, neither of the twomost prominent sectoral unions in Guangzhou had achieved anything close to the formalsuccess of the sectoral wage agreements that are becoming more and more common inSoutheast Zhejiang.

At this point it is clear that the formation of sectoral unions is high on the agendafor the ACFTU. However, there has yet to be a critical analysis into the followingquestions: Why have Zhejiang unions been more successful than others (notably those inGuangdong) at reaching sectoral-level wage agreements? What do the particularitiesabout the Zhejiang case imply for sectoral level bargaining in other regions in China?Why were union and state alike unable to enforce the sectoral agreements that theyworked so hard to produce? And finally, has collective negotiation significantlydecommodified or incorporated labor in any of the industries in which it is prevalent? Itis these questions to which we will now turn.

Oligarchic Decommodification?

As compared to other regions of China, the volume of worker protest has beenparticularly pronounced in Zhejiang, Guangdong, and other economically dynamicregions where manufacturing industries employ huge numbers of migrant workers. Bothat the local and national levels of the state, social instability has become a major concern.In response to this, Hu Jintao’s “harmonious society” platform is simultaneously: 1) Anofficial recognition of the fact that Chinese society is actually not at all harmonious; and2) An attempt at establishing submission and consent to the re-aligned relations of classdomination which have emerged in the process of marketization. While the central stateis more concerned with maintenance of social order, local governments are anxious thatworker insurgency may affect capital accumulation and the “investment environment.”Thus, as disorganized, sporadic, and apparently apolitical as worker unrest has been inChina, this decentralized social movement has already pushed the state to makesignificant political adjustments. Because of the peculiar nature of appropriatedrepresentation, insurgent worker energy creates strength at higher levels of the unionwhich can be translated into administrative power; thus, labor appears strong. Butbecause workers in any given workplace remain atomized, the administrative success ofhigher levels of the union is frequently undermined; here, labor is quite weak. Thisdynamic has been apparent enough in the cases at hand.

164 July 21, 2009. “huanwei gongren qian jiti hetong hushen.” [sanitation workers signcollective contract for protection] Yangcheng Wanbao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (124)


And yet, we have seen that even the administrative strength of unions vis-à-visemployers varies according to region. Explaining why union and state responses havebeen different n Zhejiang and Guangdong is the purpose of this chapter.

As brief review, both Zhejiang and Guangdong have experienced similar levels ofworker insurgency but have had different capacity to address this crisis. Since there havebeen no systematic studies of labor protest in Zhejiang it is difficult to know if thecharacter of this unrest differs significantly. However, as far as government and unionofficials’ perception of the unrest is concerned, there is no significant variation betweenthe two regions. In both places there is a general sense of “chaos” (luan), “unstable laborrelations” (laozi guanxi bu wending), and lots of conflicts (jiufen). Union officials in bothplaces also understand this is a problem which must be resolved, and that it is their task topromote the development of “harmonious labor relations.” While I would not go so far asto claim that Zhejiang’s labor relations are more stable in general, the cases of thesectoral unions are an indication that Zhejiang has had some formal success that has beenelusive in Guangdong. In particular, trade unions in Zhejiang have been more successfulat setting up sectoral level unions and concluding sectoral level wage agreements. WhileGuangdong unions remained reactive to worker protest, Zhejiang unions managed, in atleast a few cases, to establish at least the formal appearance of a more stable form oflabor relations.

My explanation for why sectoral level wage negotiation has been possible inZhejiang but not in Guangdong is straightforward: it is because of the two region’sdistinct models of development and composition of capital investment. Additional butsubsidiary factors are the existence of representative employer associations, high levels ofsocial capital, and the geographic density of enterprises in a particular sector that can befound in both Rui’an and Wenling. I have gone to lengths in this chapter to demonstratethat Zhejiang, and in particular the three southeast municipalities of Wenzhou, Jinhua,and Taizhou have been largely dependent on indigenous entrepreneurs in their process ofdevelopment. While the local state’s degree of “autonomy” is up for debate, the localcharacter of capital allows for it to be highly embedded (Evans 1995). In addition, thehigh level of social capital amongst capitalists has provided the union with arepresentative bargaining partner in the form of local employer associations. Given thehigh density of a particular industry in a very small geographic region, township-levelgovernments are highly dependent on the sustainable development of the given industryin order to maintain their tax base, while the highly diversified economy of the PearlRiver Delta stands in stark contrast. The combination of these various factors meant thatthe local state had both an interest, and a capacity, to work towards an administrativeresolution to instability in labor relations.

Thus, because of the character of the Zhejiangese political economy, we appear tohave the possibility of “oligarchic decommodification” as represented by the top-downimposed sectoral-level wage agreements. In these cases, the trade union continues torespond to the demands of the state – even if the state is simply expressing the interests ofcapital. The state is interested in such a project because of the relative immobility of localcapital, and because it is in the interests of a relatively large segment of employers.Additionally, it is “oligarchic” in the sense that such a program does not involveconsultation with, or participation of workers. Membership was never asked if theythought the sectoral agreements were a good idea, nor was there any attempt to build

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (125)


relatively autonomous working class power. To return to my original conceptualframework, the development of sectoral agreements was oligarchic in both endsformulation and in the means employed to pursue these ends.

Oligarchic decommodification is appealing to the central government as itrepresents a possible method for reducing social instability and enhancing accumulationwithout actually giving dominated social groups the ability to exercise collective power.Wen Jiabao’s enthusiastic support for Wenling is a strong indication that this model is infact quite alluring to high-level authorities. Of course, given the non-enforcement of theagreement in Rui’an there are questions as to the extent that decommodification hasactually been realized in practice, an issue which we will address in a moment.

But before getting involved in an analysis of outcomes, we can see that based onthe political economy of Guangdong (as represented by its capital Guangzhou), there aresevere structural impediments to even attempting oligarchic decommodification. In linewith Atul Kohli’s argument about the difficulty of directing foreign capital towardsnational goals (2004:382), in Guangdong trade unions are completely incapable ofestablishing any standards at the sectoral level. The sectoral agreement eventually signedin Luogang district only covers six employers out of more than 600, and appears not tospecify wage levels (to say nothing of other benefits, workplace rules, etc.) Capital in thePearl River Delta is cosmopolitan and footloose, and is not organized into the type ofrepresentative organizations that exist in Zhejiang. These investors have no social bondsor sense of obligation to the locality, and are always willing to threaten capital flightwhen the state tries to increase worker protections. While such enterprises do encounterthe same issues with labor conflict that we see in Zhejiang, a pro-union strategy is notsomething that they have pursued, in large part because of the difficulty in organizingforeign capital. Even if rationalization of employment relations were something thatmade sense to some large employers, it would be very difficult for them to impose sector-wide standards on smaller employers. The inability of the state/union in Guangdong todiscipline transnational capital has negative implications not only for the politics ofdecommodification, but also for future growth potentials (Chibber 2003).

There are important consequences of this dynamic when one considers theapplicability of the Zhejiang experience to other regions in China. The question ofapplicability is significant as Wen Jiabao and ACFTU leaders have specifically called fornational promotion of the Wenling experience. Union leadership is particularly keen topromote these cases because they “prove” that it is possible to reduce conflict throughadministrative processes, rather than directly engaging with and mobilizing membership.But given the relative peculiarity of the model of development of southeast Zhejiang, theability of unions in other areas to follow in their steps is questionable. Indeed, when oneunion official from Guangzhou was asked about applicability of Zhejiangese unionactivities to his work, he responded, “If they have some experiences, I will study it. But Iwon’t accept it just because Hu [Jintao] likes it.”165 In addition to the fact that nearly allof the businesses in the Rui’an and Wenling cases are locally owned, they also had theadvantage of having a high concentration of small enterprises in a very compactgeographical area. This meant that local government had a strong incentive to focus onthe development of one particular industry, as much of their tax revenue depended on the

165 Field Notes, December 2008.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (126)


industry’s success. Such conditions do not hold in cities in the Pearl River Delta, addingyet another obstacle to establishing sectoral-level wage negotiations. The implication isthat without the unique combination of economic and political conditions that exist insoutheast Zhejiang, it will be very difficult for trade unions in the manufacturing sector inother parts of China to emulate the sectoral-level wage agreements found in places likeRui’an and Wenling.166

The question remains, however, as to what extent the Wenling and Rui’an casescount as “successes.” Specifically, to what extent were decommodification and/orincorporation of labor realized? At this point it is clear that these cases have not producedmeaningful advances in either workplace security or social protection. There are nogreater job protections, no improvement in benefits such as pensions, health insurance,housing allowances, etc., and no wage scales or seniority. The Rui’an contract evenspecifies that employers have the right to unilaterally violate the contract (and the law):delaying wage payments is deemed permissible granted that the employer is suffering“difficulty” (which is not defined) and that they “consult” the union. The union retains nolegal right to reject delayed wage payments. Additionally, there is no mechanism forresolving disputes over interpretation of the contract prior to filing with the labordepartment. The one possible exception is that there have been reports that wages formany workers in Wenling have increased somewhat substantially since theimplementation of the sectoral wage agreement (though this may very well be due to atightening of the labor market rather than union intervention). The area in which sometentative progress could theoretically exist, counter-intuitive as it may be, is in “laborprocess control.” Whereas previously, piece-rates in the Rui’an eyeglass and Wenlingwool industries were determined according to market principles, the establishment ofsectoral wage agreements submitted the determination of the price of labor power tocoordinated and conscious human action. True, this “conscious human action” ultimatelywas very undemocratic and took little account of the interests of workers themselves. Butfrom a formal standpoint, the process of negotiation by which piece-rates weredetermined implies a rejection of complete submission of the satisfaction of worker needsto the logic of the market.

However, when we move beyond an analysis of formal outcomes to interrogatethe substantive consequences of divergences in institutional responses between Zhejiangand Guangdong unions, we see that in neither province was labor actually

decommodified. The sectoral wage agreement in Rui’an’s eyeglass industry is not beingenforced, and therefore has failed to alter the methods by which the price of labor isdetermined. The rise and fall of wages in the industry are still determined by the dictatesof the free market, and coordinated and conscious human action does not play a role.Despite the efforts of officials from the trade union, government, and employer

166 It is, however, more likely that sectoral unions will have more success in the place-specific service industry. As just one example, a food and beverage union in Wuhanbargained a collective contract covering 450,000 workers in early 2011. See, May 3,2011. “Wuhan 45 wan canyin congyezhe tanpan shixian zuidi gongzi shangfu 30%.”[450,000 food and beverage workers in Wuhan collectively bargain a 30% increase inbase wage] Guangzhou Ribao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (127)


association, the pricing of labor in Rui’an’s eyeglass industry is just as described by theGuangzhou construction union cadre where “the market economy is the presupposition.”

Why is it that the attempt at oligarchic decommodification failed? The first majorfactor is, ironically, the very thing that made the institutional breakthrough possible in thefirst place, namely reliance on state support. The union was quite happy about theoutcomes from Rui’an and Wenling because they were able to win a contract by relyingentirely on the symbolic and administrative power of the state. As is well-known, theunion is categorically opposed to mobilization of its own membership, a precondition thatbecomes incredibly troublesome when trying to combat the increasing power of capital insociety. Without this resource, the union officials that do take their job as workerrepresentatives seriously are frequently frustrated by their lack of power, since they findit nearly impossible to make any demands against the objections of management. As oneGZFTU staff member said, ““we [the union] lack means to put pressure on the bosses. Ifwe say to the boss, ‘sign, sign, sign the contract!’ and he refuses there’s nothing we cando.”167 With an impotent union at the point of production, it is only through state supportthat victories can be secured. In an environment where the local state is by and largeconcerned with short-term accumulation, there is a natural tendency for a state-capitalalliance, a phenomenon which clearly bodes poorly for decommodification. What wasunique in Wenling and Rui’an is that the interests of a significant segment of capitalhappened to overlap with those of the state in bringing about rationalization, anepiphenomenon of which could have been a modicum of decommodification. Thus, theunion, working at the behest of state and capital (rather than the specifically articulateddemands of membership), was able to play a key institutional role in bringing about thesectoral agreement.

Why is it that the thing which made this development possible (i.e. state support)also a weakness? Well, as has been argued extensively, the state at the local level isstrongly beholden to capital, and is rarely capable of genuine autonomy. In fact, thereason the state supported the formation of sectoral unions in these cases was becauseworker insurgency was making life too difficult for capital, and something had to bedone. In essence, the union (backed by state power) was the means by which labor andcapital’s essentially coincidental interest in rationalization was brought about. Theimplication is that, in situations when decommodification is not in the interest of asignificant portion of capitalists within a given sector, that state support for the union willevaporate (and there is a huge amount of evidence to support this claim). The state is anunstable ally for the union; with workers continually atomized, the sustainability ofsectoral level wage agreements will be completely dependent on the state. As thechairman of the Wenling Wool Knitwear Union said to a visiting German delegation:“We can’t be like you and just go on strike. We depend on the support of the Party andgovernment.”168

Thus, the primary reason that the contract in Rui’an was not enforced was becauseof a crisis of representation in both the union and employer association – i.e. failure toincorporate labor. Claims of representation are often tenuous and subject to challenge;

167 Field Notes, December 2008168 September 22, 2008. “wenling gongzi xieshang jieya laozi maodun.” (Wenling wagenegotiations relieve pressure from labor-capital conflicts). Minzhu yu fazhi shibao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (128)


this is one of the central problematics of politics. But just as we have seen was the case inGuangdong, the Rui’an eyeglass union’s claim to represent and negotiate on behalf ofthat sector’s workers was an act that was indifferent to membership recognition. To referto eyeglass workers in Rui’an as “members” is even a bit of an overstatement, as they didnot pay dues, were not aware of the fact that they were members, and were completelyunaware of the fact that representatives were acting on their behalf. The union did exertquite a lot of effort in negotiating the contract, but this performance was directed at othergovernment agencies, the employers association, and high levels of the trade union, not

towards their supposed membership. The unsurprising consequence is that workers havenot even gotten to the stage where they can be upset with the fact that their union hasnegotiated a bad contract for them since they don’t know it exists (this would be animprovement, as it could then incite a struggle between represented and representers). Inshort, we can see that how inseparable the political and economic features of theinstitutional moment are; without incorporation of labor, decommodification can beextremely difficult to secure, even if it has the support of the state and a segment ofcapital.

The crisis of legitimacy within the Rui’an Eyeglass Employers Association issomewhat more puzzling. Members in this association pay dues, are aware of theassociation’s activities, and are generally quite supportive of these activities. Everyemployer I came into contact with in Rui’an was a member of the association, and theyfrequently had their membership certificates prominently displayed in the factory offices.And yet, these employers were unaware of the collective wage agreement to which theywere theoretically bound. From the perspective of the association chair, the successfulnegotiation of this contract was one of the, if not the, most important project theyundertook in the past few years. And yet, it was not that employers were consciouslyviolating the terms of the contract because they felt it did not suit them. Rather, they werejust like the workers in that they were – with a few exceptions – unaware of its existence.

While the challenges for the union in gaining recognition are severe, what canexplain such a failure of legitimacy in the employers association? Though any answermust remain provisional given the lack of transparency in official maneuverings in theChinese polity, it would seem that the explanation could be found through an analysis ofthe power dynamics within the association as well as its relationship to the state. First ofall, it would not be surprising to learn that the Rui’an Eyeglass Employers Associationlacked the institutional muscle to force all of their members to abide by a wageagreement that seriously harmed their bottom line. The attempt to impose sector-widewage standards clearly plays to the advantage of large employers who have investedsignificantly in worker training programs and enjoy a better economy of scale. Somecoercive power would be required to convince smaller employers to abide by such anagreement. But if such coercive power did not exist, than why would the employerassociation go to such lengths to negotiate an agreement they knew would not beenforced? It would appear that the motivation would come from the promise of somesymbolic profits in the eyes of the state. While the conclusion of such an agreement ismore clearly beneficial to the trade union (in the sense that it will please their superiors)there are also potential advantages for the employer association. Always mindful oftrying to promote “scientific development,” and “harmonious society,” the employerassociation may be eager to demonstrate to the state that they are willing to make

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (129)


compromises with the union in order to reduce labor conflicts. While this did not actuallyhappen in practice, the conclusion of a formal agreement gives them something concretethey can show to concerned government officials who will likely see it as something ofan accomplishment. Given that few agents of the state have the time or the inclination toindependently verify on-the-ground outcomes, the formal appearance of activity is likelysufficient to win accolades. Such a positive review from the state (which has certainlybeen the case for the Rui’an municipal government) can pay numerable dividends downthe line.

While my fieldwork was focused on the Rui’an eyeglasses union, it is likely thatthe same conditions hold in Wenling. Despite the huge amount of national attentionfocused on the city’s wool industry, employers and union officials alike expressed deepanxiety about the enforceability of the wage agreement. As one boss in Wenling worried,“We signed a trade-wide wage standard, we must respect it; but these workerrepresentatives were all appointed by the government and enterprises, will they be able torepresent the aspirations of all the workers? If other workers do not respect the standardand continue to demand a wage increase what can be done?”169 The chairman of theWenling Wool Knitwear Employers Association expressed nearly identical concerns: “Ican unite everyone to go and control those employers that do not follow the wageagreement. But who can guarantee that workers will follow the agreement and won’t goon strike again?”170 These comments reveal that employers understand the lack oflegitimacy the union has with workers. Interestingly, a trade union official from theTaizhou Federation of Trade Unions (the body immediately superior to that of theWenling Federation of Trade Unions) expressed the mirror image of this concern: “Thesectoral union represents workers in concluding the wage agreement with the enterprise.But what happens if the boss ignores it? There are no specific measures, and no legalprotection, we just have to rely on the good will of the employer.”171 So, while employersare concerned that workers will not respect the contract because of the union’s lack oflegitimacy, union officials fear that employers will violate the agreement because there isno check on their power.

What then, are the consequences for theory? At first glance, it appeared as ifoligarchy and the countermovement did not stand in stark opposition to each other, butrather that a degree of decommodification could be oligarchicaly administered by stateand trade union if worker resistance was exacting a high enough cost on capitalaccumulation. The sectoral-level collective wage agreements reached in both Wenlingand Rui’an indicate a small degree of “labor process control” on the part of the unions,

169 January 25, 2008. “How bosses and workers can become ‘one big happy family’: areport on an inquiry into the introduction of collective wage consultation in WenlingCity, Zhejiang.” Chinese Labor and Social Security News. (translated by China LaborNews Translations)170 May 1, 2008. “wenling hangye gongzi jiti xieshang zhisu – yige laowu guanxi xin

jumian de kaichuang”[The Wenling Sectoral Collective Wage Negotiaion System – TheStart of a New Phase in Labor Relations.] guancha yu sikao.171 September 22, 2008. “wenling gongzi xieshang jieya laozi maodun.” (Wenling wagenegotiations relieve pressure from labor-capital conflicts). Minzhu yu fazhi shibao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (130)


even if such control is exercised in a non-democratic manner. After my first trip toRui’an, it seemed as if, given the unique characteristics of Southeast Zhejiangese politicaleconomy, oligarchic decommodification could be a reality.

However, on deeper investigation, it is clear that neither decommodification norincorporation of labor have been realized in the Rui’an eyeglass industry, and there is agood chance that Wenling is the same. Just as I had predicted from the outset, theemergence of the institutional moment was once again confounded by the problem ofoligarchy. Trade unions in Rui’an are, similar to their counterparts in other parts of thecountry, either unwilling or unable to involve membership in contractual negotiations, orany other features of political life that impact employment relations. Workers remainunaware of the actions that are taking place in their name, and so they have no allegianceto (or cognizance of) contracts to which they theoretically are party. Unions failed toincorporate workers, which then undermined their attempts at decommodification. As wehave seen previously, labor is strong enough to win formal concessions, but not strongenough to enforce those concessions. While the focus of this study is trade unions, acrisis of representation within the employers association only added to the difficulties.

Why then, if the outcome with regards to the institutional moment is essentiallythe same as in Guangdong, is this research of note? First of all, one of my overarchingconcerns relates to the relationship between the insurgent and institutional moments ofthe countermovement, or in general terms, assessing the outcomes of disorganizedworker movements. Even if labor remains similarly commodified and unincorporated inboth regions, the institutional response of the union is quite different in Zhejiang than it isin Guangdong. Different models of development have generated distinct possibilities forlabor politics, the result of which is that unions in Zhejiang have the capacity to negotiatecollective wage agreements at the sectoral level, something that has by and large nothappened in Guangdong. In Zhejiang we can see, in embryonic form, the emergence ofthe institutional moment of the countermovement. The union was pushed by the state(and a certain segment of capital) to do something about instability in employmentrelations, the outcome of which was the conclusion of sectoral wage agreements. On astrictly formal level, labor was decommodified in these cases, and in this sense we have astrong divergence from what has happened in Guangdong. It is of course significant thaton a substantive level the outcomes are the same, a fact that speaks volumes about thestructural deficiencies of trade unions in China and reveals the importance ofincorporation. And yet, in both Rui’an and Wenling we see that even a highly oligarchicunion that is almost completely severed from the lives of its members is groping aboutfor an institutional response to the chaos engendered by a free labor market. Additionally,the structures and modes of negotiation that were established by the sectoral agreementsin Zhejiang hold the potential for increased decommodification and rationalization, whileGuangdong unions are mired in trying to resolve labor conflicts after the fact. ThatZhejiangese unions developed the form of an institutional response is incrediblynoteworthy.

The final reason that an analysis of trade union activity in Southeast Zhejiang issignificant is because of the high esteem afforded to them by higher levels of the tradeunion and the government. The attention showered on Wenling in particular is anindication that the state has high hopes for oligarchic decommodification-style laborpolitics. While their concerns and responses vary, different levels of the state are all

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (131)


concerned about worker unrest, and are currently searching for various methods ofdealing with the problem. However, the key precondition is that workers themselves donot attain any organized and autonomous power. The consequence is that the state (andthe union) insist that there is always a legislative response to social problems. Theenactment of the Labor Contract Law in 2008 as well as the sectoral agreements inZhejiang should be seen as just such responses. However, the crisis of representation thatderailed the implementation of the wage agreement in Rui’an should come as a warningshot to the state: the state’s goals of keeping the working class deeply cellularized whilerestoring order to employment relations may end up in contradiction with each other.

Chapter 6

Worker Insurgency and the Evolving Political Economy of The Pearl River Delta

In previous chapters, we have seen how the union’s responses to workerresistance have encountered severe challenges in realizing decommodification or the re-alignment in power relations at the point of production necessary to politicallyincorporate migrant workers. Each case has revealed the state and union’s deep-seatedfear of social instability – a fear so profound that they would rather have laws gounenforced than devolve organizational capacity to workers. Even when higher levels ofthe state are interested in promoting decommodification in the service of rationalizingproduction, these efforts have been confounded by the non-enforcement of laws andcontracts within the enterprise. This has been made evident in cases where sympatheticelites first provide a degree of support to workers but then backtrack due to a fear ofworkers beginning to formulate political demands. In Zhejiang, we have seen instances

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (132)


where intermediate bodies of the union exert great effort in establishing the form, but notthe substance, of top-down, rationalized, and decommodifying collective contracts.However, the strong alliance between state and capital at the local level has left mangerswith incredibly high levels of autonomy, the result of which has been continual non-enforcement of laws and contracts, poor working conditions, and high levels ofcommodification and resistance among migrant workers. Paradoxically, the one way outof this conundrum – the emergence of an organized countervailing force at the point ofproduction – is categorically opposed by all levels of the state and union. This paradox iswhat I have termed “insurgency trap.” But by the spring of 2010 it became clear that not all incidents of workerinsurgency are treated equally. In this chapter I will compare two strikes which in manyways are similar – the little known Otis elevator strike and the famous Nanhai Hondastrike – but which produced very different outcomes. In the Otis case we see insurgentworkers resisting commodification, encountering inaction and passive repression fromthe union, and ultimately failing to win any victories. In this case, we see that unionoligarchy has caused the countermovement to stall at the insurgent moment. In the Hondacase, on the other hand, the strikers were able to “boomerang” (Keck and Sikkink 1998)around repressive local authorities to win the support of the provincial and central state,which then opened up the space for better organization and greater militancy. As a strikewave engulfed the entire foreign-owned auto industry in China, the capacity for the localstate and union organizations to maintain low-cost and highly repressed labor – alreadychallenged by the myriad daily labor conflicts around the country – began to crumble.Significantly, the position of the higher levels of the state had evolved between late 2007and 2010, affecting both the process and outcome of the strikes. The Honda strikedissolved the ideological image of a unified state edifice, as the central and provincialgovernment showed relative tolerance towards the workers while the local state resortedto coercion in a failed attempt to bring the deadlock to a close. The 2010 strike wavemarked an important turning point in the history of the Chinese labor movement, and inthe trajectory of the nation’s political economy more broadly.

My main argument in this chapter is as follows. The transition from the insurgentto institutional moment of countermovements can – particularly in non-democratic states– only be accomplished with the accumulation of countless, often relatively anonymousinstances of class struggle. As we can see in China, it is not until the unstable equilibriumhas been severely upended that the capitalist state, or at least significant segments of it,will be ready for compromise. For years, a strong alliance between capital and the lowestlevels of the Chinese state resulted in strikes being dealt with either through policerepression or through an ad-hoc system of mediation by union and government officialswhich was focused almost exclusively on resuming production, regardless of the outcomefor workers. But by 2010, the Chinese central government and Guangdong provincialauthorities were not only ready to seek a new model of accumulation in the Pearl RiverDelta, but were willing to (indirectly) ally with insurgent workers in attempting to realizethis goal. Just such an alliance, conditional and ephemeral as it may have been, emergedin the course of the Nanhai Honda strike, which in turn allowed the strikers to wineconomic demands and to begin to develop political goals. In large part because of thissmall political opening, the character of protest in the 2010 strike wave displayed someunusual (if not unprecedented) tendencies, most significantly that demands were

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (133)


offensive rather than defensive in nature. By comparing the strikes at Otis and NanhaiHonda, we can see that the position of the provincial and central authorities shifted in theintervening two and a half years, even if the lowest level of the state remained committedto low-cost labor. However, even if Honda employees won economic gains that eludedthe Otis strikers, we see that all levels of the state and union remain vigilant about thedevelopment of autonomous bases of worker power. Although economic gains weremade in the 2010 strike wave, worker disillusionment with unions from the enterpriselevel on up persists, and the state’s exit from insurgency trap remains murky.

The Countermovement in China

To briefly review, I originally posited that countermovements as originallyconceived of by Polanyi are too mechanistic, and that he fails to account for the complexrelationship between anti-commodification social struggles and the consequences forstate action. Subsequent theorists have tended to conceive of the double movementprimarily either as resistance pure and simple (Chin and Mittelman 1997), or asrepresented by state policy. By pointing to expanded social spending in China since thebeginning of the 21st century, Wang Shaoguang (2008) falls into the latter category. Withlittle discussion as to why such spending has increased, how/if it has been distributed, orwhat the implications are for decommodification, Wang has implied that the doublemovement has already been realized in China. But in this work we see a conflation ofsocial rejection of commodification and actual decommodification. Because in Chinathere is hardly any space in which workers can develop autonomous power to counter thehegemony of market and state (Burawoy 2003), worker rejection of commodification isconfined to cellular activism (Lee 2007) which has yet to cohere into the type of politicalforce that can articulate specific political demands.172 In Wang’s view, we are led tobelieve that simply because there were some negative consequences of marketization, thestate decided to initiate the double movement. Society, or in this case cellular resistance,and the deeply entrenched interests of state and capital are eliminated from the formula,i.e. the political aspect of the countermovement is obliterated in the analysis. Thus we areleft with no explanation for why the state (and which levels of the state) might or mightnot initiate reform, aside from the fact that it is necessary.

In the existing literature, it is clear that the new Chinese working class has beenengaged in frequent acts of resistance against the incursion of the market (Chan 2010;Chan and Pun 2009; Lee 2007; Pun, Chan, and Chan 2009; Pun and Lu 2010). I haveshown that higher levels of the state are sometimes supportive of anti-commodificationstruggles, particularly when they threaten rationalized accumulation, and that they arekeen to integrate workers into regularized and legal modes of contention. And yet, evenwhen pro-worker legislation is adopted at the national or regional level, actualimplementation is frequently confounded by vaguely worded laws (Cooney 2006) and the

172 My perspective is quite different from that of Teresa Wright who contends thatvarious social groups in China “accept” authoritarianism (2010). Although Wrightacknowledges that this acceptance is perhaps most tenuous among migrant workers, Iwould counter that the lack of a well-articulated political opposition is due not toacceptance of the status quo but rather results from the continual threat of repression.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (134)


alliance between the local state and capital. Commodification on a national scaleproduces new interests and new power arrangements in society. When certain segmentsof society (and possibly the state) revolt against further commodification, they muststruggle to realize these goals; decommodification severely impacts the interests ofcapital and various levels of the state, and so is not easily won. As a result, it is ofincredible importance to analyze specific cases of worker insurgency and the response ofstate and union to see the dialectic between resistance and institutionalization in action.Additionally, while national level policies on increasing social spending are worth noting(particularly because they indicate a segment of the state is interested in classcompromise), they cannot be accepted at face value as large swaths of migrant workershave been excluded from the potential benefits (Frazier 2010).

Owing to this, I have argued that in the early stages of capitalist development,countermovements against the market must be broken down into two constituentmoments: the “insurgent moment” when workers and others negatively affected bymarketization rebel and engage in non-regularized forms of contention; and the“institutional moment” when decommodification of labor comes to be institutionalized atthe social level and the working class is politically incorporated by the state. The centralargument up until now has been that persistent oligarchy within the ACFTU – workers’only legal representative – has caused the countermovement in China to stall at theinsurgent moment (i.e. fall into an insurgency trap). Until we descend to the shopfloor tosee how and why workers organize strikes, and what the concrete responses of unionofficials are when such an event happens, we cannot fully understand the relationshipbetween the insurgent and the institutional moments of the countermovement.

Particularly in a country as vast in size as China, it is important to understand howinsurgent workers interact with various levels of the Chinese state. Kevin O’Brien andLianjiang Li developed the influential concept of “rightful resistance” (1996; 2006) todescribe how protestors use the language and concepts of the state in attempting tocircumvent local officials in attracting attention from higher ups. Ching Kwan Lee hasdiscovered the political economy undergirding similar forms of resistance amongworkers, in what she terms “decentralized legal authoritarianism.” This allows us to seethat the ideological image of a benevolent central authority counter-posed to corrupt andrapacious local officials is produced through a specific economic and politicalarrangement – one which not coincidentally allows for a more stable form of domination.And Yongshun Cai has argued that one of the key determinants in protestors getting theirgrievances addressed is whether or not they cause enough of a disturbance so that higherlevel officials (especially the central government) cannot maintain the illusion ofignorance (2010). I concur with these scholars that the lowest levels of the state oftenappear to be, and in fact are, aligned with powerful interests against workers, and thathigher levels of the state (especially the central government, but sometimes the provincialor even municipal authorities) are more likely to be sympathetic. But as we will see inthis chapter, there is an ongoing dialogue between the central authorities and insurgentworkers. During the first thirty years of China’s transition to capitalism, increasingiterative flashes of worker insurgency have not merely dissipated into the ether. Rather,in a complex and highly mediated process, residual particles of insurgency have settledon the political field, eventually re-making the landscape and changing the attitudes ofthe state towards wildcat strikes. Gaining the attention of the central government through

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (135)


social/economic disruption is helpful not just in resolving a particular case (as argued byCai), but has ongoing consequences for contours of resistance itself. This dynamic will beclearly illustrated in the following empirical account.

The Otis and Honda cases make for a good comparison. Both are capital-intensivewholly foreign-owned operations, with Otis located in Guangzhou and Nanhai Hondaless than 25 miles away in the adjoining municipality of Foshan. Production in bothfactories requires skilled labor, and most workers are graduates of technical schools.Workers enjoy somewhat better working conditions than in most labor-intensiveindustries in the Pearl River Delta and other export-oriented regions. And Honda and Otisalike employ the labor force dualism so prevalent in capital-intensive industries in China(Zhang 2008), in which a large segment of employees are hired with fewer jobprotections, lower wages, and worse benefits then the regular workers (Yu 2009). Andyet, we will see that different organizational capacity and labor market dynamics, as wellas different overall political conditions forced the union to adopt varying responses to therespective strikes. Additionally, we will see that various levels of the union had verydifferent responses to each strike.

Commodification, resistance… and more commodification

I first heard about the Otis Elevator factory from Guangzhou Federation of TradeUnions (GZFTU) chair Chen Weiguang during an official visit he made to the U.S. As isthe case with most trade union leaders from China, Chen is aware that many foreign laboractivists take a rather dim view of the ACFTU and its subordinate unions. Thus, Chenwas eager to demonstrate the efficaciousness, and perhaps potential advantages, of theChinese model of trade unionism. In pursuit of such an end, he brought up the case of theOtis Elevator factory several times on his trip in the U.S. in the spring of 2008. Nearly ayear and a half later, in October 2009, another GZFTU delegation came to the U.S., thistime headed by vice-chair Zheng Yiyao. Zheng repeated the story of Otis Elevator to agroup of labor scholars and activists at UCLA,173 indicating that it was still considered aprime example of union efficacy.

The story as Chen and Zheng related it went like this: Otis Elevator is anAmerican owned company, which has a production site in the Baiyun district ofGuangzhou. In the late fall of 2007 the company announced to workers that they wouldbe switching from an hourly-rate system to a piece-rate system in order to increaseproductivity. They announced how much workers were going to be paid per-piece anddemanded that all employees sign a letter of intent that indicated their acceptance of thischange in factory rules. Workers were quite upset with this because, according to theircalculations, their salaries would be dramatically reduced by such a change in wagesystem. They tried to get assistance from their enterprise-level union chair, but he wasultimately unresponsive.174 Feeling that they had no choice, the workers went on strike todemand that the hourly-rate system remain in place. After the strike occurred, theGZFTU got involved to mediate. Union leadership met with workers and with managers,

173 Field Notes, October 2009174 The fact that Chen was willing to publicly admit that the enterprise level chair didnothing is noteworthy and unusual for Chinese union leaders.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (136)


and eventually reached a resolution. Although the piece-rate system was to beimplemented, the union successfully negotiated an increase in the piece-rates. Severalmonths after the strike, productivity was up, workers’ wages were up, and everyone wassatisfied. It was a “win-win” outcome, and it revealed that Chinese unions were capableof defending workers’ interests without being antagonistic towards management.

Chen told this story because he thought it was a good demonstration of the factthat Chinese unions are quite capable of promoting worker interests. In reality, not onlywas the union completely incapable of realizing any decommodification, their negotiatedresolution was unable to prevent further outbreaks of worker insurgency. The actualevents of this case reveal the immense challenge Chinese unions will face if they are tosignificantly decommodify labor without challenging the interests of state or capital.

First of all, it is not at all insignificant that the Otis Elevator story was widelycovered by the Guangzhou press. The GZFTU did not get involved until after the storybegan gaining a lot of public attention. An unfortunate but common phenomenon is thatthe trade union (and other government agencies) refuse to address mounting socialconflicts until they pose a real threat to their image. Time and time again, the union hopesthat conflicts can be glossed over and that they will just go away. Workers understandthis logic of the state as is reflected in the clichéd phrase, “create a big disturbance, get abig resolution, create a small disturbance, get a small resolution, create no disturbance,get no resolution.” Otis Elevator employees had tried to resolve their grievance byapproaching their enterprise level trade union chair, and some workers reported trying toget assistance from the labor bureau. But it was not until they went on strike and receivedwidespread coverage in the media that the municipal level union decided to intervene.This consistently reactive stance of the union is in line with general practice.

While Chen’s account was more or less accurate in describing the events up to thefinal resolution, it is still important to fill in some additional details. Tensions in thefactory began building up in mid-December when the company fist announced the switchto a piece-rate system. Workers were quite angry about the announced switch, sinceaccording to their calculations, their salary would be decreased by 60%, and they wouldbe forced to do more overtime. On December 18th, 128 of the workers went on strike,completely shutting down operations at the entire production facility. Two days later, thecompany sent a “gentle reminder” saying that anyone who refused to sign a letter ofintent accepting the piece-rate system by the following day (December 21st) would havetheir contract terminated at the end of the month.

Unbeknownst to the workers, enterprise level trade union chair Cheng Weiji hadalready discussed the new piece-rate system with management and claimed to haveexpressed the interests of the workers. In order to further grasp the farcical nature ofCheng’s discussion with management, it is worth quoting at length a section of anewspaper article which appeared on December 21st:

Yesterday around 1p.m. this reporter was finally able to see the chairman of OtisElevator Company, Mr. Cheng Weiji. It was surprising to find that every one of hisstatements came under fierce attack from the 128 workers at the scene of the incident.

Cheng Weiji said, “Today we convened a staff and workers congress. Workers werepresent.”

“This didn’t happen,” said the workers.Cheng Weiji said, “The union contacted the representatives that you all selected to

discuss the piece-rate issue.”

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (137)


“No, who are you saying participated?” the workers said.Cheng Weiji said, “The night before last, the union held a meeting to discuss this

issue. We informed all of the workers that there is going to be a negotiation conferencetoday at 1:30.”

“We didn’t receive any notification,” the workers said.Cheng Weiji said, “Just because you didn’t receive the notification does not mean that

it wasn’t sent. We sent it out by email.”“We don’t even have computers, how can we receive something by email?” responded

the workers.In an interview with this reporter, Cheng Weiji said that the union had discussed the

matter of the piece-rate system with Otis Elevator Company two times, and that theyconvened a union representative congress. The union representatives transmitted the spiritof the congress to the employees. This reporter inquired what the union was prepared to doif the company terminated the contract of employees, some of which have been at thecompany for 10 or more years, if they refused to sign an agreement on accepting the piecerate system. Cheng Weiji said that he is not an administrator, so it wouldn’t be appropriatefor him to give a response.175

Clearly, the enterprise level union was not up to the task of resolving the impasse,as their representation was formalistic at best. At this point the GZFTU and the Electricaland Machine Industrial Union dispatched officers to Otis Elevator Company to do aninvestigation. As a symbolic indication of the gravity of the event, the ACFTU sentsomeone from Beijing to attend the investigation. The goal of the investigation was todetermine if the company had gone through the appropriate legal procedures forimplementing the new system, to determine if workers’ legal rights were being protected,and to decide if the enterprise level union chair had fulfilled his appropriate duties. At thesame time, the union began its own evaluation of the proposed changes, with the intent ofcoming up with an alternative piece-rate system. Chairman Chen, legal experts, and othertrade union leaders planned to personally visit the company the following week to engagein negotiations.

The labor bureau also jumped into the fray. Though initially giving cautiouswarning to the company that their approach could violate labor laws, they quickly did anabout face. Chief of the Baiyun District Labor and Social Security Bureau Xie Xijian wasquoted in the newspaper as saying, “the enterprise has sovereignty over the determinationof labor contracts. As long as it is above the minimum wages for the province and thecity, the labor bureau cannot intervene.”176 Since the bureau understood that a reductionof wages to levels below that of the legally mandated minimum was not under discussion,this comment was equivalent to removing the government from adjudication of the issue.Management’s confidence in the support of the government was evident when they saidthat if they were unable to come to a resolution, they would “encourage” workers to filefor arbitration. The pro-capital position of the state was not lost on workers, as isevidenced by this comment from a temp worker about the strategy of workers in a secondstrike that took place in August 2008: “some people said we should go report it to the

175 December 21, 2007. “Otis guangzhou company plans to take away 60% of workerssalary; those who refuse to sign will be fired.” Xin kuai bao.176 December 22, 2007 “GZFTU to investigate Otis’s contract to protect the rights andinterests of workers.” Guangzhou ribao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (138)


labor bureau, but others opposed this. Those in opposition said that when there was aconflict last time, some people went to the labor bureau, but it didn’t have any effect. Thelabor department doesn’t help workers, it protects bosses. So if you go it will beuseless!177”

After garnering much attention in the print media and on various blogs (where theoverwhelming majority of people supported the striking workers), a gag-order wasimplemented and the story disappeared from public discourse. It is of course significantthat this gag-order was issued before the resolution of the conflict, since the likelihood ofthe union or other government agencies having their image tarnished was greatlyreduced. However, the Otis Elevator incident did appear in the papers one more time atthe end of the following January. Chen Weiguang, displaying some of the characteristicswhich have given him a good reputation among foreign trade unionists, publiclycriticized the inaction of the enterprise level union in Otis Elevator: “On the surface, theOtis incident is a conflict between labor and capital, but in the background it magnifiesthe crisis of our enterprise level trade unions, something which is worth uscontemplating.”178 Chen went on to make a more general critique of these enterprise levelunion chairs, saying that, “some enterprise level union chairs are taking the wrongposition and are completely servile towards capital. They have completely forgotten theirprimary responsibility as a worker representative.”179 The fact that Chen’s leadership andinsistence that unions more effectively represent workers was unable to affect managerialauthoritarianism reveals some of the fundamental weakness of the trade union.

So what was the actual outcome of the Otis Elevator incident? In interviews withworkers conducted more than a year after the strike, there was universal displeasure withthe results. Most of these workers participated in the strike, but said that in the end theyhad no choice but to sign the letter of intent. Otherwise, they said they would lose theirjobs. Contrary to the promises of management and the reassurances of the union, wagesfor workers declined significantly after the implementation of the piece-rate system. Oneworker claimed that 90% of workers’ salaries had declined since the strike. He said thatbefore the strike “temp workers” such as himself earned between 2000 and 2500 RMB,but that their wages had since been reduced to 1300-1600. Another temp worker said thathe made only 1100 RMB the previous month. For regular workers, their original salarieshad been between 4000 and 4500, but had since been reduced to 3000-3500. While Icould not verify these numbers with official documents, all interviewees confirmed thattheir wages had been significantly reduced.

This leads to a secondary point which was not addressed in any of the mediacoverage: Otis Elevator Company’s employment of temporary workers. Theimplementation of labor force dualism has become increasingly common in manyenterprises throughout China,180 both as a method to break employee solidarity and in

177 Interview, May 2009178 January 1, 2008. “Some union chairs are taking the wrong position.” Guangzhou

ribao.179 Ibid.180 Zhang, Lu. 2008. "Lean Production and Labor Controls in the Chinese Automobile

Industry in An Age of Globalization." International Labor and Working-Class

History 73:1-21.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (139)


order to maintain greater flexibility in human resource management. This was the case inthe Otis factory, with a very significant number of employees hired as “temp workers”through a hiring agency. The long-term maintenance of this sort of labor regime is inviolation of the legal requirement of “equal pay for equal work.” Indeed, intervieweesfrom Otis Elevator expressed that as temp workers they did exactly the same work asregular workers but were compensated at a much lower level. This system was notaddressed at all in the negotiations at the end of 2007. The failure to address this issue issignificant, because temp workers went on strike briefly in August of 2008, protestinglow wages and unequal treatment. I was unable to get more detailed data on this event,but two workers mentioned it in interviews.

Finally, there is little evidence that the union did anything to secure increasedrecognition from workers. Many interviewees, despite participating in the strike, wereunaware that the factory had a union:

Interviewer: Before the strike, what actions did the Otis company union chair take?Respondent: I didn’t know our company had a union!Interviewer: Well then, how about the higher level trade union?Respondent: Higher level trade union? Is that the labor bureau?181

In response to the same question about the behavior of the union chair, a differentrespondent said, “my friend said that the union chair and the boss ‘wear the same pants,’he just speaks for the boss!”182 This respondent said that the union chair eventually toldthem that management’s actions were legal and that they would either have to obey thenew rules or leave the company.

But perhaps the most succinct summary of the outcome of the Otis Elevator casecame from a temp worker who was asked whether there was a “win-win” resolution:“Where is there win-win? It was an absolute and thorough defeat for the workers. Anabsolute defeat!”183

The Honda Conflagration

On May 17, 2010, more than two years after the Otis Elevator strike, hundreds ofworkers at Nanhai Honda184 walked off the job to demand higher wages. Over the nextthree weeks, all of Honda’s production facilities in China would be shut down, and astrike wave would have cascaded throughout the auto industry, plunging the union intoperhaps its deepest political crisis of the reform era. Although the original organizers inNanhai did not have the intention of sparking a string of worker revolts, that is preciselywhat they accomplished. As we will see in a moment, the changing dynamics of workerinsurgency produced a response from the union that simultaneously revealed continuityand evolution; in short, a dialectical progression. The labor movement in China would bechanged irrevocably.

181 Interview, May 2009182 Interview, May 2009183 Interview, May 2009184 The official name of this company is China Honda Auto Parts Manufacturing Co.,Ltd. However, the plant has frequently been referred to as Nanhai Honda in the media,and so I will use this terminology.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (140)


In the Honda case we will see both similarities and differences with Otis. Similarin that the enterprise unions were completely unable to either effectively represent theirmembership or even to stave off insurgency. Although on the surface the initial reasonsfor going on strike were different, underlying both incidents was the general process ofcommodification of labor. But the outcomes of the two strikes were quite different.Although Honda workers did not win everything they asked for, it would be impossibleto describe the outcome as an “absolute defeat.” The organizational capacity of workerswas much greater at Honda, and they made specifically political demands. Both thesechanges in the dynamics of resistance as well as changes in the overall political economyresulted in a different response from both the union and the state than what we have seenpreviously.

Nanhai Honda

Honda’s production chain in China consists of a somewhat convoluted system ofownership. The most significant company is Guangzhou Honda, a 50-50 joint venturewith the state-owned Guangzhou Automobile Group Corporation, where a majority ofunits are produced. Additional assembly plants include Honda Automobile (China),which produces for foreign markets, and the joint-venture Dongfeng Honda located inWuhan. These assembly plants are served by a variety of parts manufacturers, includingthe wholly Japanese-owned Nanhai Honda. Starting production in March of 2007 with aninitial investment of USD $98 million,185 the company was Honda’s fourth integratedautomatic transmission production plant in the world.186 Aside from producingtransmissions, the plant also makes drive shafts and connecting rods for engines.187 Inpart because Honda believed that work stoppages were highly unlikely in authoritarianChina, the Nanhai plant was established as the sole supplier of several key parts for theentire China operation. As stated in the press release accompanying the company’sestablishment, “The start-up of production at [Nanhai] enables Honda to secure anadequate supply of powertrain components to support expansion of Honda’s automobileproduction in China, and also to further increase local content of powertrain components,which will help cost reduction efforts and strengthen Honda’s competitiveness in themarket.” By sourcing from within China rather than from Japan or Southeast Asia, costswould be reduced by saving on transportation and labor.

In part because of the key position that auto manufacturing plays in the economy,the government put a high premium on maintaining good labor relations in this sector. Asa result, all of the Honda assembly and parts manufacturing plants in Guangdong hadunions established. The union at Guangzhou Honda had been awarded several officialaccolades for its good work, and frequently hosted visiting delegations of foreign tradeunionists. But there were strict limits on how much even this model union would do for

185 June 11, 2010. “dongfeng bentian: wufa duoshan de hudie xiaoying.” [DongfengHonda: No way to hide from the butterfly effect] E’shang zazhi.186 March 8, 2007. “Honda Auto Parts Manufacturing Co., Ltd. Begins Operations inChina.” [accessedJuly 3, 2010]187 ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (141)


its workers. During a lunch meeting in December 2008 between the chair of GuangzhouHonda and visiting union leaders from the U.S., talk turned to international cooperationbetween auto unions. The union chair said that they had visited Japan previously to holdexchanges with auto union representatives, and that he felt they had much in common.Alluding to the difficulties American auto manufactures were facing at the time, he jokedthat he had told his Japanese counterpart, “we have a strong union, like you. But we don’twant to be too strong, just look at all the problems they have in the U.S.!”188 In fact, itturned out that the very weakness of the union at the Nanhai supplier plant would make itimpossible for workers to have their demands heard without going on strike; not just theNanhai plant, but Honda’s entire China operations would be shut down as a result.

Although workers at Nanhai had long been unhappy with the wages and haddiscussed going on strike, hardly anybody knew that Tan Guocheng was going to initiatethe strike when he did.189 One week before the strike Tan met with 15 people from theassembly department were he worked, and previously they, “had random talks on theshuttle bus to work.”190 One worker from this department said that the idea had beendiscussed but that nobody wanted to lead it. In separate interviews, workers from otherdepartments confirmed that they had heard nothing of the strike until it had begun. Butaccording to Tan, more than 20 people, most of them from Hunan, had been in on theplan by the time it was put into action.191

On the morning of May 17th, just as production normally began at 7:50a.m., Tanhit the emergency stop button and both production lines in the assembly department wereshut down. Tan and co-organizer Xiao shouted out at each assembly line, “Our wages areso low, let’s stop working!”192 For most of the plant’s nearly 2000 workers, this was to bethe first they heard of the strike. Even one worker who was from the assemblydepartment and had heard discussion about the possibility of the strike was caughtunaware: “I didn’t know the strike was going to happen… I wasn’t there at the time[because I went to the bathroom]. When I was finished in the bathroom I came out andthere weren’t any people. I stood there looking, ‘huh, how come they aren’t at work?’”193

As workers from the assembly department fanned out throughout the facility, theyshouted to their co-workers to stop work and join them in fighting for higher wages. Theyinitially received a somewhat cool reception in the other departments, and eventuallybegan a sit-in in front of the factory with only about 50 workers. But given the criticalposition of the assembly department in the production process, the other departmentswere forced to shut down in a matter of hours. By that afternoon, management had set upsuggestion boxes and pleaded with the workers to resume production, promising them

188 Field Notes, December 2008.189 This is a pseudonym. The same person also appeared as “Tan Zhiqing” in otherreports.190 Barboza, David. June 13, 2020. “In China, Unlikely Labor Leader Just Wanted aMiddle-Class Life.” The New York Times.191 June 2, 2010. “zhongguo bentian nanhai chang jin fugong, cong yuangong jiaodu kan

laozi shijian.” [China Honda Nanhai factory resumes production today, looking at laborconflicts from the perspective of employees]. Zhongguo Xinwen Zhoukan.192 Ibid.193 Interview July, 2010.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (142)


that they would consider their demand for higher wages and provide a full response infour days. Perhaps because of their relatively small numbers, the strikers tookmanagement at their word, and production was resumed that very day.

On the 20th, management, government officials, union officials, and workerrepresentatives engaged in negotiations. The workers demand at this point was simply toraise all wages by 800 RMB. In the meantime, the strikers returned to work, thoughproduction was greatly reduced during these few days. On the 21st, negotiations brokedown and the strike continued. Over the weekend, organizers continued their outreachand the number of strikers in front of the factory grew to over 300. Then, on the 22nd

management announced that Tan and Xiao, the two original strike leaders, were havingtheir contracts terminated. But this attempt at repression completely backfired, as thefollowing day the strike only grew in strength. Now concerned for their livelihoods,workers covered their faces with surgical masks, but continued to hold the line.

Throughout this process, the enterprise union alternated between passivity andhostility. Workers complained that during the bargaining session the union representativedid not say anything at all, but merely observed the proceedings. When the strike initiallybegan on May 17th, a team of investigators from the district labor department and tradeunion were dispatched to the factory. Leaving no doubt which side of the struggle theywere on, the officials announced, “according to relevant regulations, we did not find thatthe factory is in violation of any laws.”194 One worker who was selected as arepresentative was quite disappointed with the behavior of the enterprise union chair, WuYouhe, in the first round of negotiations:

[The enterprise union chair] invited a lawyer [to the first round of negotiations].The lawyer said that our strike was illegal. He [the union chair] didn’t have any views ofhis own, and couldn’t make any decisions. He always asked the general manager what todo. At bottom he is a chairman, and isn’t controlled by the company, he has this power.But for him, everything had to go through the general manager, and he would help thegeneral manager refute the things we said.”195

On the 24th, worker representatives were convinced to come back to the table in anegotiation session chaired by the enterprise union head. Still trying to serve as anintermediary, the union chair attempted to persuade the workers to accept management’soffer of a RMB 55 increase in food subsidies – a far cry from the RMB 800 they weredemanding. This ineffectiveness was not lost on the workers, with one strikercommenting, “The union said it stood for our interests. They said us employees couldgive them any demands and they would pass them on to management, and they wouldresolve things for us. But they didn’t do this in the slightest.”196

The strikers refused management’s offer on the 24th, and the situation escalated.On the 25th things became much more tense when all of Honda’s assembly plants inChina were completely shut down due to lack of parts. Originally counting on a well-disciplined workforce, Honda only had one supplier for transmissions in the country, and

194 May 27, 2010. “bentian duzi lingbujian chang bagong zhi guangben tinggong.” [strikein Honda-owned factory results in stoppage for Guangzhou Honda] Caixin.195 Interview, July 18, 2010.196 Interview, July 2010.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (143)


all four assembly plants in China were therefore highly dependent on Nanhai. Thecombined daily loses of the five plants was estimated to be RMB 240 million.197

Management further yielded by producing a second offer for wage increases on the 26th.This proposal called for increasing regular workers’ salaries by RMB 200/month, alongwith RMB 155 in living expense subsidies, and a wage increase of RMB 477 for internswho had been at the plant for more than three months.198 But workers rejected this offeras well, and the strike continued. At this point, workers formalized their demands, and inaddition to the primary demand of increasing wages for all employees by RMB 800, theyalso demanded that the fired workers be re-hired, that there would be no retributionagainst strikers, and that the enterprise union be “re-organized.” (chongzheng) Accordingto some strikers, the demand for union re-organization emerged after seeing that theunion had failed to actively represent them in the previous negotiations sessions.

With the losses mounting, management became desperate and did their best to tryto break the resolve and unity of the strikers. The most direct attack was on the 28th whenthey attempted to force workers to sign a pledge saying that they would, “not lead,organize, or participate in slowdowns, work stoppages, or strikes anymore.”199 But thistactic completely backfired as almost nobody agreed to sign it, with one worker saying,“as soon as I saw it [the agreement] I threw it away. We won’t sign.”200 One group offemale workers said that, “nobody moved a hand.”201 When asked if they were afraid ofrefusing management’s demand, one worker insisted, “Nobody was afraid! Who wouldbe afraid? If they want to fire us, then they’ll have to fire all of us!”202

The strike was entering a decisive stage. Likely already the longest strike everwaged by migrant workers in the reform era, the situation had become a political crisisfor the local state. Despite the mounting economic and political costs, the events of May31st took everyone by surprise.

The Union as Strikebreaker

When workers arrived at the factory on the morning of May 31st, they wereinformed that each department would be holding meetings to further discuss strikeresolution.203 As the workers were waiting in various rooms of the main administrationbuilding, a large contingent of vans and buses pulled up in front. The vehicles were filledwith dozens of men, all of whom were wearing yellow hats and badges reading “Shishan

197 June 2, 2010. “zhongguo bentian nanhai chang jin fugong, cong yuangong jiaodu kan

laozi shijian.” [China Honda Nanhai factory resumes production today, looking at laborconflicts from the perspective of employees]. Zhongguo Xinwen Zhoukan.198 May 31, 2010. “bentian jiaxin shi bagong chongji di chengben zhizaoye moshi.”[Honda wage strikes are a shock to the low-cost manufacturing model] Caixin.199 May 30, 2010. “nanhai bentian tichu di’san ge tixin fang’an.” [Nanhai Hondaannounces third proposal to raise wages] Caixin.200 Ibid.201 Interview, July 2010202 Interview, July 2010203 In addition to interviews, much of the information from this section is derived from adetailed account written by a Honda worker which was posted on-line.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (144)


Township Federation of Trade Unions,” which is the union organization immediatelysuperordinate to the enterprise union branch. Shortly thereafter, the assembly department,crucial to reviving production, met with the general manager of the plant, during whichtime he made a new offer for a wage increase. Although still dissatisfied withmanagement’s new offer, they were persuaded to return to their assembly lines.Indications began to emerge that the strikers’ unity was crumbling as some departmentsbegan to start up their assembly lines. People from the union dispersed to each of thedepartments and encouraged workers to immediately resume production.

When some workers from the assembly department moved to return to the area infront of the factory where they had been demonstrating over the previous nearly twoweeks, a confrontation with the union group emerged. As confirmed from multipleindependent sources, the union people began filming the workers and demanded that theyreturn to the factory and end the strike. A tense situation quickly escalated and soondevolved into a physical confrontation during which several workers were struck bypeople from the union. This infuriated the workers, and a strike which appeared to belosing steam was quickly re-invigorated. Workers from other departments who hadresumed production rushed to the scene as soon as they received news of the violence,and a large crowd quickly gathered. Another physical confrontation occurred, and thistime the union side was even more violent than before, with several workers sufferinglight wounds. Among those that were hit were some women workers. The aggressorsquickly retreated to their vehicles and refused to come out.

At this point, the government decided things had gone too far, and took steps tosettle the conflict. Riot police were deployed, though they never engaged the workers.The authorities additionally cordoned off the road into the factory and nobody wasallowed entrance. Whichever government agencies had supported the peaceful strikewere not interested in more violent confrontations or the possibility of the strikers leavingthe production grounds.

It is certain that most of the strikebreakers were not actually union officers. Thefirst thing mentioned by many workers is that it seemed preposterous that the townshiplevel federation, with only a few paid members on staff, could recruit so many officersfrom other union branches. One worker involved in the scuffle said that some of thestrikebreakers (all of whom were male) had earrings and tattoos, items which unionofficials would be very unlikely to sport. But if most of the thugs were not actually unionofficers, it is simultaneously undeniable that the district union federation had a hand inorganizing the strikebreakers, a point made obvious in a letter they wrote to workers (seebelow). A foreman from the assembly department was blunt in his assessment: “of courseit was the union’s idea. Who else would have such a stupid idea? Only Chinese unionswould think of this.”204 It is, however, unclear to what extent the union federation wasacting at the behest of management or whether they were taking independent action.

When workers received an open letter from the Shishan Township and NanhaiDistrict Federations of Trade Unions the following day, the local union leadershipprovided a tepid apology and did not denounce the violence that had occurred theprevious day, nor did they attempt to deny that they had organized the strikebreakers. Theletter is worth quoting from at some length:

204 Interview, September 29, 2010.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (145)


Yesterday the trade union participated in mediation talks between the workers andmanagement of Honda. Because a portion of Honda employees have refused to return towork, factory production has been severely curtailed. In the process of discussions withforty or so employees, at one point there occurred some misunderstandings and verbalimprudence from both sides. Due to the impulsive emotional state of some of theemployees, a physical conflict ensued between some employees and representatives fromthe union. This incident has left a negative impression on employees. A portion of theseemployees, after receiving word of the incident, seem to have misinterpreted the actions ofthe union as siding with management. Yesterday’s incident came entirely as a shock to us.If people feel that some of the methods used in yesterday’s incident were a bit difficult toaccept, we apologize.

* * *The behavior of the above mentioned group of forty or so workers has already damaged theinterests of the majority of employees. In addition, such behavior harms factoryproduction. The fact that the union has stood up and admonished these workers is entirelyin the interests of the majority of employees. This is the responsibility of the union!

* * *It would be unwise for workers to behave in ways that go against the interests ofthemselves and others because of impulsive emotions. Some employees are worried thatrepresentatives who are willing to stand up and enter into talks with management wouldlater receive the reprisals of management. This is a misunderstanding.205

The letter went on to admonish workers for refusing to accept the offer that managementhad made. In a final attempt at damage control, the letter closed by saying, “Please trustthe union. Trust each level of Party officials and government. We will definitely upholdjustice.”206

Unsurprisingly, the letter from the Shishan and Nanhai union federations wasunsatisfactory to the strikers. As one worker activist put it, “Their apology letter wasn’tan apology letter at all, so we were pretty enraged.”207 An open letter from workerrepresentatives which appeared two days after the union’s apology letter was defiant:“The union should protect the collective rights and interests of workers and lead theworkers in the strike. But up until now, they have been looking for excuses for the unionpeople’s violence against striking workers, and we seriously condemn this.” 208

Additionally, the letter went on to express “extreme rage” at the union’s claim that it wastheir hard work which had caused management to increase their offer of wage increases,arguing rather that these were, “won by the blood and sweat of striking workers facing

205 “Open Letter from the Nanhai District Trade Union and Shishan Town General TradeUnion to the Workers of Honda Motors Nanhai Component and Parts Factory.” Ibid.207 Interview, July 18, 2010.208 The open letter from the worker representatives was posted online, available at:

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (146)


extreme pressure.”209 Relations between the strikers and the township level union couldnot have been worse, and certainly heightened the tension of the unfolding drama.


If the tactics of the township level union failed to break the deadlock, higherlevels of the union and Party were much more sympathetic to the strikers. I heard fromGZFTU leadership that Guangdong Party secretary Wang Yang supported the strike andthe workers’ wages demands, and even that there was support in the central government.The Central Propaganda Department did not issue a reporting ban until May 29th, nearlytwo weeks into the confrontation, at which point the strike wave had spread to otherfactories. 210 But this was an indication that the central government was willing to allowmore pressure to build on management, as it is rare for coverage of strikes to go on for solong. GDFTU vice-chair Kong Xianghong took an active role in the negotiations, andwas supportive of the wage demands. Particularly following the confrontation betweenthe Shishan union and workers, the provincial level authorities were eager to resolve theconflict quickly.

In order to find an orderly resolution, the various government agencies that hadbecome involved in the strike demanded that the workers select representatives. Althoughthere had been a hastily arranged set of negotiators selected in the first round of talks,strikers had become reluctant to produce representatives, particularly after the two peoplewho initiated the strike were fired. This unwillingness to negotiate was unacceptable tothe state, and they brought in Guangzhou Automotive CEO and National People’sCongress delegate Zeng Qinghong to speak with the workers. Through gentle andpaternalistic persuasion, Zeng convinced the strikers to select representatives and to begina conditional resumption of production late on June 1st. In their open letter, the workerrepresentatives said that if management had not met their demands within three days, thestrike would be resumed. Furthermore, the letter stated that, “bargaining representativeswill not accept anything less than the above listed demands without the authorization of ageneral meeting of employees.”211 Finally, negotiations began on the 3rd.

On June 4th, the worker representatives were joined by Chang Kai, a well-knownlabor scholar from Beijing, who served as their legal counsel. Negotiations went late intothe night, and eventually an agreement was struck. Regular workers were to receive wageincreases of approximately RMB 500, bringing their monthly wages above RMB 2000.The underpaid “interns” that work alongside regular workers saw their wages increase bymore than RMB 600, an increase of more than 70%. Such large wage increases inresponse to strikes were unprecedented in China.

Strike Wave

209 ibid.210 Thanks to Jonathan Hassid for advising on this; the leaked Department of Propagandadirective can be found here:

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (147)


Before discussing the political fallout from the Nanhai strike, it is important tonote that the ongoing coverage of the struggle had sparked widespread resistance amongindustrial workers around the country. The central government had given the ok formedia coverage to continue, as it fit in with their inclination towards raising wages forskilled workers. Additionally, the fact that Honda is a Japanese owned factory wasimportant, as it appeared possible to divert anger about worker mistreatment innationalistic, rather than class-based, directions.212 And yet, the central governmentcertainly got more than it was bargaining for as strikes ripped through the auto industry,even spilling over into other sectors. While many strikes targeted Japanese-ownedfactories, this was certainly not true in all cases. With workers experiencing generalizeddissatisfaction, a small relaxing of media controls was all it took to ignite an outbreak ofintensified insurgency. It is important to keep this uptick in class struggle in mind tounderstand the context in which both the Nanhai and subsequent strikes were resolved.

In each strike, the primary demand was for large wage increases. While theinsistence on union re-organization was not highlighted as much in these other struggles,it was included as a demand by many strikers. Perhaps most noteworthy among thecopycat strikes was that initiated on June 21st at Denso in the Nansha district ofGuangzhou. Employing 1100 workers, the Japanese-owned Denso is a major auto partssupplier, in particular for the Toyota assembly plant located in Guangdong. On severaloccasions, workers in the plant had brought up issues related to pay and living conditionswith the enterprise union, but nothing changed. Indeed, a shop-floor union representativeadmitted that the “small group” (xiaozu) meetings that are supposed to happen once amonth had only been held once ever, and that there had never been an employee congress(meeting for all enterprise employees).213 Just as had been the case at Nanhai, stagnatingwages and the inability/unwillingness of the union or management to take workergrievances seriously were the impetus for the strike. Although they were clearlyfollowing the lead of the Nanhai workers, one Denso employee said, “we wanted to go onstrike for a long time.”214

The most significant thing about the Denso strike was the militant and tightlyorganized nature of the strike. During the weekend before the 21st, a group of up to 200workers gathered in secret to discuss plans. At this meeting, the “three nos” strategy wasdecided upon by workers. The three nos required that for three days there would be nowork, no representatives, and no demands. Workers knew that within three days lack ofparts would cause a shutdown in the neighboring Toyota assembly plant. Additionallythey had strong reason to believe that representatives would face retaliation or co-optation. By waiting three days before issuing any demands, they would be bargainingfrom a position of great power, as they anticipated that losses would be mounting not justfor Denso, but for the entire Toyota supply chain.

212 It is important to note that in interviews with Nanhai strikers, there was hardly anyevidence of anti-Japanese or overtly nationalist sentiment. And many strikers expressedgreat opprobrium for their Chinese managers.213 June 24, 2010. “riben dianzhuang zai hua qiye bufen fugong, laozi tanpan reng zai

jinxing.” [Japanese Denso Company in China has Partial Resumption of Work,Negotiations Continue] Reuters.214 Ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (148)


Their calculation was correct. They started the strike on the morning of the 21st,and workers blocked trucks from leaving the plant. By that afternoon, six of the otherparts plants in the “auto city” industrial zone where Denso was located were also shutdown. The following day, the nearby Toyota assembly plant closed for lack of parts. 215

And on the third day, workers elected 27 representatives and went into negotiations.Their central demand was simply a wage increase of RMB 800. The negotiations wereattended by worker representatives, the CEO of Denso (who had flown in from Japan),officials from various levels of the union, as well as Zeng Qinghong (the general managerof Guangzhou Automotive and National People’s Congress delegate who had played akey role in resolving the Nanhai strike). Following the negotiations, management andZeng pleaded with the workers to return to work and promised them a resolution by the25th. Worker solidarity broke down somewhat at this point as many began to return towork, although output on the 24th was far below usual levels. And on the morning of the25th it was announced that they would be granted an RMB 800 wage increase.

The evolving political position of the union was evident in the Denso strike,which took place within the jurisdiction of Chen Weiguang’s GZFTU. When the workersfirst went on strike, they rejected the intervention of the higher levels of the trade union,likely with the Nanhai incident fresh in their minds. But under the leadership of ChaimanChen, the union proclaimed that it would not play a “mediator” role as requested by thegovernment, but that it would represent the workers’ interests and only the workers’interests (rather than the typical rhetoric about win-win outcomes). Additionally, andquite significantly, the local public security bureau told the union they wanted to speakwith worker representatives, but the union refused to comply.

Workers struck in another nearby Honda supplier, again severely disruptingproduction. At the Guli Lock Factory in Zhongshan, workers marched in public andblocked roads. As early as May 28th, Hyundai Automobile workers walked off the job.And an absolutely massive strike wave engulfed a development zone in the northern cityof Dalian. From the end of May through August, 70,000 workers in a variety of sectorswent on strike, affecting 73 enterprises in the zone. As reported by union officials,workers won an average wage increase of 34.5%.216 This was the third time the Daliandevelopment zone had experienced a strike wave since 1994. Some Japanese companiesthat had not experienced strikes decided to give their workers a preemptive raise, withNational increasing wages by RMB 500 and Panasonic 200-300.217

It is not clear how many workers went on strike in the summer of 2010, but it iscertain that the dozens of reported cases are merely a small portion. Around the country,workers were granted unprecedented wage increases. In the fallout from the strike wave,union officials and media commentators declared the end of low wage labor in China.

215 Ibid.216 September 20, 2010. “dalian tinggong chao 7 wan ren canyu boji 73 jia qiye, yi

gongzi zhang 34.5% gaozhong.” [70k participate in Dalian strike wave affecting 73enterprises, ends with 34.5% wage increases.” Caixin.217 June 12, 2010. “yi lang gao guo yi lang, bentian zhi suo gongsi ye baofa bagong.”[One wave is higher than the next, strike also erupts at Honda lock factory] Lianhe


Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (149)


Response to the 2010 Strike Wave

In the wake of Nanhai and other strikes, the public position of many unionofficials, particularly in Guangdong, began to change. Chen Weiguang, always somewhatmore daring than his other union colleagues, publicly stated that the strikes had been aforce for encouraging trade union reform and that it was allowing labor negotiationsystems to become more mature.218 He additionally expressed concern that unions hadcontinued to be passive and that they were “tailing” workers, publicly stating that unionscould not “hide forever.”219 In October, the GZFTU issued the “Advice on StrengtheningUnion Work and Developing the Organizational Function of the Union,” whichemphasized that in the course of resolving labor conflicts, “the union must be clear on itsrole, that it cannot be a mediator but rather only represent the side of workers.”220 KongXianghong was quite explicit in his criticism for the lower levels of the union, asindicated by the following statement: “…the union must not be antagonistic to strikingworkers, but the local unions were precisely standing in an antagonistic position. So theworkers said that they want to disband the union, that the union is a running dog and atraitor.”221 While hardly a radical position, Kong’s solution to the problem indicated aremarkable divergence from earlier positions held by most union officials: “I think thatcontrollable strikes are a right which should be enjoyed in order to govern a stable andharmonious society. What is harmony? Harmony is admitting conflicts and integratingthe conflicts into a systematized path for resolution. Many of us Party cadres are lackingprecisely the correct recognition of this problem.”222 Even GDFTU chair Deng Weilong,not known for his progressive politics, admitted that very few enterprise-level unionchairs in Guangdong are democratically elected. Deng went on to say that oneconsequence of this is that, “in the eyes of many workers, unions are an organizationsubordinate to the boss… when labor conflicts become acute, unions represent theinterests of the boss.”223 The position of ACFTU leadership was considerably moreconservative, as they mostly repeated old exhortations to increase unionization in privateenterprises. However they did issue a new specific call for increasing the wages ofproduction workers.

In some ways, the central Party leadership was ahead of the ACFTU in theirposition. Zeng Qinghong – the former Politburo Standing Committee member and vice-president of China, not the Guangzhou Automotive CEO – wrote a long statement on the

218 June 10, 2010. “Shenzhen ji tiao zuidi gongzi zhi 1100 yuan.” [Shenzhen makesemergency adjustment to minimum wage, now 1100 Yuan] Yikuo Meiri Caijing.219 Ibid.220 October 12, 2010. “guangzhou shizong: gonghui yao zuowei gongren liyi daibiao

canyu xietiao laodong guanxi.” [GZFTU: the union should serve as workers’ interestrepresentative in participating in labor relations.”221 June 22, 2010. “kekong de bagong shi hexie de yingyou zhi yi.” [Controllable strikesare a right that should be enjoyed for harmony] Zhongguo Qiyejia.222 Ibid.223 July 3, 2010. “Guangdong sheng zong gonghui zhuxi: qiye gonghui zhuxi duo bu shi

minzhu xuanju.” [GDFTU chair: most enterprise union chairs are not democraticallyelected] Yangcheng Wanbao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (150)


value and necessity of expanding and strengthening collective bargaining. While hestopped short of calling for the legalization of strikes, his argument implied that it wassomething worth considering:

In other countries, the strike is only the last resort and frequently is used as athreat; domestically, employees strike first and then negotiate. That is to say, they usestrikes and other extreme measures to win the right to negotiate. First, the production linestops, and then they consider how to resolve the problem. This sort of pre-emptive coerciveapproach expends a large volume of social resources, giving workers a high cost to protecttheir rights, employers a high operating cost, and the government high costs in protectingstability. This constitutes an unfortunate lose-lose situation.224

Zeng is referring in quite precise terms to the costs that worker insurgency exacts onaccumulation; a situation which, as he argued, requires an institutional response in theform of substantive collective bargaining. Staking out a more reform-oriented position,the other Zeng Qinghong (writing in his capacity as a National People’s Congressdelegate) said in a published article that in the process of resolving the Honda strikes he,“deeply felt the necessity of legislating rights for economic strikes.”225 Although theParty did not immediately embrace his call for national-level legislation, the topic was nolonger taboo.

Even representatives of capital began to question the sustainability of deeplyauthoritarian management systems. Li Chunbo, an auto industry analyst at investmentbank Citic Securities, argued that in light of the strike wave,

The government should intervene and help enterprises establish an internal system.For example, they should allow employees to have interest representatives, and allow laborand capital to have a platform for communication. There may be fears that people will usethis platform to argue, but it can also be an emotional outlet and used for resolvingproblems. This may allow for a reduction in extreme actions.226

The unambiguous implication of such a statement is that it was precisely the lack ofmeaningful interest representation which led to the strikes in the first place.

The moment, it seemed, was ripe for an institutional re-alignment. In pursuit of astabilization of labor relations, the Guangdong provincial government re-issued aproposed set of laws entitled, “Regulations of Democratic Enterprise Management,”which had been considered in 2008 but put aside when the economic crisis deepened thatfall. The regulations included an article which allowed for collective bargaining if 1/5 ofenterprise employees demanded it. Once such a decision was made, employees coulddemocratically elect their negotiating team, albeit under the supervision of the enterpriseor regional trade union. Once the request for collective bargaining was submitted to theemployer, bargaining had to begin within 15 days. In addition to these seemingly pro-labor provisions, the law also explicitly forbade workers from engaging in strikes orslowdowns while negotiations were taking place. However, as the government was

224 (accessed March 10, 2011)225 (accessed March 10, 2011)226 May 31, 2010. “bentian jiaxin shi bagong chongji di chengben zhizaoye moshi.”[Honda wage strikes are a shock to the low-cost manufacturing model] Caixin.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (151)


seeking public opinion on the draft regulations, they encountered stiff resistance fromrepresentatives of capital, most notably the Hong Kong General Chamber of Commerce.Before the regulations could be passed into law, it was once again shelved, althoughunion officials claimed it was out of concern for disrupting the Asian Games being heldin Guangzhou in November, rather than because of pressure from employers. In Octoberof that year, Chen Weiguang said that this was just a minor hiccup and that theregulations would certainly be passed in the future.

Even if voices calling for reform were gaining the upper hand, it is important tonote important limitations. While Chen Weiguang argued that strikes could be legalized,he consistently affirmed his opposition to “political” strikes. Even for relatively pro-labortrade unionists such as Chen, the discourse of accepting the “reasonable demands” ofworkers continued to hold sway. “Reasonable” is meant to imply that worker demandscannot go beyond immediate workplace issues, and that the authorities retain thepaternalistic authority to determine what constitutes reasonableness even among strictlyeconomic demands. Additionally, trade unionists by and large maintained the obsessionwith “stability maintenance” that characterized all government agencies during thisperiod. In September of 2010, a conference was held in the Luogang district ofGuangzhou, which in part was focused on how unions can effectively respond to strikes.Following the conference, the district union published a book summarizing theexperiences of various enterprise unions where it was argued:

…when handling spontaneous mass incidents, enterprise level unions definitely mustobjectively analyze and recognize the cause and substance of the mass incident. When theconflict is just sprouting they must quickly communicate with the relevant governmentdepartments, and quickly devise a strategy to contain the incident from expanding andintensifying.227

Even if collective bargaining was coming to be embraced by more unionists, it seemedlikely that most enterprise unions would continue to be reactive to strikes.

One Year Later – Assessing the Outcomes

But if major political changes were beginning at the provincial and even thenational level, what happened at Nanhai Honda in the months following the strike? Ofparticular importance is whether the promised wage increase actually materialized,whether union elections took place and collective bargaining mechanisms established,and what workers’ perceptions were of subsequent changes. Although there are continualreasons for pessimism, some important changes did in fact take place at Nanhai.

The announced wage increases of approximately RMB 500 for regular workersand RMB 600 for interns did in fact materialize, thereby distinguishing this case frommany others. But few workers were satisfied with the settlement, feeling that their wageswere still too low (and indeed they remained lower than those of workers in the joint-venture Guangzhou Honda). And with the possible exception of union re-organization

227 P.91 2010. Jiyu yu tiaozhan: gonghui gongzuo 100 li. [Opportunity and Challenge:100 cases of union work] Guangzhou: Guangzhou shi luogang qu zong gonghui.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (152)


(see below), the other 100+ demands put forth by workers had seemingly been brushedaside, something that was quite clear in this comment from one worker:

Actually, I’m not satisfied with the results of the strike. We also demanded an hour forour lunch break, and they didn’t agree to this. And us women wanted a day of rest forwhen we’re menstruating, but they didn’t agree to this. Menstruating is so uncomfortablefor girls, and they wouldn’t agree to even one day.228

It is thus apparent that even if management was willing to accede to some wage demands,other workplace issues continued to be ignored.

Even during the strike, some worker activists had been aware of the pressing needto build collective power, with some claiming that this was even more important thanwage increases. While such a politicized point of view was not generalized among theworkers, it was certainly an important current, particularly amongst the leaders. Althoughthere were no open calls for an independent union during the course of the strike, it is ofutmost importance to recognize that this was something that had been under discussionamongst the organizers:

Worker A: I think the union issue is more important than raising wages now. Actually, atthe time we considered a huge issue, whether to establish China’s first union. A democraticunion.Worker B: If it really happened, we’d be quite proud.Worker A: Really proud. When we wrote that open letter, we discussed establishing thefirst one in China… there isn’t anyone that really does things for us, and other unions ofcourse are the same, or even worse. So we wanted to take the union… because at the end ofthe day, the union is the representative of workers.229

That workers did not end up calling for an independent union is of course due to theirrecognition that this would certainly be met with harsh repression. A less contentiousframing was to call for union “re-organization,” which implied holding new elections forenterprise-level officers while remaining within the structure of the ACFTU.

What then was the outcome with regard to the demand for union re-organization?Such an important problem was not to be resolved by the consistently pro-capitaltownship union federation, but rather by the provincial authorities. Kong Xianghong, whohad been involved in strike resolution negotiations, took up responsibility for respondingto the Nanhai strikers’ demand for new union elections. Kong was personally tornbetween simply appointing a new enterprise chair from above and allowing the workersto hold a democratic election, as they had demanded. Indicating consultation with thecentral government, Kong announced shortly after returning from Beijing that ademocratic election would in fact be held in Nanhai, though it was notably onlyannounced in the Hong Kong media.230 In an unprecedented move, union leadership saidthat they would allow students and researchers from Sun Yat-sen University to observethe election proceedings. However, there was some subsequent backtracking. Rather than

228 Interview, July 4, 2010.229 Interview, July 18, 2010.230 June 14, 2010. “nanhai bentian ni minxuan gonghui zhuxi.” [Plans for Nanhai Hondato hold democratic elections for union chair] Ta Kung Pao.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (153)


allow for workers to immediately hold a general election to replace the discreditedenterprise chair Wu Youhe, elections were only held for team (ban) representatives – thelowest level of the enterprise hierarchy, roughly equivalent to a shop steward in the U.S.In each of the 30-40 person teams, there were relatively democratic elections, althoughinterns were not allowed to run. While it is unclear if management or governmentintimidation occurred, many were surprised that Li Xiaojuan, one of the lead organizersand negotiators who was the sole signatory of the bargaining team’s open letter, was notelected team representative.

From the perspective of most workers, there was not any significant change in theunion following the strike. More than a month after the strike concluded, one workerexpressed continual dissatisfaction with the enterprise union:

If you’re an employee it’s impossible for you to be heard by the higher ups. It’sdoesn’t matter who you talk to, it won’t work because you can’t get in a word with theoffice [management]. The union doesn’t come down to the shopfloor at all, they justcollect the dues and whatever, or organize one outing a year. To be honest, I don’t knoweven a couple people from the union. It was just that when we first came, there was awoman and she said she was from the union. So they just collect the dues each month andthat’s it.231

When a different worker who had been at the plant since 2008 was asked whether hewould participate in elections for union representatives, he said, “no, I won’t, because theunion is just those government people, it’s just for the managers.”232 Even more alarming,one of the workers who was elected as a union representative (at the team level) wascompletely dismissive about the possibility of union re-organization:

Anyway, they have this idea, but… the union is useless. Go ahead and change [theunion officers], but I’m indifferent to whether they change or not. If they change or not,there isn’t any use… I think they can’t help me with anything… So if this company has aunion or not, it makes no difference.233

This representative even said that the union was unnecessary for collective bargainingand that having a union, “is just another expense.”234 But the difficult position of theworkers vis-à-vis the union was revealed by one worker’s frustration when asked if hestill trusted the provincial union federation: “we have to trust them, even if we don’t trustthem we have to trust them. Who else is there to trust now? Nobody.”235

If workers remained suspicious and/or indifferent towards the union, by latewinter 2011, it was clear that power relations at Nanhai Honda had shifted in the wake ofthe strike. Starting in February, the enterprise union, working with the direct supervisionof Kong Xianghong from the provincial union federation, entered into collective wagenegotiations with management. The democratically elected team-level representatives

231 Interview, July 13, 2010.232 Interview, July ?, 2010.233 Interview, September 29, 2010.234 Ibid.235 Interview, July 18, 2010.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (154)


were able to take part, and their initial wage demand was for a monthly wage increase ofRMB 800. In three rounds of negotiations which appeared not terribly dissimilar fromsimilar processes in liberal democracies, offers and counter-offers were exchanged. Bothbecause of the intervention of the provincial union federation and an awareness of thepossibility of workers to mobilize independently, management bargained in good faith. Inthe end, the union got management to increase their initial offer quite significantly, andworkers won a monthly increase of RMB 611.236 It appeared as if a meaningful system ofcollective bargaining – one in which real bargaining takes place – was beginning to takeshape.

This seeming breakthrough must of course be qualified with many significantprovisos. The first is that wages at Nanhai were very low when compared to the joint-venture assembly plants. Strikers had specifically cited the much higher wages atGuangzhou Honda as a reason for their actions, and post-strike their wages remainedlower than those at the joint-venture. Second, the structural power of workers at Nanhai ismuch more significant than workers in either the service industry or other manufacturingsectors. Perhaps only workers in transportation or energy exercise a comparable capacityto inflict damage on the economy by withholding their labor. A third and directly relatedpoint is that the wage increases coming from the strike resolution and from the winter2011 round of collective bargaining were presided over by the provincial authorities. Inless critical industries and in less economically developed regions of the country, thelikelihood that workers would have backing from provincial authorities is slim. Andfinally, it is worth emphasizing again that non-wage issues were completely unaddressedin the subsequent round of collective bargaining. Following the second wage raise, KongXianghong proclaimed that, “collective wage negotiations at Nanhai Honda will becomea classic case of Chinese unions for MBAs.”237 Perhaps this is so; but its replicability isdubious.

Insurgency and Institutionalization

What does the comparison between Otis and Nanhai Honda tell us about therelationship between insurgency, decommodification and incorporation, anddevelopment? Why did strikes in similar factories in neighboring cities producedramatically different results?

To review briefly, the strike at Otis was, in the view of one of the participants, an“absolute and thorough” failure. Worker sentiment aside, the implementation of a piece-rate system fits my definition of commodification of labor, and specifically compromisesworkplace security. Management argued that a piece-rate system needed to beimplemented because fierce competition in the industry required that they increaseproductivity. Piece-rates are a key strategy of capital to shift risk onto individualizedworkers and to dissolve potential bases of solidarity, and in this sense increases thesubmission of needs satisfaction to the logic of the market. Additionally, the dramatic

236 March 3, 2011. “nanhai bentian tinggong shijian, rifang zuizhong jieshou zengjia

gongzi fang’an.” [Nanhai Honda work stoppage, eventually the Japanese agree to a wageincrease] Nanfang Ribao.237 Ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (155)


decrease in wages that were experienced by workers across the board means that theyhave fewer financial resources at their disposal to satisfy their needs. This means thatworkers were increasingly compelled to sell their labor power at whatever pricemanagement offered. Increased financial reserves that come with a higher salary decreasethe urgency of this situation. And finally, the manner in which the union, capital, and thestate combined forces to reject any worker participation in the determination of the laborprocess reinforces a lack of labor process control, another measure of the commoditycharacter of labor. Though it would be difficult to argue that labor process controlsuffered as a result of the strike (since there was none to start with), the completerejection of workers’ ability to have meaningful participation or representation wasreinforced by this incident. In short, while workers rejected and rebelled against increasedcommodification, state, union, and capital allied in repressing this mini-countermovement. The commodity character of labor was not only maintained but enhanced.Unsurprisingly, a second attempted strike took place less than a year later. Through thisprocess, workers came away with little reason to believe that the union at any level wouldeffectively represent them in the future.

In Nanhai and other workplaces that were affected in the 2010 strike wave, theresults were quite different. Workers at Nanhai won a huge wage increase (in percentageterms) in June of 2010, and gained an even larger wage hike (in absolute terms) withoutstriking in the winter of 2011. This increase in wages enhances the workplace security forthe employees, albeit to a limited extent. Of course, two worker activists were firedwithout the union protesting, indicating continuing managerial authoritarianism. Therewas no agreement to transfer temp workers into regular workers, implying ongoingprecarity for a huge portion of the workforce, as well as compromised social protection(since these workers do not receive social insurance). Demands related to workplaceissues continue to be ignored by union and management alike, indicating that there havenot been improvements in labor process control. Thus, from an economic standpoint,workers in Nanhai were able to make gains in workplace security while other indicatorsof decommodification remained essentially unchanged. Otis workers, on the other hand,experienced increased commodification.

When we look at indicators of the political incorporation of workers into theunion, we see some important differences as well. At Otis, GZFTU and ACFTU officialsintervened to try to broker a resolution to the strike. But there was no attempt made to re-organize the union or to build a sustainable organization capable of bargaining withcapital. As soon as union higher ups moved on, management reneged on the deal, andworkers were subject to the very enhanced commodification that they had originallyresisted. In this case, the union was acting merely to ameliorate the immediate crisis ofsuspended production; as soon as that issue was resolved, they returned to passiverepression which allowed managerial authority to continue unabated. Workers continueto hold the union in contempt, and are unlikely to be able to resolve future grievancesthrough the enterprise-level union organization.

Nanhai represents a somewhat different outcome in this regard. It is true thatregular workers continue to think that the union is useless, and that even some electedrepresentatives are highly doubtful of its capacity to bring about important change. Thatbeing said, a number of team-level representatives were elected, and subsequently wereallowed to partake in collective wage negotiations. While I have already elucidated the

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (156)


many conditions that make Nanhai exceptional, these negotiations produced a verysignificant wage increase for the workers. Although it is unclear what will happen withthe Nanhai enterprise union once the provincial union federation withdraws from directsupervision, a precedent for continual rationalized modes of contention has nonethelessbeen established. Depending on the specifics of how collective bargaining takes place(e.g. are rank and file involved in formulating demands, mobilized during contract fights,kept in regular contact with leadership, able to democratically ratify contracts, etc.?) thismay increase worker identification with the union. In the aftermath of the strike, it did notappear that workers had any greater interactions with the union or hopes that it couldimprove things for them in the future. Even when the provincial unions preside oversubstantial wage increases, such actions may be insufficient to result in the politicaloutcome of worker incorporation. This suggests that union oligarchy may be a greaterobstacle to political incorporation than it is to decommodification.

When trying to explain the different outcomes in these two incidents of workerinsurgency, one could refer to the differences between elevator and auto partsmanufacturing. Certainly, elevator production is not as crucial a feature of the economyas is auto, and the construction industry does not depend on just-in-time manufacturing(as is the case in the auto sector). That being said, with real estate speculation coming tobe a key element in national development in China (Hsing 2009), elevators are hardlyunimportant. Because of the structure of the supply chain, the stoppage at Nanhai hadimmediate effects for a number of other plants, something which was not true in the sameway at Otis. And Otis workers started with wages which were closer to those ofGuangzhou Honda, and therefore considerably higher than those of the Nanhai workers.

Without discounting the differences between auto parts and elevatormanufacturing, I argue that these differences cannot account for the divergent outcomesbetween our two cases. Rather, between late 2007 and early 2010, two inter-relatedrelated currents – one political and one economic – led the provincial and centralauthorities to take a very different approach in responding to worker insurgency. On thepolitical level, 2008 and 2009 saw unprecedented levels of labor conflict (with the PearlRiver Delta as the epicenter), in which labor arbitration courts were completelyoverloaded and worker insurgency frequently spilled out into the streets. Economically,the onset of crisis increased the central government’s determination to shift away from amodel of development so wedded to exports. And in Guangdong, the provincialauthorities gained confidence that they occupied a less subordinate position vis-à-vistransnational capital than had previously been the case.238 When Otis workers went onstrike in late 2007, the impetus of provincial and central authorities was merely to avoidinstability and to get workers back on the line, regardless of the outcome; but by spring of2010, they were interested in resolving the strike and improving the consumptioncapacity of workers – even if they were still a long ways away from realizing the Fordist

238 Without a doubt, the position of the provincial government vis-à-vis capital varies byindustry. Even if the government became willing to sacrifice low value-added industries,the threat of capital flight both to other countries and other regions of China remains aconcern. But given the dense cluster of auto suppliers in Guangzhou and the relativelyhigh costs of relocating production, the government likely felt greater confidence indealing with foreign automakers.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (157)


ideal of being able to purchase the cars they were producing. As a result, once the Nanhaistrike started the higher-level authorities decided to open up more space for insurgentworkers to overcome the repressive tendencies of the local state, a tendency that wasmade apparent when the township union federation physically attacked the strikers. Andthe provincial union’s continual attention to the plant is an indication that they would liketo build up institutional capacity at the enterprise level which will be resilient enough toovercome the opposition of capital and the local government. Whether they will be ableto do this while keeping the genie of autonomous worker power in the bottle remainsunclear.

Although in this project I am largely not concerned with the specific dynamics ofmigrant worker resistance, the topic is worth briefly addressing here. Two relativelyunique features appeared in the Nanhai strike and other strikes in the summer of 2010.The most significant difference with most previous outbursts of insurgency (includingthose during the major factory closures in 2008-9) was that workers were on the offensiveand that they won economic gains. And second, the issue of union re-organizationbecame more central, an indication of the germination of political consciousness. Butthere were many ways in which older patterns of resistance continued, both of which aresomewhat counter-intuitive. The first is that worker protest remained highly cellular.Despite the fact that there was a major strike wave in which workers were clearlyinspired by what was occurring in other factories, there was no coordination betweenstrikers from different factories. The other is that workers’ sense of fidelity to the lawremains quite strong. In the course of an interview, one Nanhai striker revealed thecontradictory consciousness that is prevalent among insurgent workers. On the one hand,he recognized the efficacy of legally ambiguous strikes: “That’s right, [the company’sinternal complaint system] was not at all effective. Stopping work is effective, as soon aswe stop work we see the Japanese people, as soon as we stop working the Japanese willimmediately ask what the problem is.”239 And yet, just a few minutes later, the sameworker said, “I hope [the union] will teach us about labor laws, so in the future we’ll bebetter able to defend our rights according to normal procedures.”240 None of the workersfrom Nanhai expressed discontent with the laws themselves, a remarkable confirmationof the persistence of highly legalistic modes of thought, if not always action, amongworkers.

In conclusion, a comparison between Nanhai Honda and Otis allows us to see theinsurgent-institutional dialectic unfolding during the key years of 2007-2011. Up to andduring the economic crisis of 2008-9, oligarchy had blocked a substantive institutionalresponse from the union. However, with the 2010 Nanhai strike the higher-levelauthorities allowed for the strike to continue in order to overcome the strong state-capitalalliance at the local level and to raise workers’ wages. A combination of continualinsurgency as well as a degree of elite support produced gains that serve as an indicationof the institutional moment of the countermovement. And yet, even if material conditionsimproved, continual worker disillusionment and the involvement of the provincial unionare indications of persistent oligarchy, a phenomenon which should cause us to considerhow sustainable and “exportable” the compromise at Nanhai really is. Indeed, as one

239 Interview, July 25, 2010.240 Ibid.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (158)


striker at Nanhai implied, it is not just economic but also political issues that underliegeneralized worker discontent:

Of course we don’t take the same position as those in power… a lot of people have thesame difficulties as us, and have the same thoughts. Actually, the reason our Honda factoryended up like this [causing a strike wave], is because this is already a social problem, it’snot just as simple as raising wages.241

The state and union at all levels maintain a steadfast dedication to the isolation ofpolitical and economic struggles; whether such a separation will continue to be feasible inpractice is unclear.

Chapter 7


This work has provided an analysis of the dynamics of labor politics in Chinaduring capitalist industrialization.242 I began with an empirical observation: during theeight years after the Chinese central government began making symbolic and materialmoves towards class compromise in 2002, migrant worker insurgency grew rapidly.While there are a number of factors behind this, the overarching reason is that while thecountry experienced an economic boom of historic proportions things did notsignificantly improve for migrant workers. For China scholars intimately familiar withthe internal tensions and conflicts within the state apparatus, this phenomenon may nothave come as such a surprise. Indeed, the past three decades of reform have resulted inthe emergence of a well-documented alliance between the local state and capital, ascenario that has resulted in many potentially pro-labor laws and collective agreementsgoing un-enforced. But if this is the case, my research has asked how we can explain asituation in which labor is strong enough to force concessions from the central state butnot strong enough to significantly benefit from (i.e. enforce) these concessions.

At the risk of oversimplification, my answer has been that it is because the onlyunions that are allowed to exist – those subordinate to the ACFTU- are characterized by

241 Ibid.242 China of course experienced several decades of capitalist industrialization beginningin the late 19th century. However, this development was limited in geographic and socialscope and was suspended once the Communists prevailed in the civil war.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (159)


persistent oligarchy which grants high levels of access to the state but prohibits theconstruction of autonomous collective power for workers. As a result, I have argued thatChina has fallen into an “insurgency trap,” in which efforts by the central and sometimesprovincial and municipal governments to improve conditions for workers aresystematically undermined by the specific political economy of China in which theinterests of the lowest levels of the state are tightly aligned with the short-term interestsof capital. The one way to break this alliance and ensure legal compliance – opening upspace for the development of a potentially coercive countervailing force at the point ofproduction, i.e. independent worker organization – remains anathema to higher levels ofthe state since they fear it would threaten their monopoly on power. Hence, theconundrum of insurgency trap.

Based on the experience of earlier industrializers, there is ample evidence thatsome form of organization is necessary to decommodify labor and incorporate theworking class, both of which are likely preconditions for reducing insurgency. I haveemphasized the distinctiveness of the Chinese case by tracing the historical trajectory ofthe ACFTU and its relationship to the contemporary class of migrant workers. We haveseen that contemporary Chinese labor is characterized by appropriated representation,implying very different political dynamics than was the case for the labor movements thatPolanyi might have been familiar with. At the most general level, working classformation in the West resulted in the construction of unions that while prone to oligarchictendencies over time, nonetheless were originally constituted in opposition to capital andthe state. In China, on the other hand, an entirely new urban proletariat has emergedunder conditions of appropriated representation in which the sole and universalrepresentative of the working class is already deeply integrated into the state. In thissense, it is evident that dynamics of representation between insurgent but dispersedworkers and union officialdom becomes a key problematic for understanding theposition, trajectory, and capacity of Chinese labor as a political actor.

These peculiarities of labor politics in China called for a re-configuration of thePolanyian countermovement. Based on his study of similar movements in the West,Polanyi conflated resistance to the market with actual decommodification. But when welook at the experience of China over the past three decades, increasing levels of workerunrest have by and large not resulted in the decommodification and incorporation that wemight expect based on such theory. Thus, I argue for a distinction between the insurgent

and institutional moments of the countermovement. While ongoing worker unrest in theface of deep commodification of social life confirms the existence of the first “moment”in China, we have also seen that oligarchy within the union structure has complicated theprocesses of political incorporation and decommodification that constitute theinstitutional moment.

That a highly repressive state seeks to keep the working class atomized underconditions of capitalist development is hardly surprising; what is less expected is that, ina striking confirmation of Polanyian theory, we have seen that even highly oligarchicunions are actively searching for a response to the chaos engendered by the free market. Ihave shown that unions in the most developed provinces in China (Zhejiang andGuangdong) have felt the pressure of worker unrest quite acutely, and are attemptingvarious different types of reforms. In Guangzhou, we saw how union federation chairChen Weiguang argued forcefully that unions must represent workers and only workers, a

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (160)


dramatic rhetorical departure from the norm. Chen also successfully won manylegislative and regulatory victories for workers. When we look at the development ofsectoral unions, a type of union that the national leadership is pinning high hopes on, wesee that different regional models of development produce different political possibilities.While the globalized economy of the Pearl River Delta makes it very difficult for unionsto engage in sector-wide bargaining, the locally-driven economy of Zhejiang presentsgreater opportunities. Despite all of these changes in union form and rhetoric, we see thateven in the most likely cases, unions had little success with advancing the economicinterests of migrant workers or increasing their own legitimacy amongst membership. Intheoretical terms, I argue that unions by and large failed to incorporate or decommodifylabor, implying that the countermovement remained stalled at the insurgent moment.

But between 2007 and the spring of 2010, some of the dynamics of insurgencytrap began to shift. Intensified worker unrest in 2008 and 2009 and a credit crunch in theWest caused the central government to become more assertive in supporting an expansionof domestic consumption. When a strike broke out at Nanhai Honda in May 2010, thecentral government provided implicit support, thereby affecting both the process andoutcome of this particular instance of insurgency. As a strike wave spread throughoutGuangdong and other regions of the country, workers began to make offensive, andsometimes political, demands, many of which they won. While the political outcomes ofthe strike wave are still not entirely clear, it has resulted in some important changes in theunion. One year after the strike wave broke out, the union at Nanhai Honda presided overa round of collective bargaining in which substantive bargaining took place – asignificant departure from previous experience. While workers did win a large wageincrease, other workplace issues were unaddressed and the workers remain suspicious ofthe union. That being said, it was clear that the severity of the insurgent outburst began toresult in some institutional changes, as union officials and legislators from Guangdongbegan openly calling for the legalization of strikes. While it would be impossible to arguethat relative success in one factory qualifies as institutionalization, the complexinteractions between workers, unions, the employer, and the various levels of the staterevealed different dynamics from just a year or two earlier. Particular as the case mayhave been, it revealed important underlying structural changes, and indicated what thecontours of labor and the countermovement may look like moving forward.

What have we learned about labor in China?

There are many features of labor politics in China that I have discussed which willnot come as a surprise, e.g. unions on the shopfloor are weak, workers are suspicious ofunion leadership, unions work with state and capital to repress resistance, etc. But thereare several advances, both conceptual and empirical which are worth emphasizing.

The first is that I have sought to show how factors other than control by the Partyconstrain and/or enable union activity. These include both internal organizational andexternal economic factors. I have demonstrated that the historical context of theformation of the ACFTU led the union both to its alliance with (and subordination to) theParty and to the development of a particular organizational logic. I argued that aparticular practical orientation was established at the founding of the ACFTU in the1920s which has persisted to the present – despite instances in which individual leaders

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (161)


sought reform. This basic principle has been to promote the inter-related goals of ethno-national autonomy and an expansion in productive capacity (which are of course sharedby the Party). Such an organizational logic developed because capitalist development inChina took place in the context of semi-colonial subjugation, a marked difference fromthe political situation for labor in Western countries. Thus, the interests of the workingclass were always seen as a particular subset of the general interests of the nation, aformula that persists to this day. With this perspective, we can see the continuity betweenthe ACFTU’s radical past and its deep conservatism since the revolution. While theapproach to advancing ethno-national autonomy and an expansion in productive forceshave changed (mass mobilization in the 1920s, the iron rice bowl in the Mao era, andmarketization since the late 1970s), the basic organizational ends of action have not. Thuswe can see that ACFTU conservatism derives not just from the external constraint ofsubordination to the Party (a relationship that they have been unable to break free from),but also because of a particular internal logic that results in a very different orientationvis-à-vis capital and the nation than was the case for Western unions.

In addition to historical trajectory, I have argued that external economicconditions can be as crucial in determining union activity as the political constraintswhich have received much more attention in the literature. Through an analysis of casesthat vary across space (Zhejiang vs. Guangdong) and time (Otis vs. Honda strikes), wesee that different models of development and the dynamics of the global economy have aprofound impact on the possibilities for organizational innovation and collectivebargaining. Because of these opportunities and constraints posed by the various regionalpolitical economies in China, I have highlighted how a variety of markedly differentsystems of labor relations may emerge within a single country.

Empirically, I have presented new material on ACFTU unions which revealspreviously under-studied trends. Most significantly, I have documented the emergence ofsectoral unions in China, an organizational form that will be an increasingly centralfeature of Chinese trade unionism and industrial relations in the future. By linkingdevelopments in sectoral-level collective bargaining to regional political economy, I haveshown that the success of such efforts will be geographically uneven. It seems that aEuropean-style region-based approach to collective bargaining will be more likely inZhejiang than in Guangdong, while the latter appears headed toward a more American-style firm-based approach. Perhaps most importantly, the glowing terms in which tradeunion officials around the country refer to the perceived success of Zhejiang sectoralunions reveals a fundamental feature of the organizational logic of contemporary ACFTUunions – the pursuit of what I have termed “oligarchic decommodification.” While newforms of organization and bargaining are proliferating around the country, the oldproblem of weak enterprises unions emerges as an issue here as well. In any event,sectoral-level unionization and bargaining are sure to remain high on the agenda ofChinese unionists, and these organizational forms are likely to shape the contours ofworker resistance in the future.

The organizational innovation of sectoral unions is a response to workerresistance in general; however I have also documented union responses to a number ofspecific incidents of strikes. I concur with existing literature that unions are essentiallywithout exception reactive to strikes and other forms of bottom-up activism, even in theleast likely cases (such at the Guangdong Hotel). However, the comparison in chapter six

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (162)


reveals that changing dynamics in political economy imply that, at least in certainindustries and in certain regions, unions may begin to leverage autonomous workerresistance in pushing for increased wages and benefits.

Conceptually, I have argued that the research calls for a re-conceptualization of“labor,” as worker-union relations in China’s current period of capitalist development aremarkedly distinct from those in earlier industrializers. Because China’s new workingclass has emerged under conditions of appropriated representation, it is of crucialimportance to problematize relations of representation. I have argued that “labor” shouldbe regarded as the dialectically constituted whole of the distinct entities of workers and

unions in their political activity. While representation necessarily “connect and cuts,attaches and separates,” (Hardt & Negri 2004:241) conditions of appropriatedrepresentation lock representatives and members together in an uncomfortablerelationship in which the “attachment” is coercive and the “separation” is profound.While emphasizing that there is no normative assessment involved in the term, I arguethat the ACFTU does represent the working class if only because the authoritarian statedemands it of them, a scenario which has important practical consequences. Thus, workerinsurgency results in organizational responses from the union, the best examples of whichare the experiments with sectoral unions. At the same time, shifts in union activity in turnshape the dynamics of worker resistance. This was eminently apparent in the case of theHonda strikes, where workers initially started with simple wage demands, but inresponding to union repression and inactivity quickly came to insist on union re-organization. Subsequently, democratic elections of limited scope took place whichallowed some rank and file to participate in two rounds of collective bargaining. Finally,one of the consequences of the Honda strikes has been a renewed push by union officialsfor legalizing strikes which, if enacted, will certainly yet again alter the dynamics ofworker resistance. The dialogue between insurgency and institutionalization within laboris apparent enough.

Making the distinction between unions and workers helps us to understand why itis that labor is strong enough to win concessions from the state (at various levels), but isnot strong enough for migrant workers to significantly benefit from these concessions.On the one hand, dispersed worker insurgency has become a powerful political force atthe aggregate level, since it represents a strong threat to the stability of the system. Thishas strengthened the hand of unions in pushing for pro-labor legislation, widespreadincreases in the minimum wage, and sectoral level collective bargaining agreements,instances of which have all been discussed in the preceding chapters. Especially given thedeclining fortunes of workers in the developed world, these gains seem exceptional andare certainly an indication of the strength of labor. On the other hand, we find that at theparticular level, i.e. on the shopfloor, labor is quite weak. Workers are weak because theyare disorganized and there is no credible strike threat, while unions are weak because oftheir heteronomy vis-à-vis capital and the state. The specter of generalized insurgency isa concern for the higher levels of the state, but not nearly so much for a particularemployer or local officials. Finally, we can see that in instances in which workers buildstrength on the shopfloor, as at Honda, there emerges the possibility of drawing on thegeneral power of labor (winning support from higher levels of the union and state) inbolstering unions within the enterprise.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (163)


And yet, we have seen that unions respond positively to worker strikes only inexceptional situations. The tension between workers and unions, between represented andrepresenter, has already become and will continue to be a key site of class struggle. Muchas Ching Kwan Lee argued that legal reforms are both shaped by and end up shapingworker struggles, the same is increasingly true for organizational reforms within theunion. Again, this is quite different from most prior experiences of labor politics duringindustrialization, during which time unions and the working class developed relatively intandem. 243 Given conditions of appropriated representation and the much greaterrepressive capacity of the state, migrant workers in China are forced to fight traditionalbattles against employers, but also must confront and transform their own unions to amuch greater degree than was true previously. Whether or not dispersed workerinsurgency will create enough political pressure to force changes in the union sufficientto reduce unrest is unclear; however, struggles over representation, i.e. political struggles,have emerged within labor and are likely to increase.

In concluding, I would like to highlight the implications of this study for globalcountermovements against the market and the future of global capitalism. I will firstsituate China within the world system while indicating potential systemic risks. Next, Iwill return to the domestic level in discussing how China may play a key role in avertingsuch risks, and will then assess the likelihood of various outcomes. Finally, I will providea discussion on the relationship between autonomous worker movements and traditionalunions, and what sort of politics might allow for the construction of a powerfulcountermovement at the global level.

China and Global Capitalism

As early as 1999 Giovanni Arrighi and Beverly Silver argued that the core of thecapitalist world system was shifting away from United States and towards East Asia –China in particular (1999:286-9). With the U.S. accumulating a massive current accountdeficit (due in large part to trade with China) and descending into expensive militaryquagmires overseas, the passage of the “Belle Époque” of American hegemony becameincreasingly evident in the first decade of the 21st century (Silver and Arrighi 2003). Butif a China-centered global economy is becoming more of a reality every year, theconsequences of such a massive shift are not immediately evident:

How drastic and painful the transformation is going to be – and, indeed, whether itwill eventually result in a commonwealth rather than in the mutual destruction of theworld’s civilizations – ultimately depends on two conditions. It depends, first, on howintelligently the main centers of Western civilization can adjust to a less exalted status,and, second, on whether the main centers of the reemerging China-centered civilization can

243 I do not want to appear Pollyannaish about union democracy in the West or otherdeveloped countries during the period of industrialization. While being aware of all sortsof corruption, “yellow” unionism, etc. in other countries, the point is that there are quitedistinct general tendencies in China, which means that representation is a more acuteproblem at the class level.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (164)


collectively rise up to the task of providing system-level solutions to the system-levelproblems left behind by U.S. hegemony.” (Arrighi and Silver 1999:286)

There is of course debate about what precisely will result from China’s rise.Arrighi, while adding several conditions and provisos, maintains a great deal of optimismabout the possibilities that an emergent China holds for the stability and fairness of theworld economy. He dusts off the Smithian theory of a “commonwealth of civilizations”in arguing that a declining U.S. hegemon will face greater restraints with an empoweredChina acting as a counter-weight (2007). Additionally, he maintains a degree of optimismthat China’s rise will not necessarily result in the ecological disaster that many anticipate.Drawing on the work of Kaoru Sugihara (2003), we learn that China experienced an“Industrious Revolution” starting in the 16th century, which resulted in a labor intensive(rather than energy intensive) development path, a set of circ*mstances which resulted indivergence from Britain and the Western industrial revolution. Somewhat dubiously,Arrighi argues that this experience may help China avoid the energy intensivedevelopment strategy pursued in the West, thereby potentially averting the impendingdestruction of the biosphere.

Li Minqi is far less sanguine as is evident in the title of his book, The Rise of

China and the Demise of the Capitalist Economy (2009). Although he does providepotential alternate outcomes, Li argues that the world system cannot bear the fullintegration of China into global capitalism. In addition to concerns about ecologicalcatastrophe (Li 2007), he fears both internal social chaos and the possibility of“downward convergence,” (Li 2005:436) in which other countries in the semi-peripherycontinue to push down wages to compete with China. Both Li and Ho-fung Hung (2008)express concern that high levels of investment (particularly in infrastructure) and relianceon foreign consumer markets leaves China vulnerable to severe economic shocks.

This then brings us back to Arrighi and Silver’s second condition for avoidingsevere upheaval in the world system: the necessity for a “system-level solution” from the“China-centered civilization.” The systemic problem to be addressed is – among others –precisely the issue of under-consumption in China that Hung identified even before theintensification of the financial crisis in 2008. It is here where the specifics of laborpolitics become of central importance. The introduction of some form of classcompromise, including enhanced social protection and workplace security, will benecessary to reduce China’s high savings rates and to develop a new class of domesticconsumers. Ongoing economic troubles in the world’s developed countries only enhancethis imperative. But making such a transition is of course a political problem, one whichwill likely require the incorporation of the working class. In my terms, the institutionalmoment represents a potential solution to one of the central “system-level problems.” Atthe risk of grandiosity, the future of global capitalism may hang in the balance.

A New Class Compromise?

The problem as defined above of course echoes the situation found during anearlier systemic crisis – the Great Depression. Fordism and the construction of thewelfare state were primary means for addressing this crisis, steps which allowed for the

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (165)


brief flowering of American hegemony.244 In the West, social democratic and other non-revolutionary left parties allied with unions in demanding a greater share of profits fromcapital, which in the post-war era led to three decades of relatively high growth and areduction in inequality. But contemporary China is of course very different from Euro-American countries of the early-mid 20th century, as is the global context. I have soughtto address the question of, under what conditions has the Chinese state attempted torealize the sort of class compromise and decommodification (that we would expect fromPolanyian theory) through the means of a highly oligarchic trade union structure? Andthere are many reasons to believe that such a program will continue to encounter severeobstacles.

The first reason, which has been an empirical focus of this project, is the crisis ofrepresentation within the ACFTU. We have seen that heteronomy vis-à-vis both state andcapital has rendered union organizations in the workplace – the space were formal lawseither have material effects or dissipate into the ether – frequently impotent and incapableof enforcing their own agreements. Model trade unions in Guangzhou and sectoral unionsin Zhejiang bearing the blessing of the Premier have by and large failed to incorporate ordecommodify labor (even if the formal mechanisms seem to be in place, as in Zhejiang).In enterprises rocked by militant and autonomously organized strikes during the spring-summer of 2010, we see that workers are able to make material gains; but it is unclear towhat extent union organizations will be forced to engage in substantive reform. Certainly,persistent union oligarchy continues to present a challenge to institutionalizing classcompromise.

A second and related challenge to the realization of class compromise in China isthe lack of a developed radical political agenda. Particularly in many of the Europeancases, the threat of socialist revolution appeared quite real. The potential for theelimination of private property was of course quite unsettling for capitalists (and theirallies in the state) and so labor unions and parties were able to exact relatively highconcessions in exchange for abandoning a revolutionary agenda. It is important to recallthat class compromise, as traditionally conceived, implies mutual concessions by both

labor and capital. Labor agrees to “trade-off the abolition of private property of the meansof production” (Przeworski 1980:56) in order to secure lower rates of exploitation bycapital and expanded social protections from the state. Additionally, labor consents toefficiency producing arrangements in production so as to maintain the profitability of thefirm. However, worker activism in China is not a traditional social movement but aninsurgency; thus, few if any formal political demands (e.g. abolishment of privateproperty) are articulated by representatives. The trade unions that claim to representworkers sometimes ask capital for some improvements; but without the capacity to stopproduction, such requests are by and large ignored. Thus, while particular instances ofworker insurgency can result in heavy losses for capital, there is no credible threat tocapital at the aggregate, i.e. political level. Without a unified effort from various levels of

244 As is the case for Arrighi and Silver, I use this term in the specifically Gramsciansense, i.e. domination constituted primarily by consent rather coercion. In the view ofArrighi and Silver, the period of American hegemony was notable for its brevity, andquickly lapsed into coercion.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (166)


the state, something which has yet to emerge, it is difficult to imagine how capital couldcollectively agree to some sort of compromise.

Related to this last point is third difficulty, namely the problem of the strongalliance between the local state and capital and the resulting non-enforcement of relevantlabor laws. This is of crucial importance, because in order for a class compromise towork, “The state must enforce the compliance of both classes with the terms of eachcompromise and protect those segments of each class that enter into a compromise fromnon-cooperative behavior of their fellow class members.” (Przeworski 1985:202) Even ininstances when relatively decommodifying legislation or collective bargainingagreements are enacted, implementation remains a severe problem. The mutual lack oftrust resulting from such an anarchic situation was amply apparent with the sectoralunions in Zhejiang. Given the significant individual incentives for violating suchagreements, it will be difficult to enforce a substantive compromise at the class level untilthe state can act as a credible and relatively disinterested245 broker.

Finally, and perhaps most vexingly, is the extent to which China is deeplyintegrated into the global economy. International trade and investment is not a newphenomenon, and was certainly quite prevalent in the period of British hegemony in the19th-early 20th centuries. But given China’s heavy reliance on exports and foreigninvestment, not to mention WTO agreements, the imposition of institutionalizeddecommodification at the national level in today’s world would be tenuous at best. Asremarked upon by Peter Evans (2000; 2005; 2008), it is likely that a contemporarycountermovement against the market must be as globalized as capital. Although therehave been overtures towards greater international solidarity between the ACFTU andforeign unions, little substantive cooperation has developed thus far (David-Friedman2009; Luce and Bonacich 2009). Thus, the globalized nature of the contemporaryeconomy will likely continue to present challenges to institutionalizing thecountermovement.

If a class compromise of the type that industrialized nations pursued from themiddle of the 20th century seems unlikely in contemporary China, that does not mean thatcertain segments of the state won’t continue to try. The central government and certainlower levels of the state (particularly in highly industrialized areas with lots of laborconflict) seem to have become conscious of the “reverse J” problem described by EricOlin Wright (2000). That is to say, the interests of capital are, in certain regions andsectors, being negatively affected by the lack of associational power for Chinese workers.An increase in organized worker power could have the effect of coordinating collectiveaction on the part of capital, thereby leading to rationalization and efficiency gains. Inother words, “[capital] must find ways to have its own needs put forward by its enemies.”(Tronti 1971)246 Nowhere is such a situation clearer than in the small commodity sectorsin Southeast Zhejiang.

245 The state is never “disinterested” in the strong sense of the term; I merely mean that inan instance in which labor and capital have agreed to maintenance of the general structureof capitalism, the state does not strongly favor either side.246 I have cited the original Italian version of the book where this passage appears. I drewthis quote from an unofficial English translation of “The Strategy of Refusal” which canbe found at:

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (167)


The failure to secure substantive compromise in China has meant that workerinsurgency continues. It is likely that labor conflict will not continue to grow at the samefantastic pace of the first decade of the 21st century – if only because it is already so high.But it appears that insurgency will not decline significantly in the short-mid term, and itis therefore likely that the state and union will continue to experiment with institutionalresponses. Since only the formal outlines of compromise have begun to emerge, whatmight we reasonably expect from the ongoing dialectic between the insurgent andinstitutional moments of the countermovement?

Future Developments

I have postulated that marketization in China has resulted in the development ofan “insurgency trap.” Divergent interests between the central and local state, combinedwith the center’s categorical ban on autonomous worker organization, have resulted in asituation in which worker insurgency resulting from commodification has remained atvery high levels. But if the state is in a “trap” at present, the situation is not static. Howmight things develop in the future?

The first possibility is stalemate. The state maintains its ban on independentunions, workplaces remain in a legal vacuum dependent on highly commodified labor,and worker insurgency remains high but stabilizes. This insurgency remains highlycellular and therefore doesn’t threaten basic social order. As a result, workers do notexact a high enough cost on the state and capital to force reforms, and the basic set-upremains unaltered. High levels of economic and social polarization would persist, andChina would fail to develop significant domestic consumer markets. Over time, such astalemate could sap the strength of economic growth as labor conflicts would continueand capital would not be forced into actively pursuing gains in efficiency. In other words,the state is unable to exit insurgency trap.

The second potential development is reform. In this scenario, segments of thestate that are committed to compromise will promote the development of associationalpower for the working class. The existing union structure would be deeply reformed,and/or new and independent unions would be legalized. Such organizations would haveto be much more democratic than the ACFTU, and would have the potential to confrontstate and capital over issues of economic, and potentially political, concern to theirmembers. These relatively autonomous unions could play a key role in forcingcompromise from their own constituency (something the ACFTU has not been capableof), while simultaneously wielding potentially coercive power vis-à-vis employers in theform of a credible strike threat. This may push the state to force employers to participatein interest-aggregating business associations, thereby increasing the possibility ofregional or national-level collective bargaining. Additionally, it is possible to imaginesuch unions pushing for greater citizenship rights for migrant workers and an expansionof formal democracy,247 although this scenario would not result in social revolution.248

(accessed April 19, 2011) Since it was drawn from the internet, there is no listed pagenumber.247 Working class organizations have played an important role in democratization andredemocratization (Valenzuela 1989) in Europe (Rueschemeyer, Stephens, and Stephens

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (168)


Currently the state is unwilling to make the political compromises necessary forsubstantial reform. The central government remains deeply anxious about socialinstability, and local governments are in most instances too concerned with short-termaccumulation.249 But if worker insurgency exacts a high enough economic and politicaltoll over the mid-long term, this may change and the central government could need thehelp of relatively autonomous – but potentially co-optable – trade unions. Additionally,local governments may come to see that expanded associational power for workers couldhelp with the long-term development and stability of particular industries, as has alreadybeen the case in some sectors in Zhejiang. They will need unions with higher legitimacyamong workers (i.e. incorporation of labor) to enforce rationalizing and efficiency-producing reforms. Additionally, inability to expand domestic consumption may come toseriously hamper China’s economic development. In short, reform implies a Keynesian-style solution to the problem, and may hold the greatest promise for the long-termstability and continuity of capitalist development in China.250

A third scenario is that of expanded radicalism which could result inrevolutionary outcomes. Here, the state at various levels digs in its heels and continues torepress any attempts at substantive union reform, while worker insurgency continues toexpand in scope, intensity, militancy, and organization. Strikes and various forms ofdirect action are directed against employers, but increasingly against the state as well aslabor struggles become explicitly political. In other words, the cell walls separatingsparks of insurgency are eventually worn town, and horizontally organized “tissues” ofresistance begin to take form.251 Worker organization develops outside the legal auspicesof the state, and therefore is constituted in opposition to it.252 I do not wish to make anywildly speculative remarks on what sort of politics might characterize such a movement.But it is possible to imagine a variety of forms of working class radicalism that couldseriously destabilize current power arrangements, potentially leading to either libratoryoutcomes or social breakdown.

1992; Therborn 1977), Asia (Buchanan and Nicholls 2003), and Latin America (Collierand Mahoney 1997; Collier 1999), though it is worth noting this role has beeninconsistent across cases (Bellin 2000).248 As defined in the classic formulation by Skocpol (1979), a very different scenario thansimple “democratization” as conceived of by Huntington (1991) and others.249 This is true both because of so-called “local protectionism” and because officials’ owncareer advancement can depend in large part on success with economic growth; see (Liand Zhou 2005; Zhou 2004).250 It is important to note that even if such a solution could address the immediateeconomic crisis, continual expansion of consumption in China poses an incredibly graveecological threat, (Guan et. al. 2008)251 I want to emphasize that I am not advocating a functionalist “social body” metaphor inthe vein of Durkheim (1997), but rather simply want to indicate that previously(relatively) autarkic social units begin to be fused together, thereby increasing theirstrength.252 Similar to Seidman’s (1994) account of the militancy of South African and Brazilianlabor movements.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (169)


While I do not think this third scenario is particularly likely, as a thought exerciseit does reveal a key characteristic of labor politics in China. Namely, that strikes anddirect action by workers are currently organized, essentially without exception,autonomously from existing union structures, thereby making it very difficult for the stateto co-opt. Depending on one’s perspective, this is either a significant advantage orindicative of fundamental weakness. In concluding I will now turn to a discussion of thisproblem and its relevance to the global political economy.

Autonomous Worker Movements and the Future of Capitalism

Antonio Negri and Michael Hardt have argued that the contemporary era ischaracterized by the development of a, “decentered and deterritorializing apparatus ofrule,” which they term Empire (Hardt and Negri 2000). Drawing on Negri’s autonomistMarxist roots, they contend that the “withering of civil society” (Hardt 1995) is actuallyan advantage for the working class, since we now have, “the set of all the exploited andthe subjugated, a multitude that is directly opposed to Empire, with no mediation betweenthem.” (Hardt and Negri 2000:393) From this perspective, the marginalization of theunions in the West that helped construct the welfare state in the post-war era is to bewelcomed since, “Working-class power resides not in the representative institutions butin the antagonism and autonomy of the workers themselves.” (ibid:269)

Whatever one’s assessment of such a position, the autonomist perspectivecaptures something about contemporary worker insurgency in China. Although it isunclear to what extent he is engaged in analysis vs. exhortation, Mario Tronti holds thatclass consciousness emerges with, “the working class refusal to present demands tocapital, the total rejection of the whole trade union terrain, the refusal to limit the classrelationship within a formal, legal, contractual form.” (Tronti 1971)253 Consciousmaintenance of such a political position does not exist for Chinese workers who ingeneral remain hopeful about the capacity of the state/laws to deliver justice; and yet, ona very practical level a variety of political formations have forced workers into just suchautonomous resistance. As has been well-documented, workers – like other protestors inChina – generally use the official language of the state (O'Brien 1996; O'Brien and Li2006) and attempt to resolve grievances through proper legal channels (Gallagher 2005;Lee 2007), while not infrequently seeking the help of the union (Chen 2004); however, itis the systemic failure of these channels to resolve conflict that has resulted in anexpanding, diffuse and decentered, extra-legal insurgency. A media description from a2004 worker strike and riot in Dongguan could easily be applied to any number ofincidents: “In the chaos the workers had no leaders, no representatives, no organizationand no concrete demands. They dispersed as quickly as they rose up.” (quoted in Chan2011:31) Additionally, lack of substantive representation means that when workers dohave demands they are immediate and particular. Although the state is forced to respondto the political crisis of generalized insurgency, workers themselves do not aggregatedemands at the political or class level.

We have already identified the present systemic risk in the capitalist worldsystem, but how are we to assess the possibilities of autonomous worker insurgency

253 See footnote 246.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (170)


generating a more sustainable and equitable political economy? First of all, it must benoted that the crisis of trade unionism is not particular to China, even if it exists in moreextreme form there. Around the world, trade unions have sustained decades of assaultfrom both state and capital, and the capacity of these organizations to defend, to saynothing of advancing, either the narrow interests of membership or more general socialinterests has become highly circ*mscribed. What’s more, if the response to neoliberalismmust be global, we find that existing international organizations, e.g. the InternationalLabor Organization, International Trade Union Confederation, etc., are eitheruninterested in or woefully unprepared for such a struggle. Given the current overbearingpower of transnational capital, it is highly unlikely that such institutions will be thebirthplace of renewed counter-hegemony.

The question then becomes, how can diffuse and decentered insurgency have aglobal impact? With the Eurozone in crisis and ultra-nationalism on the rise from the U.S.(Blee and Creasap 2010) to Europe (Mudde 2007) and East Asia (f*ckuyama 2007) andmany places in between, it is increasingly clear that theories of the “eclipse” of the statewere exceedingly premature (Evans 1997). National institutions, laws, culture, etc., stillheavily structure movements, reinforcing the position that efficacious transnationalismmust be built on the foundation of robust local/national movements. Since traditionaltrade unions in most countries are heavily integrated into existing nationally-basedstate/party structures, it is unlikely that they will – of their own accord – fight to constructa counter-hegemonic foundation at the national level. This is particularly the casebecause such a project could, and probably would, threaten their integration into existingpower arrangements. This is not to say that such institutions should or realistically can bewholly circumnavigated; even in the case of China where the ACFTU is a non-factor inmost workers’ lives, they are time and again forced to confront official unions. In PeterEvans’s view, counter-hegemony can be constructed through some amalgam of “trees” –traditional organizations like trade unions – and “rhizomes” – the diffuse insurgentscelebrated by Hardt and Negri (Evans 2008:291). My argument is that in many cases,typical “trees” are decayed, and without being forced into mobilization by insurgent“rhizomic” activity, they will continue to rot in irrelevance. In short, the insurgent-institutional dialectic can be the motor force behind the establishment of counter-hegemonic blocs capable of confronting capital at the transnational level.

It goes without saying that any countermovement against neoliberal globalizationcannot succeed without a strong presence in China. China’s increasing centrality in theworld economy means that any domestic shift in economic and political organization willhave immediate ripple effects throughout the globe. I have argued throughout thatChinese workers are already rejecting commodification en masse and are beginning toforce major concessions from individual employers. Unions remain reactive to workerresistance and highly illegitimate amongst their constituency; thus the countermovementin China remains stalled at the insurgent moment of the countermovement. But thesituation is fluid, and China continues to be the world’s epicenter of labor unrest (Silverand Zhang 2009). Whether this diffuse insurgency will be capable of exercising greaterpolitical power is unclear, but no global countermovement will succeed without it.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (171)



1958. Zhongguo gonghui lishi wenxian [Historical Documents of Chinese Trade Unions].Beijing: Gongren chubanshe.

1980. shenggang da bagong ziliao [Materials from the Guangdong-Hong Kong General

Strike]. Guangzhou: Guangdong renmin chubanshe.1988. jianguo yilai zhonggong zhongyang guanyu gongren yundong wenjian xuanbian

[CCP Central Committee Documents on the Workers Movement Since the

Establishment of the Nation], Edited by A. C. F. o. T. Unions. Beijing: gongrenchubanshe.

Andrews, Kenneth T. 2004. Freedom is a Constant Struggle: The Mississippi Civil Rights

Movement and its Legacy. Chicago: University of Chicago Press.Arrighi, Giovanni. 2007. Adam Smith in Beijing: Lineages of the Twenty-First Century.

New York: Verso.Arrighi, Giovanni and Beverly J. Silver. 1999. Chaos and Governance in the Modern

World System. Minneapolis: University of Minnesota Press.Averill, Stephen C. . 1987. "Party, Society, and Local Elite in the Jiangxi Communist

Movement." The Journal of Asian Studies 46:279-303.Bellin, Eva. 2000. "Contingent Democrats: Industrialists, Labor, and Democratization in

Late-Developing Countries." World Politics 52:175-205.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (172)


Blecher, Marc. 2002. "Hegemony and Workers' Politics in China." China Quarterly

170:283-303.Blee, Kathleen and Kimberly Creasap. 2010. "Conservative and Right-Wing

Movements." Annual Review of Sociology 36:269-286.Block, Fred. 1977. "The Ruling Class Does not Rule." Social Revolution 33:6-28.—. 2003. "Karl Polanyi and the Writing of The Great Transformation." Theory and

Society 32:275-306.Bianco, Lucien and Muriel Bell. 1971. Origins of the Chinese Revolution, 1915-1949.

Stanford: Stanford University Press.Bickford, Thomas. 1994. "The Chinese Military and Its Business Operations: The PLA as

Entrepreneur." Asian Survey 34:460-474.Bourdieu, Pierre. 1984. Distinction: A Social Critique of the Judgment of Taste.

Cambridge: Harvard University Press.—. 1985. "The Social Space and the Genesis of Groups." Theory and Society 14:723-744.—. 1989. "Social Space and Symbolic Power." Sociological Theory 7:14-25.—. 1998. Acts of Resistance: Against the Tyranny of the Market. New York: The New

Press.—. 2003. "Political Representation: Elements for a Theory of the Political Field." in

Language and Symbolic Power. Cambridge: Harvard University Press.—. 2005. "The Mystery of the Ministry: From Particular Wills to the General Will." in

Pierre Bourdieu and Democratic Politics, edited by L. Wacquant. Cambridge:Polity Press.

Bramall, Chris. 1989. "The Wenzhou 'Miracle': an Assessment." Pp. 43-76 in Market

Forces in China: Competition and Small Business - The Wenzhou Debate, editedby P. Nolan and F. Dong. London: Zed Books.

Broadman, Harry G. and Xiaolun Sun. 1997. "The Distribution of Foreign DirectInvestment in China." The World Bank.

Buchanan, Paul G. and Kate Nicholls. 2003. "Labour Politics and Democratic Transitionin South Korea and Taiwan." Government and Opposition 38:203-237.

Burawoy, Michael. 1982. Manufacturing Consent: Changes in Labor Process Under

Monopoly Capitalism. Chicago: University of Chicago Press.—. 2003. "For a Sociological Marxism: The Complementary Convergence of Antonio

Gramsci and Karl Polanyi." Politics & Society 31:193-261.Cai, Yongshun. 2002. "The Resistance of Chinese Laid-off Workers in the Reform

Period." The China Quarterly 170:327-344.—. 2010. Collective Resistance in China: Why Popular Protests Succeed or Fail.

Stanford: Stanford University Press.Caraway, Teri L. 2009. "Labor Rights in East Asia: Progress or Regress?" Journal of

East Asian Studies 9:153-186.Chan, Anita. 1993. "Revolution or Corporatism? Workers and Trade Unions in Post-Mao

China." The Australian Journal of Chinese Affairs 29:31-61.—. 2001. China's Workers Under Assault: The Exploitation of Labor in a Globalizing

Economy. Armonk, N.Y.: M.E. Sharpe.—. 2007. "Organizing Wal-Mart in China: Two Steps Forward, One Step Back for

China's Unions." New Labor Forum 16:87-96.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (173)


—. 2008. "Challenges and Possibilities for Democratic Grassroots Union Elections inChina: A Case Study of Two Factory-Level Elections and Their Aftermath."Labor Studies Journal 34:293-317.

—. 2011. "Strikes in China's Export Industries in Comparative Perspective." The China

Journal 65:27-51.Chan, Anita and Jonathan Unger. 2009. "A Chinese State Enterprise Under the Reforms:

What Model of Capitalism?" The China Journal 62:1-26.Chan, Chris King-Chi. 2010. The Challenge of Labour in China: Strikes and the

Changing Labour Regime in Global Factories. London: Routledge.Chan, Chris King-Chi and Ngai Pun. 2009. "The Making of a New Working Class? A

Study of Collective Action of Migrant Workers in South China." China Quarterly

198:287-303.Chan, Jenny Wai-ling. 2009. "Meaningful Progress or Illusory Reform?: Analyzing

China's Labor Contract Law." New Labor Forum 18:43-51.Chen, Calvin. 2008. Some Assembly Required: Work, Community, and Politics in China's

Rural Enterprises. Cambridge: Harvard University Press.Chen, Chung, Lawrence Chang, and Yimin Zhang. 1995. "The Role of Foreign Direct

Investment in China's Post-1978 Economic Development." World Development

23:691-703.Chen, Ding and Junyong Huang. 2008. "zhejiang wenlingshi tuixing hangye gongzi jiti

xieshang zhidu de jingyan yu qishi (The experience and inspiration of Wenling'simplementation of a collective wage negotiation system)." Tianjinshi gonghui

guanli ganbu xueyuan xuebao 16:30-32.Chen, Feng. 2000. "Subsistence Crises, Managerial Corruption and Labour Protests in

China." The China Journal 44:41-63.—. 2003. "Industrial Restructuring and Workers' Resistance in China." Modern China

29:237-262.—. 2004. "Legal Mobilization by Trade Unions: The Case of Shanghai." The China

Journal 52:27-45.—. 2007. "Individual Rights and Collective Rights: Labor's Predicament in China."

Communist and Post-Communist Studies 40:59-79.—. 2009. "Union Power in China: Source, Operation, and Constraints." Modern China

35:662-689.Chen, Shengjun. 2009. "dui "laodong hetong fa" shishi guocheng zhong ruogan wenti de

sikao [On some issues in the implementation of the Labor Contract Law]." hunan

sheng gonghui ganbu xuexiao 23:30-32.Chesneux, Jean. 1968. The Chinese Labor Movement, 1919-1927. Stanford: Stanford

University Press.Chibber, Vivek. 2003. Locked in Place: State-building and Late Industrialization in

India. Princeton: Princeton University Press.Chin, Christine and James Mittelman. 1997. "Conceptualising Resistance to

Globalisation." New Political Economy 2:25-37.Choi, Young-Jin. 2003. "Managerial Styles, Workforce Composition and Labor Unrest:

East Asian-Invested Enterprises in China." Comparative Sociology 2:321-354.Clarke, Simon, Chang-Hee Lee, and Qi Li. 2004. "Collective Consultation and Industrial

Relations in China." British Journal of Industrial Relations 42:235-254.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (174)


Cohen, Youssef. 1982. “‘The Benevolent Leviathan’: Political Consciousness amongUrban Workers under State Corporatism.” The American Political Science Review

76:46-59.Collier, Ruth. 1999. Paths Toward Democracy: The Working Class and Elites in Western

Europe and South America. New York: Cambridge University Press.Collier, Ruth and David Collier. 1991. Shaping the Political Arena: Critical Junctures,

The Labor Movement, and Regime Dynamics in Latin America. Princeton:Princeton University Press.

Collier, Ruth and James Mahoney. 1997. "Adding Collective Actors to CollectiveOutcomes: Labor and Recent Democratization in South America and SouthernEurope." Comparative Politics 29:285-303.

Cooney, Sean. 2006. "Making Chinese Labor Law Work: The Prospects for RegulatoryInnovation in the People's Republic of China." Fordham International Law

Journal 30:1050-1097.Cooney, Sean, Sarah Biddulph, Kungang Li, and Ying Zhu. 2007. "China's New Labor

Contract Law: Responding to the Growing Complexity of Labour Relations in thePRC." University of New South Wales Law Journal 30:786-801.

David-Friedman, Ellen. 2009. "U.S. and Chinese Labor at a Changing Moment in theGlobal Neoliberal Economy." Working USA 12:219-34.

Deyo, Frederic C. 1987. "State and Labor: Modes of Political Exclusion in East AsianDevelopment." Pp. 182-202 in The Political Economy of the New Asian

Industrialism edited by F. C. Deyo. Ithaca: Cornell University Press.—. 1989. Beneath the Miracle: Labor Subordination in the New Asian Industrialism.

Berkeley: University of California Press.Dong, Fureng. 1989. "The Wenzhou Model for Developing the Rural Commodity

Economy." Pp. 77-96 in Market Forces in China: Competition and Small

Business - The Wenzhou Debate, edited by F. Dong and P. Nolan. London: ZedBooks.

Duckett, Jane. 2001. "Bureaucrats in Business, Chinese-Style: The Lessons of MarketReform and State Entrepreneurialism in the People's Republic of China." World

Development 29:23-37.Durkheim, Emile. 1997. The Division of Labor in Society. New York: Free Press.Edelstein, J. David. 1967. "An Organizational Theory of Union Democracy." American

Sociological Review 32:19-31.Edelstein, J. David and Malcolm Warner. 1976. Comparative Union Democracy.

London: George Allen & Unwin.Eng, Irene. 1997. "The Rise of Manufacturing Towns: Externally Driven Industrialization

and Urban Development in the Pearl River Delta of China." International Journal

of Urban and Regional Research 21:554-568.Esping-Andersen, Gosta. 1990. The Three Worlds of Welfare Capitalism. Princeton:

Princeton University Press.Evans, Peter. 1995. Embedded Autonomy: States and Industrial Transformation.

Princeton: Princeton University Press.—. 1997. "The Eclipse of the State? Reflections on Stateness in an Era of Globalization."

World Politics 50.1:62-87.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (175)


—. 2000. "Fighting Marginalization with Transnational Networks: Counter-HegemonicGlobalization." Contemporary Sociology 29:230-241.

—. 2005. "Counter-Hegemonic Globalization: Transnational Social Movements in theContemporary Global Political Economy." Pp. 655-668 in Handbook of Political

Sociology, edited by T. Janoski, A. M. Hicks, and M. Schwartz. CambridgeUniversity Press: Cambridge.

—. 2008. "Is an Alternative Globalization Possible?" Politics & Society 36:271-305.

Fewsmith, Joseph. 2005. "Chambers of Commerce in Wenzhou Show Potential andlimits of 'Civil Society' in China." China Leadership Monitor 16:1-9.

Forster, Keith. 1990. "The Wenzhou Model for Economic Development: Impressions."China Information 5:53-64.

Forster, Keith and Xianguo Yao. 1999. "A Comparative Analysis of Economic Reformand Development in Hangzhou and Wenzhou Cities." Pp. 51-102 in Cities in

China: Recipes for Economic Development in the Reform Era, edited by J. H.Chung. New York: Routledge.

Frazier, Mark. 2010. Socialist Insecurity: Pensions and the Politics of Uneven

Development in China. Ithaca: Cornell University Press.Friedman, Eli and Ching Kwan Lee. 2010. "Remaking the World of Chinese Labour: a

30-Year Retrospective." British Journal of Industrial Relations 48:507-533.f*ckuyama, Francis. 2007. "The New Nationalism and the Stategic Architecture of

Northeast Asia." Asia Policy 3:38-41.Gallagher, Mary. 2002. "Reform and Openness: Why China's Economic Reforms Have

Delayed Democracy." World Politics 54:338-372.—. 2004. "Time is Money, Efficiency is Life: The Transformation of Labor Relations in

China." Studies in Comparative International Development 39:11-44.—. 2005(a). Contagious Capitalism: Globalization and the Politics of Labor in China.

Princeton: Princeton University Press.—. 2005(b). "Use the Law as Your Weapon: Institutional Change and Legal Mobilization

in China." in Engaging the Law in China: State, Society, and Possibilities for

Justice, edited by N. Diamant, K. O'Brien, and S. Lubman. Stanford: StanfordUniversity Press.

Giugni, Marco. 1998. "Was it Worth the Effort? The Outcomes and Consequences ofSocial Movements." Annual Review of Sociology 98:371-393.

Giugni, Marco, Doug McAdam, and Charles Tilly. 1999. How Social Movements Matter.Minneapolis: University of Minnesota Press.

Gold, Thomas. 1990. "Tiananmen and Beyond, the Resurgence of Civil Society inChina." Journal of Democracy 1:18-31.

Goldfield, Michael. 1989. "Worker Insurgency, Radical Organization, and New DealLabor Legislation." The American Political Science Review 83:1257-1282.

Goldman, Rene. 1962. "The Rectification Campaign at Peking University: May-June1957." The China Quarterly 12:138-153.

Gramsci, Antonio. 1971. Selections from the Prison Notebooks. New York: InternationalPublishers.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (176)


Greene, Anne-Marie, John Hogan, and Margaret Grieco. 2003. "E-collectivism andDistributed Discourse: New Opportunities for Trade Union Democracy."Industrial Relations Journal 34.

Guan, Dabo, Klaus Hubacek, Christopher L. Weber, Glen P. Peters, and David M.Reiner. 2008. "The Drivers of Chinese C02 Emissions from 1980 to 2030."Global Enviromental Change 18:626-634.

Guha, Ranajit. 1983. Elementary Aspects of Peasant Insurgency in Colonial India. Delhi:Oxford University Press.

Guo, Yingjie. 2009. "Farewell to Class, except the Middle Class: The Politics of ClassAnalysis in Contemporary China." The Asia-Pacific Journal 26.

Hardt, Michael. 1995. "The Withering of Civil Society." Social Text 45:27-44.Hardt, Michael and Antonio Negri. 2000. Empire. Cambridge, Mass.: Harvard University

Press.—. 2004. Multitude. New York: Penguin Books.Harper, Paul. 1969. "The Party and Unions in Communist China." The China Quarterly

37:84-119.Hart-Landsberg and Paul Burkett. 2005. China and Socialism: Market Reforms and Class

Struggle. New York: Monthly Review Press.Hassid, Jonathan. 2008. "Controlling the Media: An Uncertain Business." Asian Survey

48:414-430.Hinton, William. 1966. Fanshen: A Documentary of Revolution in a Chinese Village.

New York: Monthly Review Press.Hishida, Masaharu, Kazuko Kojima, Tomoaki Ishii, and Jian Qiao. 2010. China's Trade

Unions: How Autonomous are They? . New York: Routledge.Howell, Jude. 2008. "All-China Federation of Trade Unions beyond Reform? The Slow

March of Direct Elections." China Quarterly 196:845-863.Hsing, You-tien. 2009. The Great Urban Transformation: Politics of Land and Property

in China. Oxford: Oxford University Press.Hung, Ho-fung. 2008. "Rise of China and the Global Overaccumulation Crisis." Review

of International Political Economy 15:149-179.—. 2009. China and the transformation of global capitalism. Baltimore: Johns Hopkins

University Press.Huntington, Samuel P. . 1991. "Democracy's Third Wave." Journal of Democracy 2:12-

34.Hurst, William. 2004. "Understanding Contentious Collective Action by Chinese Laid-

Off Workers: The Importance of Regional Political Economy." Studies in

Comparative International Development 39:94-120.—. 2009. The Chinese Worker after Socialism. Cambridge: Cambridge University Press.Jacobs, Paul. 1963. The State of the Unions. New York: Atheneum.Jin, Hehui, Yingyi Qian, and Barry R. Weingast. 2005. "Regional Decentralization and

Fiscal Incentives: Federalism, Chinese Style." Journal of Public Economics

89:1719-1742.Katznelson, Ira. 1986. "Working-Class Formation: Constructing Cases and

Comparisons." in Working-Class Formation: Nineteenth-Century Patterns in

Western Europe and the United States, edited by I. Katznelson and A. Zolberg.Princeton: Princeton University Press.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (177)


Keck, Margaret and Kathryn Sikkink. 1998. Activists Beyond Borders: Advocacy

Networks in International Politics: Cornell University Press.Keith, Ronald C. 1991. "Chinese Politics and the New Theory of "Rule of Law"." The

China Quarterly 125:109-118.—. 1994. China's Struggle for the Rule of Law. New York: Macmillan.Kohli, Atul. 2004. State-Directed Development: Political Power and Industrialization in

the Global Periphery. Cambridge: Cambridge University Press.Koo, Hagen. 2001. Korean Workers: The Culture and Politics of Class Formation.

Ithaca: Cornell University Press.Kornhauser, William. 1959. "The Structure of Mass Society." in The Politics of Mass

Society. New York: Free Press.Kwan, Daniel Y.K. 1997. Marxist Intellectuals and the Chinese Labor Movement: A

Study of Deng Zhongxia, 1894-1933. Seattle: University of Washington Press.Lau, Raymond W.K. 2001. "Socio-political Control in Urban China: Changes and

Crisis." British Journal of Sociology 52:605-620.—. 2003. "The Habitus and 'Logic of Practice' of China's Trade Unionists." Issues &

Studies 39:75-103.Lee, Ching Kwan. 1995. "Engendering the Worlds of Labor: Women Workers, Labor

Markets, and Production Politics in the South China Economic Miracle."American Sociological Review 60.

—. 1998. Gender and the South China Miracle: Two Worlds of Factory Women.Berkeley: University of California Press.

—. 2002. "From the Specter of Mao to the Spirit of the Law: Labor Insurgency in China."Theory and Society 31:189-228.

—. 2007. Against the Law: Labor Protests in China's Rustbelt and Sunbelt. Berkeley:University of California Press.

Lee, Lai To. 1984. The Structure of the Trade Union System in China, 1949-1966. HongKong: Hong Kong University Press.

—. 1986. Trade Unions in China: 1949 to the Present. Singapore: Singapore UniversityPress.

Leung, Pak Nang and Ngai Pun. 2009. "The Radicalisation of the New Chinese WorkingClass: a case study of collective action in the gemstone industry." Third World

Quarterly 30:551-565.Levi, Margaret, David Olson, Jon Agnone, and Devin Kelly. 2009. "Union Democracy

Reexamined." Politics & Society 37:203-228.Li, Guofang. 2001. "tongyi zhanxian zhengce zai shenggang bagong zhong de zuoyong

[The Function of the United Front Policy during the Shenggang Strike]." xingtai

shifan gaozhuan xuebao [Xingtai Normal Institute Journal] 16.Li, Hongbin and Li-An Zhou. 2005. "Political Turnover and Economic Performance: The

Incentive Role of Personnel Control in China." Journal of Public Economics

89:1743-1762.Li, Minqi. 2005. "The Rise of China and the Demise of the Capitalist World-Economy:

Exploring Historical Possibilities in the 21st century." Science & Society 69:420-448

—. 2007. "Peak Oil, the Rise of China and India, and the Global Energy Crisis." Journal

of Contemporary Asia 37:449-471.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (178)


—. 2009. The Rise of China and the Demise of the Capitalist World Economy. NewYork: Monthly Review Press.

Li, Sishen and Zhikun Liu. 2005. Li Lisan zhi mi: yi ge zhongcheng gemingzhe de quzhe

rensheng [The mystery of Li Lisan: The winding life of a loyal revolutionary].Beijing: renmin chubanshe.

Li, Zheng and Ji Chen. 1988. "quanmian shenhua gaige zhong de gonghui he gonghui degaige [The unions in deepening reforms and union reform]." Beijing: GongrenChubanshe.

Lipset, Seymour Martin, Martin Trow, and James Coleman. 1956. Union Democracy.New York: Free Press.

Liu, Alan P. 1992a. "The "Wenzhou Model" of Development and China'sModernization." Asian Survey 32:696-711.

Liu, Mingwei. 2009. "Chinese Employment Relations and Trade Unions in Transition[Dissertation]." Cornell University.

—. 2009. “Bottom-up Change: The Future of Chinese Trade Unions?”Liu, Yia-Ling. 1992b. "Reform from Below: The Private Economy and Local Politics in

the Rural Industrialization of Wenzhou." The China Quarterly 130:293-316.Lubman, Stanley. 1999. Bird in a Cage: Legal Reform in China after Mao. Stanford:

Stanford University Press.Luce, Stephanie and Edna Bonacich. 2009. "China and the U.S. Labor Movement." Pp.

153-172 in China and the Transformation of Global Capitalism, edited by H.-f.Hung. Baltimore: Johns Hopkins University Press.

Ma, Zhining. 2011. The ACFTU and Chinese Industrial Relations. New York: PeterLang.Magrath, C. Peter. 1958. "Democracy in Overalls: The Futile Quest forUnion Democracy." Industrial and Labor Relations Review 12:503-525.

Marx, Karl. 1977. Capital: A Critique of Political Economy, vol. 1. New York: VintageBooks.

—. 1978. "The Communist Manifesto." in The Marx-Engels Readers, edited by R. C.Tucker. New York: Norton.

MacFarquhar, Roderick. 1960. The Hundred Flowers Campaign and the Chinese

Intellectuals. New York: Praeger.McAdam, Doug. 1982. Political Process and the Development of Black Insurgency,

1930-1970. Chicago: University of Chicago Press.McAdam, Doug, John D. McCarthy, and Mayer N. Zald. 1996. Comparative

Perspectives on Social Movements: Political Opportunities, Mobilizing

Structures, and Cultural Framings: Cambridge University Press.Michels, Robert. 1962. Political Parties: A Sociological Study of the Oligarchical

Tendencies of Modern Democracy. New York: Dover.Milkman, Ruth. 2006. L.A. Story: Immigrant Workers and the Future of the U.S. Labor

Movement. New York: Russel Sage FoundationMontinola, Gabriella, Yingyi Qian, and Barry R. Weingast. 1996. "Federalism, Chinese

Style: The Political Basis for Economic Success." World Politics 48:50-81.Mudde, Cas. 2007. Populist Radical Right Parties in Europe. Cambridge: Cambridge

University Press.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (179)


Offe, Claus and Helmut Wiesenthal. 1980. "Two Logics of Collective Action:Theoretical Notes on Social Class and Organizational Form." Political Power and

Social Theory 1:67-115.O'Brien, Kevin. 1996. "Rightful Resistance." World Politics 49.1:31-55.O'Brien, Kevin and Lianjiang Li. 2006. Rightful Resistance in Rural China. Cambridge:

Cambridge University Press.—. 1999. "Selective Policy Implementation in Rural China." Comparative Politics

31:167-186.Parris, Kristen. 1993. "Local Initiative and National Reforms: The Wenzhou Model of

Development." The China Quarterly 134:242-263.Parsons, Talcott. 1964. "Evolutionary Universals in Society." American Sociological

Review 29:339-357.Pearson, Margaret 1994. "The Janus Face of Business Associations in China: Socialist

Corporatism in Foreign Enterprises." The Australian Journal of Chinese Affairs

31:25-46.Pei, Minxin. 2006. China's Trapped Transition: The Limits of Developmental Autocracy.

Cambridge, Mass.: Harvard University Press.Perry, Elizabeth. 1993. Shanghai on Strike: The Politics of Chinese Labor. Stanford:

Stanford University Press.—. 1994. "Shanghai's Strike Wave of 1957." The China Quarterly 137:1-27.Piven, Frances Fox and Richard Cloward. 1977. Poor People's Movements: Why They

Succeed, How They Fail. New York: Vintage.Polanyi, Karl. 1944. The Great Transformation: The Political and Economic Origins of

Our Time. Boston: Beacon Press.Poulantzas, Nicos. 1973. Political Power and Social Classes. London: NLB.Przeworski, Adam. 1980. "Social Democracy as a Historical Phenomenon." New Left

Review 122:28-58.—. 1985. Capitalism and Social Democracy. Cambridge: Cambridge University Press.Pun, Ngai. 2005. Made in China: Women Factory Workers in a Global Workplace.

Durham: Duke University Press.Pun, Ngai, Chris King-Chi Chan, and Jenny Chan. 2009. "The Role of the State, Labour

Policy and Migrant Workers' Struggles in Globalized China." Global Labour

Journal 1:132-151.Pun, Ngai and Huilin Lu. 2010. "A Culture of Violence: The Labor Subcontracting

System and Collective Action by Construction Workers in Post-Socialist China."The China Journal 64.

Roomkin, Myron. 1976. "Union Structure, Internal Control, and Strike Activity."Industrial and Labor Relations Review 29:198-217.

Rueschemeyer, Dietrich, Evelyne Huber Stephens, and John D. Stephens. 1992.Capitalist Development and Democracy. Chicago: University of Chicago Press.

Sarti, Roland. 1971. Fascism and the Industrial Leadership in Italy, 1919-1940.Berkeley: University of California Press.

—.1974. The Ax Within: Italian Fascism in Action. New York: New Viewpoints.Schmitter, Philippe C. 1974. “Still the Century of Corporatism?” The Review of Politics


Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (180)


Seidman, Gay. 1994. Manufacturing Militance: Workers' Movements in Brazil and South

Africa, 1970-1985. Berkeley: University of California Press.Sheehan, Jackie. 1998. Chinese Workers: A New History. London: Routledge.Silver, Beverly J. 2003. Forces of Labor: Workers' Movements and Globalization Since

1870. Cambridge: Cambridge University Press.Silver, Beverly J. and Giovanni Arrighi. 2003. "Polanyi's "Double Movement": The Belle

Epoques of British and U.S. Hegemony Compared." Politics & Society 31:325-355.

Silver, Beverly J. and Lu Zhang. 2009. "China: Emerging Epicenter of World LaborUnrest? ." in China and the Transformation of Global Capitalism, edited by H.-f.Hung. Baltimore: Johns Hopkins University Press.

Skocpol, Theda. 1979. States and Social Revolution: A Comparative Analysis of France,

Russia, and China. Cambridge: Cambridge University Press.Smith, S.A. 2000. A Road is Made: Communism in Shanghai, 1920-1927. Honolulu:

University of Hawai'i Press.—. 2002. Like Cattle and Horses: Nationalism and Labor in Shanghai, 1895-1927.

Durham: Duke University Press.Snow, David and Daniel Cress. 2000. "The Outcomes of Homeless Mobilization: The

Influence of Organization, Disruption, Political Mediation, and Framing."American Journal of Sociology 105:1063-1104.

Solinger, Dorothy J. 1995. "The Chinese Work Unit and Transient Labor in theTransition from Socialism." Modern China 21:155-183.

—. 1999. Contesting citizenship in urban China peasant migrants, the state, and the logic

of the market. Berkeley: University of California Press.—. 2001. "Why We Cannot Count the 'Unemployed'." The China Quarterly 167:671-688.Soule, Sarah A., Doug McAdam, John McCarthy, and Yang Su. 1999. "Protest Events:

Cause or Consequence of State Action? The U.S. Women's Movement andFederal Congressional Activities, 1956-1979." Mobilization 4:239-256.

Stepan-Norris, Judith. 1997. "The Making of Union Democracy." Social Forces 76:475-510.

Stepan-Norris, Judith and Maurice Zeitlin. 1991. "'Red' Unions and 'Bourgeois'Contracts?" American Journal of Sociology 96:1151-1200.

—. 1995. "Union Democracy, Radical Leadership, and the Hegemony of Capital."American Sociological Review 60:829-850.

—. 1996. "Insurgency, Radicalism, and Democracy in America's Industrial Unions."Social Forces 75:1-32.

—. 2002. Left Out: Reds and America's Industrial Unions. New York: CambridgeUniversity Press.

Stranahan, Patricia. 1998. Underground: The Shanghai Communist Party and the Politics

of Survival, 1927-1937. New York: Rowman & Littlefield.Stratton, Kay. 1989. "Union Democracy in the International Typographical Union: Thirty

Years Later." Journal of Labor Research 10:119-134.Streeck, Wolfgang. 1991. "Interest Heterogeneity and Organising Capacity: Two Class

Logics of Collective Action?" Pp. 161-198 in Political Choice: Institutions, Rules

and the Limits of Rationality, edited by R. Czada and A. Windhoff-Heritier.Boulder, CO: Westview.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (181)


Sugihara, Kaoru. 2003. "The East Asian Path of Economic Development: A Long-termPerspective " Pp. 78-122 in The Resurgence of East Asia: 500, 150 and 50 Year

Perspectives, edited by G. Arrighi, T. Hamash*ta, and M. Selden. New York:RoutledgeCurzon.

Tang, Chunliang. 1989. Li Lisan zhuan [Biography of Li Lisan]. Ha'erbin: HeilongjiangRenmin Chubanshe.

Tarrow, Sidney. 1998. Power in Movement: Social Movements and Contentious Politics.Cambridge: Cambridge University Press.

Therborn, Goran. 1977. "The Rule of Capital and the Rise of Democracy." New Left

Review 103:3-41.Thireau, Isabelle and Linshan Hua. 2003. "The Moral Universe of Aggrieved Chinese

Workers: Workers' Appeals to Arbitration Committees and Letters and VisitsOffices." The China Journal 50:83-103.

Traxler, Franz. 1993. "Business Associations and Labor Unions in Comparison:Theoretical Perspective and Empirical Findings on Social Class, CollectiveAction and Associational Organizability." The British Journal of Sociology

44:673-691.Tronti, Mario. 1971. Operai e Capitale. Turin: Einaudi.Tsai, Kellee. 2002. Back-Alley Banking. Ithaca: Cornell University Press.Tseng, Wanda and Harm Zebregs. 2003. "Foreign Direct Investment in China: Some

Lessons for Other Countries." Pp. 68-88 in China: Competing in the Global

Economy. Washington, D.C.: International Monetary Fund.Unger, Jonathan and Anita Chan. 1995. "China, Corporatism, and the East Asian Model."

The Australian Journal of Chinese Affairs 33:29-53.Valenzuela, J. Samuel. 1989. "Labor Movements in Transitions to Democracy: A

Framework for Analysis." Comparative Politics 21:445-472.Voss, Kim and Rachel Sherman. 2000. "Breaking the Iron Law of Oligarchy: Union

Revitalization in the American Labor Movement." American Journal of Sociology

106:303-349.Walder, Andrew G. 1983. "Organized Dependency and Cultures of Authority in Chinese

Industry." The Journal of Asian Studies 43:51-76.—. 1984. Communist Neo-Traditionalism: Work and Authority in Chinese Industry.

Berkeley: University of California Press.—. 1991. "Workers, Managers and the State: The Reform Era and the Political Crisis of

1989." The China Quarterly 127:467-492.Walder, Andrew G. and Xiaoxia Gong. 1993. "Workers in the Tiananmen Protests: The

Politics of the Beijing Workers' Autonomous Federation." The Australian Journal

of Chinese Affairs 29:1-29.Wang, Fei-Ling. 2010. “Conflict, Resistance and the Transformation of the hukou

system.”Pp. 80-100 in Chinese Society: Change, Conflict, and Resistance, editedby Elizabeth J. Perry and Mark Selden. New York: Routledge.

Wang, Haiyan, Richard P. Appelbaum, Francesca Degiuli, and Nelson Lichtenstein.2009. "China's New Labour Contract Law: is China moving towards increasedpower for workers?" Third World Quarterly 30:485-501.

Wang, Hui. 2006. "Depoliticized Politics, From East to West." New Left Review 41:29-45.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (182)


Wang, Hong. 1990. "zhongguo gonghui zhongyao wenjian xuanbian [selection ofimportant Chinese union documents]." Beijing: Jixie Gongye Chubanshe.

Wang, Jian. 2002. "gonghui gongzuo zhongyao wenjian xuanbian [Important trade unionwork document collection]." Beijing: zhongguo jinrong chubanshe.

Wang, Shaoguang. 2008. "The Great Transformation: The Double Movement in China."boundary 2 35:15-47.

Wang, Shizong. 2004. "hangye zuzhi de cunzai jichu he quanli laiyuan: dui wenzhoushanghui de shehui hefaxing kaocha (The Origin of Power and Foundation ofExistence for Industrial Organizations: An Investigation of the Social Legitimacyof Wenzhou Employer Associations)." zhonggong zhejiang sheng weidangxiao

xuebao 2.Wang, Yongxi, Bufeng He, and Yanping Cao. 2005. jianming zhongguo gonghui shi.

Beijing: zhongguo gongren chubanshe.Warner, Malcolm. 1995. The Management of Human Resources in Chinese Industry.

Basingstoke: Macmillian.Weber, Max. 1946. "Bureaucracy." in From Max Weber: Essays in Sociology, edited by

H. H. Gerth and C. W. Mills. New York: Oxford University Press.—. 1978. Economy and Society. Berkeley: University of California Press.Wei, Yehua Dennis, Wangming Li, and Chunbin Wang. 2007. "Restructuring Industrial

Districts, Scaling Up Regional Development: A Study of the Wenzhou Model,China." Economic Geography 83:421-444.

Weil, Robert. 1996. Red Cat, White Cat: China and the Contradictions of 'Market

Socialism'. New York: Monthly Review Press.Wright, Erik Olin. 2000. "Working-Class Power, Capitalist-Class Interests, and Class

Compromise." American Journal of Sociology 105:957-1002.Wright, Teresa. 2010. Accepting Authoritarianism: State-society Relations in China's

Reform Era. Stanford: Stanford University Press.Wu, David Yen-ho. 1991. "The Construction of Chinese and Non-Chinese Identities."

Daedalus 120:159-179.Wu, Ganying. 2004. "wenzhou shanghui zuoyong tuxiang: qiangda minjian liliang zhutui

jingji fazhan (Promotion of the Function of Wenzhou Employer Associations:Strong Civil Power Promotes Economic Development)." qiye jingji 11:152-153.

Wu, Tien-Wei. 1968. "Chiang Kai-shek's March Twentieth Coup d'Etat of 1926." The

Journal of Asian Studies 27:585-602.Ye, Xinyue and Yehua Dennis Wei. 2005. "Geospatial Analysis of Regional

Development in China: The Case of Zhejiang Province and the Wenzhou Model."Eurasian Geography and Economics 46:445-464.

Yu, Jianxing, Yueqian Xu, and Hua Jiang. 2007. "wenzhou shanghui de liwai yu buliwai: zhongguo gongmin shehui de fazhan yu tiaozhan (The Exceptionalness andUnexceptionalness of Wenzhou Employer Associations: The Development andChallenges of Civil Society in China." zhejiang daxue xuebao 37.

Yu, Jing. 2009. "lun shixisheng laodong baozhang de zerenren ji xiangguan zeren [On theobligations to Labor and Social Security of Intern]." Zhongguo Laodong Guanxi

Xueyuan Xuebao 23:98-101.Zhang, Li and Aihwa Ong. 2008. Privatizing China : socialism from afar. Ithaca: Cornell

University Press.

Rupture and Representation: Migrant Workers, Unions and ...· 2/12/2011 · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (183)


Zhang, Lu. 2008. "Lean Production and Labor Controls in the Chinese AutomobileIndustry in An Age of Globalization." International Labor and Working-Class

History 73:1-21.Zhang, Yunqiu. 1997. “From State Corporatism to Social Representation.”Pp. 124-148 in

Civil Society in China. Armonk, NY: M.E. Sharpe.Zhao, Dingxin. 1998. "Ecologies of Social Movements: Student Mobilization during the

1989 Prodemocracy Movement in Beijing." American Journal of Sociology

103:1493-1529.—. 2001. The Power of Tiananmen: State-Society Relations and the 1989 Beijing Student

Movement. Chicago: University of Chicago Press.Zhao, Minghua and Theo Nichols. 1996. "Management Control of Labour in State-

Owned Enterprises: Cases From the Textile Industry." The China Journal 36:1-21.

Zhou, Li-An. 2004. "The Incentive and Cooperation of Government Officials in thePolitical Tournaments: An Interpretation of the Prolonged Local Protectionismand Duplicative Investments in China." Economic Research Journal 6.

Zhu, Shengming. 2008. "zhejiang wenling hangye gongzi jiti xieshang diaocha(Investigation of Zhejiang's Wenling sectoral collective wage negotiation)."Guangdong xingzheng xueyuan xuebao 20:82-90.

Rupture and Representation: Migrant Workers, Unions and ... · 2/12/2011  · categorical ban on independent worker organization, a dilemma I refer to as insurgency ... primacy of - [PDF Document] (2024)


How were migrant workers affected by the Great Depression? ›

Labor exceeded jobs, which further reduced wages. Traveling from crop to crop, they lived in shantytowns, squalid camps, and primitive shelters—conditions that exacerbated discriminatory attitudes toward migrant workers, and added to social frictions and the trauma of dislocation.

How does Philippine law ensure access to healthcare and social security benefits for migrant workers and their families? ›

All migrant workers deployed by a recruitment or manning agency is covered by a compulsory insurance policy which is secured at no cost to said worker. This is pursuant to Section 37–A of RA No. 8042, as amended by RA No. 10022, otherwise known as Migrant Workers and Overseas Filipinos Act of 1995.

Who is considered a migrant worker? ›

A migrant worker is a person who migrates within a home country or outside it to pursue work. Migrant workers usually do not have an intention to stay permanently in the country or region in which they work.

What is the difference between migrant workers and expats? ›

The major difference between the migrant workforce and those on assignments as expats is the permanence of their move. Generally, expats are temporary workers, whereas immigrants intend to remain in another country indefinitely and may even eventually seek out citizenship.

What were some of the struggles that migrant workers faced? ›

Migrant workers lacked educational opportunities for their children, lived in poverty and terrible housing conditions, and faced discrimination and violence when they sought fair treatment. Attempts to organize workers into unions were violently suppressed.

What problems did farmers migrant workers face during the Great Depression? ›

Farmers faced many problems during the Great Depression, such as dust storms, a surplus of crops, and a lack of electricity in rural areas. The New Deal provided solutions for each problem. The Agricultural Adjustment Act sought to raise the low crop prices by lowering production.

Do migrant workers still exist today? ›

In fiscal year 2021, roughly 258,000 H-2A workers came to the US, up from about 50,000 in 2007. Others attempt an unauthorized crossing of the US-Mexico border, in order to secure seasonal work. Some cross the border seeking asylum, and work at farm jobs while waiting for their immigration hearing.

How have migrant workers been mistreated? ›

Unequal access to employment rights, remuneration, social security, trade union rights, employment taxes or access to legal proceedings and remediation; and. Workplace racism or discrimination.

Which state has the most migrant workers? ›

Which U.S. states and cities have the largest numbers of immigrants? The U.S. states with the most immigrants in 2022 were California (10.4 million), Texas (5.2 million), Florida (4.8 million), New York (4.5 million), and New Jersey (2.2 million).

Is an expat still a US citizen? ›

A Certificate of Loss of Nationality is prepared by the diplomatic or consular officer and submitted to the Department of State. Unless and until the Department of State approves the expatriation, the expat will remain a U.S. citizen, and will be subject to U.S. tax on worldwide assets.

What qualifies you as an expat? ›

An expatriate is somebody who has left their country of origin to reside in another country. Expats may leave home for work reasons and seek more lucrative employment in other countries. Expatriates may live overseas for a while or completely renounce their citizenship in one country in favor of another.

What nationalities are migrant workers? ›

From an approved source country or region, including Bangladesh, Cambodia, Hong Kong, India, Indonesia, Macau, Malaysia, Myanmar, Philippines, South Korea, Sri Lanka, Taiwan and Thailand.

How did the Great Depression impact the Great Migration? ›

The migration only came to a halt with the start of the Great Depression. The severe economic downturn dried up virtually all employment opportunities in the North. Conditions for all Americans would not improve until the start of President Franklin D. Roosevelt's New Deal.

What problems did migrants face during the Depression? ›

During the Great Depression, migrants faced numerous challenges and difficulties. These included extreme poverty, hunger, and homelessness, as well as discrimination and hostility from locals. Many migrants had to travel long distances on foot or by hitchhiking, often with little more than the clothes on their backs.

How were migrant workers lives during the Great Depression what kind of hardships did they face of mice and men? ›

John Steinbeck's novel Of Mice and Men centers around two migrant workers, George and Lennie. Migrant workers, commonly referred to in the 1930's as hobos, moved from job to job. These men were essentially homeless, living hand to mouth and carrying their belongings in bindles slung across their backs.

What would a typical day look like for a migrant worker during the Great Depression? ›

The typical day for a migrant worker was hard. They moved around most of the day looking for jobs. The workers asked to stay at a home, but had to work for it by doing work. The workers had to do a job and once they were finished they could stay at the place for the night.


Top Articles
Latest Posts
Article information

Author: Nathanial Hackett

Last Updated:

Views: 6329

Rating: 4.1 / 5 (72 voted)

Reviews: 87% of readers found this page helpful

Author information

Name: Nathanial Hackett

Birthday: 1997-10-09

Address: Apt. 935 264 Abshire Canyon, South Nerissachester, NM 01800

Phone: +9752624861224

Job: Forward Technology Assistant

Hobby: Listening to music, Shopping, Vacation, Baton twirling, Flower arranging, Blacksmithing, Do it yourself

Introduction: My name is Nathanial Hackett, I am a lovely, curious, smiling, lively, thoughtful, courageous, lively person who loves writing and wants to share my knowledge and understanding with you.